Calculus of IT - Season 3 Episode 4 - The Verification Economy (Part 1 of 2)

In Episode 4 of The Calculus of IT, Mike Crispin and I dive deep into what happens when "proof of X" becomes the fundamental currency of digital interaction. We're entering an era where verification matters more than the technology itself and IT leaders are on the front lines whether we're ready or not.

We explore the five pillars that need verification:

• Identity (are you who you say you are?)
• Humanity (are you even human?)
• Authority (are you authorized to do this?)
• Provenance (did you actually create this?)
• Expertise (do you actually know what the fuck you're talking about?)

From deepfakes and social engineering at industrial scale to the coming wave of AI agents that need their own credentials, we break down why your company's verification gap is probably bigger than you think and what you need to do about it.

Fair warning: This is Part 1 of a two-part episode. We got so deep into the weeds (in the best way) that we had to split it. Next week, we'll tackle IT's specific role in all this chaos, the impossible balance between friction and security, and whether continuous verification is our future or our nightmare.

Mike drops his latest prediction: We're moving from a governance economy to a verification economy. Instead of asking permission upfront, we'll increasingly ask forgiveness afterward with an army of human verifiers cleaning up the mess. He even coins two new roles you'll see in 2028: Chief Truth Officer and Model Auditor (complete with an otter mascot, obviously).

Key moments you won't want to miss:

• The safe word strategy we implemented at Xilio (and why you need one too)
• Why 60% of LinkedIn accounts are now fake—and what that means for professional networking
• The uncomfortable question: If an employee uses AI to create 90% of their deliverable, did they create it? And do you care if it's good?
• Why verification fatigue is about to become your biggest user experience problem
• Sam Altman's ben-wah ball crypto solution to solve the world's identity

This episode will make you question everything about how you onboard employees, verify identities, and trust the content flowing through your systems. Because whether you're ready or not, we're building systems where proving you're real is becoming harder than faking it.

Listen now, and join us next week for Part 2 where we tackle the really hard questions about IT's role in making verification work without turning your organization into a surveillance state.

Transcript

Season 3 - Episode 4 - Final - Audio Only
===

Nate McBride: [00:00:00] Sure on, on that, on that Zoom. And so when it comes to getting like a new employee on site, like what, I don't know, I guess the question is what are we gonna do? And um, I maybe that's something we're gonna get into. Oh, well I know it's not something we're gonna get into, but um, like how do you, how do you ever verify somebody if you never met them and you only have their voice, I guess 

Mike Crispin: is the question.

Yeah, I think that, I think that's, that's the challenge and a number of things. We implicitly trust a number of things based on reference and reputation, and we really do take a shot in the dark. I think sometimes we get a, we go with our gut, Hey, we try and if people interview well and they have good examples of accomplishments and examples of how they got there and all those good things, anybody can put that together now.

Nate McBride: Well before, before we started, I was, I was trying to find that article that came out. It was either [00:01:00] last year or earlier this year about the North Koreans that had been applying and getting jobs for remote work. Um, just using some basic social engineering, uh, along with generative ai. And I don't know who wrote it, I can't put the time in place, so don't hold me to it.

But there was an article about this, uh, this event, and it occurred to me too, that nefarious individuals could theoretically get a job at a company, um, up to a certain level. Never actually having to be the employee and be there. Yeah. Or actually ever to meet anybody or entirely spoofing somebody else.

Um, sure. And, and literally going the whole nine yards. I don't think it's, uh, it might be, it might be, it might not be, might not be plausible as an approach cost-wise, effort wise, but I think theoretically it's now more possible than ever. So I was thinking about that. It definitely 

Mike Crispin: is. And I think that [00:02:00] the big component of this verification economy discussion is going to be how valuable verification is and how it will actually be a type of work that we have, as opposed to Yeah.

I think it'll be, I think there'll be two categories. We'll talk about it, but I, I've got a lot of thoughts on this. Uh, 

Nate McBride: yeah. And I was also thinking about two things. One is, you know how we get, we already get fatigued from like an eight character password or something like that. Like how soon will till we get fatigued with having to verify everybody and pretty soon You don't 

Mike Crispin: have to.

Nate McBride: Okay. You 

Mike Crispin: don't have to. I, I don't think you'll have to, I don't think you'll care as much about the people as you do. The data is, the data's good. Yeah. If you can verify the data, who cares who makes it. And I think it's gonna be interesting. We'll talk about more about it. I, I, I think. More about the results is gonna be what humans take part in as opposed to the, [00:03:00] the, uh, there will be creation.

It'll just be, there'll be a spectrum, I think, of what we do every day and how we still will need to work. And perhaps there'll be, need more, need more use for humans. Um, yeah. Or more human work and human in the loop type thinking. I think it will get popularized more and more as, as these new roles get created.

Well, so if you're already living in a, well, well think, uh, sorry to interrupt, but think about, think about this now. We, we trust people, you know, we trust people and oh, we, we, we trust we and we can very, I can verify that you are you and you can verify that I'm me. But if I'm being dishonest and I'm not doing what you say I should do, who cares who I am, or if I'm doing an excellent job, even to some extent, and you're worried about what the result is gonna be, and you can verify how well it was done and where it came from, then who cares who I am.

Um, I know it's 

Nate McBride: out there. By virtue of you being, [00:04:00] being talented doesn't mean that you're on the right side of the fence when it comes to the moral question about what it is you're, you're trying to do. Thanks. Forget about the 

Mike Crispin: moral question for a minute, and the result is, what I'm trying to get at is if that, if that's all based on what you are trying to achieve as a business.

You, you're, you're concerned about. Yes. Yeah. Maybe that's the moral component is important. It certainly is, but the effort and the way that we got to the result is going to be important. And you're, and we're going to need, be able to, need to verify that. And whether it comes from a, a person or it comes from a machine, it's the process in which we verify it.

I think we'll still need humans to do that. And, and we'll need to work, work through that. And it'll continue, it'll iterate and constantly change, which is why I think it will keep the, the human interaction part of our work ethic going forward. It's not gonna be like this complete [00:05:00] overrun of all these machines and don't be Yeah.

Think about the, 

Nate McBride: go ahead. Go 

Mike Crispin: ahead. Sorry. 

Nate McBride: I was saying, you're talking about the providence, the providence question. Um, sure. Which I want to get to. I mean, basically I'm thinking about this like almost 2026. Um, let's suppose somewhere in late 26, early 27, I get a. An email from my CO. Yep. That's, you know, I know it's actually her email address.

I'm able to prove it with instructions to do a thing, but she never, she didn't actually have to have sent it. We already let EAs send on behalf. But what if in fact, um, there was enough enough of a, a control mechanism in place that when I emailed her, that thing was designed to me, email me back with her approval automatically if certain conditions were met?

Sure. Now that's not a far farfetched idea, but not at all. No, but here's the fee then, here, let's take [00:06:00] it one more year down the road. Maybe 20 27, 20 28. Um, what if your CEO's agent, like has legitimate authority to do these things on behalf of the CEO o and what is legitimate authority even mean? Mm-hmm.

So, um, yeah. You know, the idea of digital twinning is catching on quite well over the last sort of 18 months and, and more than ever, it's getting much easier to design your twin and put it into effect. And so if my, if my person in my company says, I have this twin who's gonna answer my questions on my behalf so I can focus on their work, is it that person's authority?

That is, is that all that matters? Does 

Mike Crispin: there 

Nate McBride: need to be a higher No, 

Mike Crispin: I think, I, I, I think if you're still in the construct where the CEO is the, is the person who needs to approve everything, which I think will very much be the case, human or not human. Is, uh, you need to verify that that's the, the right person in that case, certainly.

And that, [00:07:00] that, that whoever's making that decision has the ability and the authority to do so. And I don't know, I think, uh, agent based approval says there's gonna be some traceability. And this is where the verification component continues to come in, is we're gonna need to have proof. And that's sort of things I want to, I want to talk about is, you know, ultimately right now our work product is, is artifacts and, you know, work product and reports and, you know, product.

And I think the biggest product of humans will be proofs. Yes. And as time goes forward, we will be proofers, we'll be verifiers or we will be dreamers. And there'll be two buckets. There'll be your, your dreamers, your dream staters, and there'll be your verifiers and there'll be a, a spectrum between those and with a, a large amount of, um, AI and human based think that's determining those two types of, [00:08:00] of work groups over time right now.

I mean, I think there's like your creative mind. I can state it, I can, I can dream stream right into a AI from, by typing right now or uploading a few different modes. Holy shit. Did you just say dream stream, dream streaming? That's who I'm calling it. Yeah. So dream stream, dream state, however you want to put it.

Dream state, kind of, you know, talking and typing dream stream if you're go, if you're digitally connected and yeah, on sci-fi here, but, but that's sort of the creative idea, ideation side that humans will I still have a part in. Then a large amount of us will be about, and some of, we're doing some of this today pr proving that things are real, proving that, uh, approving things.

It's kind of like what we do today. Uh, I approve that you can take a day off. I approve that this project is good. I approve that we can spend this money. I, well hold on that you are who you are. Well hold about appro. 

Nate McBride: Approving [00:09:00] and approving are two different things. Uh, if I, I go ahead and I approve your vacation.

No, that's fine. But I need to be able to have proof that I'm capable of for proving that verification. So there's, yeah, there it's semantics. But there, there, I think that there're two different things. Uh, I mean, sure, I was talking about this yesterday, which is why top of mind, but, but you, 

Mike Crispin: you're, you're proofing something and by proofing it, you are approving that it's good and that it works and that you're verifying it and you're, you're approving, you're gonna be asked to approve that the, the product is good or the product is truthful.

And, uh, it's gonna be, I think it's gonna be interesting. 'cause the question is how do we get to that point? How do we actually prove anything? And that's where all these jobs are gonna come from. I think as a, you know, a chief truth officer or whatever you wanna call it, shut the fuck up. Chief Truth Officer, somebody who's, who's focused on verification because it's gonna be, we, in some [00:10:00] respects, we don't do this well now, right?

We don't, we don't, we, we, there's implicit trust in some things that we do and, and, and not in others. Um, take software. You know, we, we trust that search results are good, or we trust that the expertise that we read in a book is correct, or we trust that just 'cause it's an expert and someone says they are, we, we, and they have a reputation, right?

And we trust that reputation. I think it'll be more mechanical and more, more efficient as that's what we focus on more and more as time goes on, is being a verifier being. And it'll be all different types of jobs that, that do these verifications. And there'll be technologies that handle this largely, sort of Web3, probably technologies of some sort now where we're talking about tokenizing, verifying, uh, verification and rewarding people for providing proofs.

And, and it's out there. But I think there's, there's a lot more to [00:11:00] this, this mean more and more crap. More and more slop, more and more AI created content. Some very, very, very good and reputable AI created content as well. That someone needs to look at it and say so, and actually verify it at will we, will we put the effort in to do that?

I don't know. I don't know if we will or not. It looks good enough to me. I mean, how many policies have you released? And people are like, yeah, I read the policy. It's, it's great. Uh, you know, it's, it's very important that they read it and understand it, but oh yeah, you, you look at the log inside the server for your cybersecurity training.

They looked at it for eight seconds and they signed it. You know, so do we verify the stuff that we read that's important? Human nature says no. And will we be forced to do more of that as time goes on? I think we will. Because, because these things are being created. This objects are being created by, [00:12:00] we don't even know where they're coming from, 

Nate McBride: and we're gonna need to verify those things.

So, uh, I wrote down some notes while you were talking. I mean, content creation is not something we need. No one has come, like in the last three years, people weren't sitting around saying, you know, what the fuck? We need more content. Yeah. So, so the fact that Gen AI can create, it's making trillions of dollars, I would say people love it.

Yeah. But you can't just manifest the need to create more content, only because the platform exists to do exactly that. And that's what the created. 

Mike Crispin: What's that? No, I don't, I don't think that's why it's getting created. I think people want to create stuff as bad as it is. Some of it. Uh, you think the company's Yeah, I think, I think 

Nate McBride: I, I think, I think there's an incentive now on a lot of, a lot of levels a ranging from proving I'm more intelligent than I actually am to just Sure.

Uh, to answering someone's request for work to doing a homework assignment, like people are using it to generate content. The problem is, or the, the area that I have difficulty [00:13:00] with is that no one ever said. We need a technology that helps us create more content. Agree. Now we, we need technologies that create better content or improve our content, but not like there's not quantitative.

Yes, we're filling, I, I hear what you're saying, but, but they, they would have you think differently,

Trance Bot: the calculus of it.

Season three,

verifying this identity.

Speaker 8: Sometimes you just have to take it.

Sometimes you just have to take it

Trance Bot: [00:14:00] because it's season three divided autonomy,

verifying identity.

The calculus of it.

Nate McBride: Let's pause right there real quick. Welcome to the calculus of it everybody. Welcome. Uh, we're back 

season, season three, 

three, episode four. Yes. Uh, I am here with Mike Crispin. My name is Nate [00:15:00] McBride. Um, I think we haven't yet to announce our show in any episode, although maybe we did. I can't remember. Um, I guess it's nice 

Mike Crispin: title though on the podcast too.

I think we have to do 

Nate McBride: a little bit better about saying what we're actually, what we actually do, so, and like why we're here. Because like, just what we always do is just get right into it and start yelling at each other. Um, that's pretty crazy how that does that. So 

Mike Crispin: I'm so glad we're not on video anymore so I can, you know, wear all different types of clothes and hats and stuff.

Is that a skin suit that you have on? 

Nate McBride: Dude, come on. 

Mike Crispin: You don't have to tell 

Nate McBride: everyone. So this podcast is about my Crispin's Dream Dream. Um, 

Mike Crispin: I'm dream streaming tonight. I'm doing it right now. He's dream streaming 

Nate McBride: tonight from his dreams. He's, he is asleep and what we're getting is a link up from, uh, his brain.

I'm a 

Mike Crispin: dream streamer, 

Nate McBride: Mike. I'm a, I'm a dream weaver man. That's what it should be. We get 

Mike Crispin: Gary Wright 

Nate McBride: in the house. Dream weaver. One of [00:16:00] Mike's dreams is we've just recently found out is to be a chief truth officer. Um, yeah. Down the road. So he's building up his resume. I'm being a chief truth officer. I got another 

Mike Crispin: one for you.

I wrote down, this is awesome. Hold on. I gotta fight it. Oh, there's another, I thought I, I, I had a big, is it Chief? Chief Asshole? Uh, chief Dink, I think is the, is the one I had? No, um, let's see. Oh God, if it doesn't have the word digital in it, I'm not buying it. Oh, no. It, it is awesome. I I was like, I, I wrote it down.

I'm like, I put a smiley face next to it and I'm like, I gotta just, oh, oh. This is it. Model auditor.

Model auditor. They'll replace all compliance roles, chief truth officers and model auditors. Okay. Well, and the, and the model. And the [00:17:00] model model auditors will have, uh, an emblem of an otter on it, like a, like this, the penguin for Lennox. There'll be an otter for model auditors. 

Nate McBride: Got it. So, um, so manscape.com, I mean, when you're hiring your chief truth officer for the first time on, there's a 

Mike Crispin: lot to verify when you're manscaping.

You know what I'm saying? You gotta check, you gotta make sure everything is done right and verify. And that's where the verification economy comes in, is that's it. And that's why man, the manser or mans, what do they call manscaped.com? Whatever, whatever the fuck they're called, they, they, they need, they need a chief truth officer and a model auditor.

That's what 

Nate McBride: they need. 

Mike Crispin: Well, 

Nate McBride: you know what it was, uh, for many years people didn't think that there would be a Chief Digital Officer because that's so, it's so absurd. But then they, they started to [00:18:00] become Chief Digital Officers. I love it. No knocking on the Chief digital officers that are out there, but really, I mean, come on.

What does that even mean? Um, we're 

Mike Crispin: talking about, you know, permissionless innovation, permissionless innovation, and ruthless verification. That's where, wow, that's my new t-shirt. It's a lot of letters. 

Nate McBride: Where, where are the dream stream folks? And you're picking up on Mike's dream waves. Um, we have, we are going to be talking about the verification economy tonight.

Um. Big, a big idea. Uh, we'll do our best to get through this tonight. May have to break it into two episodes. We'll see how it goes. But really, um, the, as we were talking before we kicked it off here, we were talking about proof and proof of anything at this point. Like proof of X, let's call it. If proof of X is the currency of digital interaction, [00:19:00] then how do you back up from me even there, how do you establish the proof to begin with?

Um, it's not like saying like, every time I see Mike's face, that's the proof because obviously we know now that that can be not the proof. Um, there needs to be, there needs to be multiple levels of linking things together to create that proof of X. Yes, we're gonna cover cover identity verification, which we've been talking about since last season.

And that's basically central thesis of this season to an extent. Are you who you say you are? Um, and it's not so easy to, to do that anymore. And then humanity verification. How do we verify people are humans? Uh, are we asking? Good question. 

Mike Crispin: Especially sales reps. 

Nate McBride: Yeah.

Mike Crispin: Yes. Hey, can we, hey, can we handle this conversation over email? 

Nate McBride: Is that okay with you? What I wanna, what I [00:20:00] wanna know is how come the, how come this, the spam marketing companies haven't figured out that if the email address starts with someone's first name. It's not real. Like, don't put Mike at shitty a spam services.com.

At least try and use a conventional email structure. But no, as soon as you've see an email with someone's first name, just delete it. Unless it's from me. If it's from nathan@longwon.consulting, then, or, or the Coit Us, which is our, our website, um, by the way, then you can answer that Peter North at the Coit us.

Yes, exactly. Well, we don't use, we don't, we don't use the DOT convention. Oh. Oh, we don't. Okay. No, we do, we do.plus equals. I like that. I like that. 

Mike Crispin: It's catching on. It really is. I I, I do. I like that. ronaldMcDonald@mcdonald.com. That's another good one I've used 

Nate McBride: before. [00:21:00] I bet you nobody's, nobody's done that one yet.

We talked about, so we talked about this just before we started, which was, um, who's actually authorized to do something and, and where does authorization come from? Does, is there only one person in the whole company that's basically capable of authorizing anybody and not being the CEO? Does it work from the middle out, top down only.

And I talked about this in the, uh, autonomy paradox book too. Like who, who watches the watchers, basically. So how is authority given to even establish proof? Then we talk about provenance. 'cause Mike was touching on this, but did you actually create this thing and how can you verify you actually created it even more than that, like.

If I use Claude for 25% of my document, basically gimme the skeleton, gimme some, you know, tree of thought, three ideas. I'll do the rest of the work. How much of the document is Claude versus me? And, and what's the provenance? Is it Nate and Claude? Just Nate or just Claude? And then the last, just Nate. I [00:22:00] think the last thing, and we talked about this last week, was do expertise verification.

Do you actually know what the fuck this is? Like, do you know what you're talking about? Or did you just read Claude five minutes before you came in the room? Like who are you the actual expert and how do you prove expertise, Mike? Um, yeah. And do you, do you need, do you need the expertise? Do you need to have the, do you need to have the expertise in order to be the expert?

Is that what you're asking? Do you need 

Mike Crispin: to have, does the, does the, does the, do you need to have the expertise? Is, does, is the expert required or is the expert verifier required? 

Nate McBride: Well, let me, let me, let me ask you this question then I'll spin it around for you because we, we talked about this, um, over the break.

Let's suppose I, um, I call plumber. I got a leak called Plumber. He comes to my house, he's got plumber on [00:23:00] the side of his van. His website says, I'm a plumber. He's guy says I'm a certified plumber. All the things right. He whips out his like plumber badge in his wall and shows it to me, and he sits down and looks at my sink and he whips out his phone, opens up YouTube, and starts googling how to solve the problem.

Um, he's still the expert. 

Mike Crispin: Yeah. No, and I think a lot of, I'm sure, I'm sure that happens. 

Nate McBride: Yeah. But yeah. But is he, is he, he's not the 

Mike Crispin: expert. 

Nate McBride: Is he an expert because he doesn't know, or is he an expert because he knows where to go to find the answer and then do that thing? Yeah. I mean, if he, if he, if he fixes your problem and it works, do you care?

Well, do you care that so much that you'd call back him as an expert? Uh, or that you would try to find another expert? I mean, let me ask you this another way. 

Mike Crispin: Uh, if I, if it was that easy, I might try and do it myself. If he, he just looked it up on YouTube. 

Nate McBride: Let's say that you then assume this plumber is an expert because, because by virtue of his Googling capabilities, he was able, and by the way, this could be a a she or they, I'm just [00:24:00] using he and the common vernacular.

But, uh, let's suppose this plumber, uh, looks at YouTube, gets the answer, solves a problem, and is great. Yeah. Is that, is that plumber an expert permanently? Forget the fact that they can, they can remember what to do or they an expert temporarily. And this is another element to that expertise verification.

Do I just need a temporary expert? Mm-hmm. Or do I need a, a permanent expert? And I 

Mike Crispin: think, I think that's part of the philosophical question here is are we are, are we gonna be so concerned with that at, at, you know, from, I'm just talking about from an objective perspective, but let's say some, uh, some robot comes in and fixes your, your sink.

Yeah. And you're, and it, it fixes it. Do do, in that case, do we care that it's an expert and that it knows what it is? Or that the, or that the plumber knows how to do it? Or did he sit and lose, watch YouTube? I mean, there's [00:25:00] still certifications and other things, right? But still you can get a certification, I imagine.

Well, or fake a certification. I've had con I, we, I think we've probably all dealt with contractors who, uh, working on our houses and things that we thought maybe had more experience in one area or another, and didn't, I mean, this, I, I think this goes, the discussion we're having is that if we can get better at verifying the experts, like you're saying, we can, we can go along, we can kind of get on that path of having a better process to do that.

I don't think we're great at it right now. And I think it's a concern that, you know, people get scammed every day. It's a huge issue. 

Nate McBride: Across the board. Well, there, there are, there are certain organizations that will give you an accreditation because you've passed an exam or something else. Right, right.

Sure. Like those people, I guess, uh, in the general cultural vernacular, because they're experts, because they've achieved a certain level of, of ex um, you know, uh, evaluated mastery of a thing. But yeah, definitely, [00:26:00] definitely. But, but even still, we've, like, we don't have to go all the way down this hole 'cause we will never get out.

But in terms of verifying expertise, we do have to have an answer of what expertise is. Is it, are you an expert for this one moment? 'cause you just watched a video and therefore I don't need to verify like the rest of your bonafides? Or do, do you, do I need to verify the fact that you are truly an expert, you're just being lazy, like maybe looking at a video like we do need to get, we need to touch on that a little bit.

Mike Crispin: I I, I, I'm only driving towards the results because I think if you hire a plumber you've never hired before, and they come in and they fix your problem right before Thanksgiving and they leave, if they fixed your problem and they didn't rip you off, or you didn't, don't know enough that they ripped you off, um, you call 'em again because the result was good.

I, I highly doubt that. Yeah. Most people will go and turn around and say, did you just watch YouTube and figure this out? Like, I, I don't think now if they broke it and they're sitting there and some filling all [00:27:00] over the floor, most people aren't gonna care. I, that's just my thought. 

Nate McBride: But it, it was, I think that was a runaway metaphor.

Let's get back to how it matters in, in terms of, I'm just, 

Mike Crispin: I'm talking about results. Results are what I think when you've got ais running around everywhere and people running around everywhere, people want to trace the result. Okay. I'm not, I'm not, I'm not, I'm not discrediting the need for verifying people's ex expertise.

I think that's important today, right now, hugely important, but it's, I think as we get these kind of result driven artifacts through systems that we don't even know how they've been created, we're gonna need to figure that out. People are gonna ask a lot of questions about that, because there's gonna be so much crap.

There's gonna be so much extra stuff, and we're gonna need to verify those things today. I think we rely on expert, we rely on expertise today, tomorrow is all what I'm trying to say. Is that tomorrow I think? No, we were, we'll spend a lot more time on the result in tracing it back. We, 

Nate McBride: we might rely on expertise, Mike, to a degree, but I think we now, we rely [00:28:00] more on who can give us the best closest right answer.

Uh, I don't know if people like I agree. I agree. Maybe, maybe, maybe, maybe in a trades craft type model. We are looking for experts to do this sort of big project for us, but I think on the whole, what we've come to condition ourselves to Yeah. Is I just need a really good answer. Like, I don't need a whole lot else.

And, and, and so Good, good enough, good enough, right? Yeah. I agree. We'll, we'll, we'll come, we'll come back to that. That's obviously gonna be a big one. But I think we have to, we about which is why proofs, which is why proofs are 

Mike Crispin: so like gonna be, if we get better at that because we, I think we both agree we have to, from an identity and up verification perspective, we have to get better at it.

That, hold on. But, but the, we there humans do. 

Nate McBride: Yeah, but it's, I think we have to break that down. I don't think you can just say humans. I think we have to break it down into, to groups like governments. Uh oh, okay. Like we have a, we have a, we have a government right now that is trying to enforce state regulations for [00:29:00] ai.

Um, and so they can, so they can control verification. We have platforms. Platforms. So right now to do a thing like to, to get into box and make a document, you need to have two different platforms working in harmony, uh, a an authentication platform and the actual platform itself, right? So there, there's, there's an authentication event that happens to them to say, oh, yes, he, Mike's asking for this.

Do you agree? Yes, Mike's can come in and do this. Okay, great. We're in agreement and I'll let Mike do it. Then there's third party services, and lastly, you have your cryptographic systems, you know, your, your biometric and your, um, you know, other sort of higher level au I mean e even like, I don't wanna get into crypto itself, but you have things that are cryptographically, um, sitting as the authenticator and then, but, but even with all those things, [00:30:00] I mean, I just, we were just talking about this a little bullet point, but is it the sort of final verifier of all things because.

It's chicken or the egg, right? I mean, it could be, it's technically the C-E-O-C-E-O verifies everyone else to make decisions. But are they, they're not the last wall of verification. 

Mike Crispin: No. No. I don't, I don't, I don't think it holds the key. I think this does some respects, that's why this does create, I think, more opportunity for, for jobs is that it's, there are gonna be verifiers in all different areas of expertise, or I don't think it's gonna be it.

Um, I think, we'll, we'll certainly have the ability to help with some of that, but that we're not the experts in a lot of these areas. So I think that's 

Nate McBride: there's, but we're expert, we're experts on proving, well, hold on. I'm not going lump everybody in this bucket, but we're generally, generally considered to be the experts on proving that [00:31:00] somebody is who they are when they authenticate to excess data.

I, 

Mike Crispin: okay. If you're talking about people and identity, yes. I think that Id, yeah, I mean that's will have a, when you're talking about verify, what do you think I was talking about? I think you're talking about verifying data, verifying truth, then that's different. Well, yeah, that 

Nate McBride: too, but, well, we have to talk about that too because we have, you start to get into identification and it and verification economy.

You now get into having to have trade-offs. You have trade offs on privacy. Mm-hmm. How much does Mike have to give about himself before I truly believe it's Mike? Uh. Are we talking like birthmarks on lower parts of the body type proof or like, what are we talking about? We have that, that would create the accessibility challenges.

So Mike needs to access this document and while he has to get four levels of verification, um, and this has been an age document. Yeah, that'll, that, that'll fail. Yep. User friction. I mean, users, [00:32:00] they're, they've already got enough. And then, then hold on. And this, 

Mike Crispin: Nate, this is why I'm saying like the whole, all this stuff we're talking about, like in terms of the, how we get to the result before we get to the result is, is too slow.

So you get to the result and then someone's gonna verify the result. Yeah. So it's almost like asking, it's almost like asking for forgiveness a job that that's all it is. Like you, you got, they're gonna, they're gonna, and in some respects we deal with this now it's like, oh, of course we do. We need governance, we need compliance, we need all these things.

And people just look at it and go, oh geez, it such slows us down. They don't understand why and they should understand why, but they don't. It's by, by default. We're 

Nate McBride: like the biggest creator of loopholes. Think of, this is my favorite loophole and I'm, I, I can't believe I'm gonna say this out loud 'cause if anyone actually listens to this, it'll be havoc.

But I hire somebody and I make 'em go through the hoops and hurdles to get access to a thing and they do. And then I hire somebody a couple months later who needs to work on that same [00:33:00] project, but they don't have the time to with through all the hoops and hurdles. So what do they do? Yep. They just have their friend.

Invite them because they're gonna get access anyway. That's right. And there's this gap in between. That's unsolvable. How, how about, how about 

Mike Crispin: we put it this way to, to your point, to your point, you know, there's, we talk a lot about governance and, and it's important today. A very, very important, and you, and we could argue that it's, you know, regulation of AI and whatnot is very important as well going forward.

But governance seems more like it's governance is asking for permission. It's proactive, right? I think that's too slow. When we start bringing all these other things into the case of certain, and I, I know it's controversial, but I'm just saying a AI and all these things that are coming and things going so fast and people just skipping the line to get things done, right or [00:34:00] wrong, and immoral or moral verification is asking for forgiveness and building the skill around being, being able to react and verify quickly.

It puts the gate at the end instead of in the middle. It's gonna create issues, it's gonna slow things down to some extent, but it will stop the very crucial things, the very crucial things that could cause real damage at the end, at the, at the gate, at the goalposts, as opposed to let's weave all this process and checkpoints in to make sure it never gets there.

'cause you know what? The people who want to get to the result. Again, morally or immorally, they're gonna skip all that shit, and they're gonna go right to the end. And we're gonna be bad at verifying things because we're trying to do all the stuff in the middle, and it's just gonna escape. Like, and that's what happens with a lot of things is I, I, I read the policy.

I I'm gonna do what I'm supposed to do and tell the shit hits the fan, and I need to get it done. I'm gonna [00:35:00] go around it. Now imagine you're running a massive project. You've spent millions of dollars of on, on ai. You've let go of half your company and you've drilled through all this stuff and, and say, I'm not saying again, I'm saying this is, could be potentially immoral.

You think that all the governance is going to protect and regulation is gonna protect in the middle. It's gonna have to stop gap at the end. And with an army of verifiers. And basically the verifiers are the only thing. And the human, they're really the, the mass human in the loop method to all the other machine learning and information that's coming down the road.

And that's why I, I get so kind of passionate about it because I feel like I, I, I feel like the only, it's, the only thing that's gonna save us from kind of eating ourselves is if we don't have this like, layer of verification that is inherently human. To some extent. You don't, we just don't want to be in a, a, a constant [00:36:00] hallucination loop until the end of the world.

To get dire, but like, we need to have this foundational but, uh, verification layer. But I think it needs to happen more. It sounds crazy, but more reactively than proactively because we're gonna get, it's just gonna get run over if 

Nate McBride: it's a gate. Well, you're just, we're, we're, what you're talking about is a major regression back to challenge response and, and so Yeah.

Kind of. Yeah. Yeah. So, so let me ask you this question, like in terms of why this all matters. I mean this, and this all of course does very much matter right now, I mean, between SOA two and how easy it's to generate audio. Yeah. And video bites the question that probably you haven't asked or you haven't considered today, or the thought you, you haven't probably haven't had this thought yet.

I've had it a few times, but I feel like that by the end of 2026 or maybe even sooner, [00:37:00] uh, every single thing I get, I will assume is fake. Until proven otherwise. I mean, I, I, I, now, as I now assume every single post on LinkedIn is fake. Mm-hmm. Um, I don't trust anybody unless it's somebody I implicitly trust that wouldn't do that.

On Substack, the people I follow, they write their own work. But there are people on Substack that you just know are absolutely fake. And then, yeah, sure. I think we, I think we have a period of time until you'll be walking around your company and even with Slack, even with that sort of, um, sec separate layer of verification, you're not gonna trust anyone's shit anymore.

And yeah, you're, it's a problem. You're gonna have to like, watch them write it in real time to verify that they actually did it. Um, and so again, I don't think this is widespread today. I can certainly, 

Mike Crispin: and, 

Nate McBride: and 

Mike Crispin: I agree and that's what I just wanna make sure is, you know, if anyone's listening [00:38:00] right, is we are talking about futures here.

Like this is, you know, getting rid of governance. Yeah. Not, not like, not like 20, 30 future. I'm talking like, but, but, but no, I mean, like, if you're talking, if I'm from, from what I was just talking about, about, you know, governance and this, you know, put that aside, move to the end of the, the line and be more verification from a reactive perspective, I think that's where it will end up, down the line.

Right now it's getting better at proving stuff, which I, you're, you're talking about right? Is you, we're, we're assuming everything that's getting written is either fake or just, you know, I, my, my, my, my thought is an I, you know, it is an idea that came from someone. They may have used a different tool to write it, that they had the idea, someone had to prompt the idea to come out and be written Well, whether it's not their, their creative language is lost, their unique writing is lost.

You can, you can still create with AI and [00:39:00] have it be unique. I mean, that can still happen, but if people are just using it to spit out the generic email Right. And then you lose a lot of personality and culture. I, I, you know, I completely agree. You there. 

Nate McBride: Meta just claimed, uh, I think 10 billion or so dollars in revenue strictly from, um, spam generated ads.

Google just recently released their estimate that 50% of web content will be AI generated by end of 26. Yeah, agreed. There are now, I think, six or seven academic paper mills that are, that are between them producing thousands of fake research papers monthly. Yeah. Um, not to mention the flood of fake books on Amazon.

You have everything that's in the news, social media, et cetera. They're all increasingly synthetic, so. Again, back to my point of what will you trust? Like this, is this purely AI generated content flooding? How, how will you find that one true signal in the noise? [00:40:00] Um, and we'll come back to that in a moment because it also has to Good do 

Mike Crispin: very good question.

Nate McBride: Yeah. It also has to do with social engineering. I mean, so we now know it takes, it takes three seconds of audio. Uh, from like anything, a TikTok video, I don't give a shit. Three seconds of audio is all it takes to create a CE Ovo CEO voice clone that's nearly perfect. Um, I mean, it used to be one scammer, one target, one attempt, good luck.

And now I wouldn't say it's unlimited, but you know, AI can research targets, it can craft personalized messages, it can respond in real time all. And so like, here's what we did, by the way, and we're about to do it. My company, which you should consider. So over Thanksgiving break, uh, when we were all around the table, I, I and the rest of the extended family created a safe word.

We created a safe word phrase. And so we all agreed to at the table [00:41:00] and if, and if that every, and a grandparent got called. Like, it's me and Nate. I'm, I need $500, I'm, I need for bail money. You would have a safe word to verify. It's a great 

Mike Crispin: thing. 

Nate McBride: Yeah. We're also, we're also doing this, uh, at my company, at the executive level, we're gonna have a safe word.

Um, we're also announcing at the bureau Learn Tomorrow the safe word you should have for making any kind of decisions. That is a way, I mean, it's basically a multifactor authentication event, but it is a way to bypass that social engineering because it's going to come, come now in big waves. Uh, I mean,

uh, if you think about, think about bots historically have been malware bots, right? Malware bots, yeah. Doing all kinds of terrible shit. Um, think about the companies that have, you know, 3000 GPTs or whatever, right? They're basically creating bots and, [00:42:00] and what's one thing that you can do with bots is you can, I mean, again, this is to your point a little bit down the road, but you have bots, um, written on the same platform.

What's sort, what's the, what's the difficulty in potentially getting those bots to swarm? What's the potential, potential problem in having, um, bots that underperform and need to be shut down or quote unquote fired? So, and what does all, what does all this stuff lead to? I think it just leads to. I mean, it's sad, it's very, very difficult to say sometimes, but it leads to almost a complete dissolution of trust.

I mean, yes, absolutely. On Twitter, I think, was it eight bucks to get a check 

Mike Crispin: mark 30? Would you, would you argue? In some respects, it's already happened. I mean, nobody knows what to believe even before ai, well, I can't even, even, even worse and worse, 

Nate McBride: [00:43:00] LinkedIn, I can't even do, 60% of LinkedIn accounts are now known as fake.

60% are fake accounts and you can, it's easy to find out who they are, you know, just see how much shit they like. And, but that means I gotta take the time to do that. Right. Um, and reviews, I mean, basically there are whole review farms sitting out in unfortunately poorer countries that are just churning out reviews for the Yelps and Googles, and they're, I mean, these companies are supporting these fake review farms.

Mm-hmm. So, to your point about proof, I would repost with trust, uh, and we should, we should kind of get into those, I think two, the core. Yeah. I think, uh, 

Mike Crispin: proofs are just artifacts that we can use to trust people or trust information, I think. And the proofs are. Just showing, and this is where, I [00:44:00] don't know how this happens or how to do this.

Yeah. But it is just showing, it's giving people something they can look at and go, okay, that's correct. Now, is that some sort of verification stamp? Is it a, is it walking through some sort of equation that proves that, I mean, it's just like, is it that type of level of information? And, and that's, and to, to me, like when, when those things are presented, I feel that we need to globally, and this is where I think Web3 and other components of sort of, uh, blockchain and stuff come in.

I think people who are able to do those proofs in a way that can be consumed by others, should be rewarded with a token of some sort. That they, as they begin to do this and they're proven, and that because those tokens can be verified by other people who agree that that proof is correct. [00:45:00] Other humans, they're, they're, they're rewarded with a global coin.

They're rewarded with a global Yeah. Stackable reward that, and they can show that they have this credibility in this specific area. Now, I don't know how you get to that, or you know, how you prove your excellence so that you know what you're doing or whatnot, but. Or, or that, do you have the skill to proof prove these things correctly?

But it's almost retroactive as AI comes up with solutions, as people build new products, or we're trying to solve major problems with these tools, someone's gonna go in and try and pick it apart and understand how they got to where they got, even though the drug might already be on the market, or the solution has already been released, or the decision has already been made.

If someone has to retroactively put proofs against that decision and be rewarded for it so that over time the decision that was made, even though it's after the fact and the damage has [00:46:00] already been done, or it's been hugely bene benevolent, that we have a way to verify that what we did was the right thing.

So we can learn from that and do it again, or not 

Nate McBride: do it again. Okay. So two things that we have to then talk about. One is weariness. If I have to do a capcha at every step of my authentication path, I'm going to quit, uh, or find a way to circumvent it. So, so there's that. But then, then we, I could, and I, I did write about this in the paradox book, um, how do you, like, why don't we continuously verify and, and how do you even do that in a way that is passive or least intrusive?

You can't, if Mike logs in, I can set a timer like eight hours from now, he has to re verify himself. Right. I'm gonna probably hear it from Mike. 'cause eight hours is, is too short maybe. Yeah. Uh, so let's say I, let's just say I leave it at eight hours though, right? Okay. So there's eight hour, eight hours of an open verification [00:47:00] session that somebody else could just slip into.

Now it's, you know, the vendors will tell you it's encrypted. Um, there's evidence to prove that these sessions are encrypted, but that means almost nothing. So, especially if the person already has, um, awareness and, uh, has taken hold of a internal place inside that, inside that session. So, I mean, how do you continuously verify identity versus just one time login and, and it is even worth doing.

But, but before you answer that question. Sure. Um, and by the way, we'll talk about the whole password's. Dead dying. In episode 39, and we are gonna take some time off of that for a while. We'll come back to it, but, okay. A new employee. All right, so here's the scenario for you. New employee starts a Monday and they're fully remote.

Somehow this person just worked it out, right? So first of all, how do you verify who they are? How do you verify that who they are is who they claim to be? That's one right? [00:48:00] Then. Then secondly, uh, the same employee calls the help desk to reset Okta. How do you verify it's that really them? And then third one is, contractor has access to sensitive data.

How often do you reverify? They're still the same person. Now, I'll answer this from my side first, which is to say if a new employee started Monday and was fully remote, my verification is a hundred percent reliant that HR has, um, checked everything they've done. Yeah, zoom, zoom calls where they verify the person's a human, they've got a driver's license and a passport, and all the other things that they need to verify This person, preferably even references from people that they physically can reach out to.

But, but I, I, I, as it right now, today, I'm a hundred percent trusting and upstream process for Yeah, I hear you. For the, the employee who calls the help desk, well, we don't have a number for our help desk, so. [00:49:00] It wouldn't work that way. But, um, yeah. When, when, whenever an employee calls, whenever an employee has a problem with resetting MFA, we will, um, ask them to tell us about what's going on, like what's the issue, and then all we will do is reset the one token that they need.

Um, yeah, they we're not giving them a password, we don't even know it. And then with the contractors, we check those every 30 days. Yep. We could check them more, but we have a pretty good sense of what they're involved in. And project wise, if we were way bigger as a company, I think it'd have to be a more frequent verification.

But we do 30 days. 

Mike Crispin: Yeah. I think, I think, uh, uh, verifying credentials and authentication is, is one thing, like you said, the two factor, not resetting the password or resetting the fact one of the factors so that they can get in. Makes sense. I think verifying who someone is, you were talking your first, your first question, uh, [00:50:00] for that you had mentioned.

I mean, they're starting to emerge more so in the consumer space than in our space in the enterprise. But, you know, people having to send in their ID or, you know, automating the process. Like DocuSign has a product in which you have to verify with a real id, you know, who you are and to sign for things and that type of thing.

It's all a pretty, uh, you know, straightforward process. It wouldn't surprise me if. Like the tech, like technologies like at the airport, like, uh, clear for example, has a more stringent enterprise product that is released more widespread for these type of things. It's just, this is what I meant by the verification economy.

Is there, there's like 40 different products integrate with Okta to try and verify you with a pin or, yeah, yeah. You know, for call centers for all these things and they've all popped up in the last year or two and it's, it's, it's all over the place. This problem that, you know, we're both discussing about making sure that [00:51:00] we can trust somebody or something and it's already emerging.

And with all these AI things, it's getting more, uh, more intricate. So I think trying to prove who someone is, is no, is more difficult now than ever because we have more contractors and consultants that work at companies temporarily that often do not come through an HR process, uh, that are always urgent, need to come on board so quickly.

It's an urgent situation. You don't know who they are, who they work for, do they have a contract in place, do they have a po? Uh, what's the, do they use, do we give them a computer? Do we not give them a computer? Now you can have a rule and set and policy for all these things. Which, which we do, you know, we have a, we go through, we make sure there's a contract, we make sure they have a, a po, we make sure their, their, their company is on the qualified supplier list.

You know, all the stuff that we try and do to [00:52:00] keep that going. But there's no saying that their identity can't be stolen or that these people, like you said, may not be experts and just have a great CV and interviewed well and a great social, social engineers. Um, you know, so that piece is, is missing.

We're not good at verifying. We trust based on a lot of times how well people socially and can, can connect and explain what they've done and their accolades. Just like interviewing people. Yeah. You know, we say, Hey, we made a great hire. We made a bad hire. Um, a lot of it is 'cause it's very hard to verify what you hear is gonna be good until the person actually joins the company and you've opened the doors to more risk.

So that's, I I I'm surprised ver this kind of verification stuff hasn't just become a, a bigger buzzword because it's a huge, it's a huge issue [00:53:00] Well think about. So, and it's now that we, it's now that we're working with machines that, that we're getting concerned about it because we see how fast it can produce stuff.

And what it can produce is sometimes good enough and sometimes incredibly wrong, but looks great. Um, yeah, so it's, I think it's 

Nate McBride: raised the alarm bells. Think of, think about what we've developed for verification. We developed, we have, we have capcha, which mm-hmm. I mean, capcha iss, if it's not dead, it's certainly dying at this point.

I mean, Google still uses it, which is funny. But, uh, attackable too. Yeah. And then we have behavioral analysis. So how you type your mouse movements. Um, you have your sort of geolocation geo, uh, transition, um, yeah. Models where, you know, you were here, now you're there. That's a typical pattern. You have your biometrics.

Yep. Uh, which is an emer still emerging, despite the fact that we've had fingers for a while. And, and, you know, some laptops will do face and thermal. [00:54:00] Uh, I, it's still not a hundred percent, but there's privacy concerns with that. You have device fingerprinting, easily spoof, like Mac IDs, no big deal. And then you have proof of work, which is fricking expensive.

Like, if you're gonna do a blockchain proof of work model like you suggested earlier, uh, you gonna have like, you gonna have some resources, some compute for that. So, you 

Mike Crispin: know, I don't want to talk about the specifics here. You, you know what I'm gonna mention. Yeah, yeah, yeah. But, uh, um, but, you know, I, I'm working on something that I think is very relevant in this space, and I would just say that there's really only one, one way that I feel like we can.

Tackle this in the future, and that is by, uh, verification of multiple peers or verification of multiple connections. 

Nate McBride: Yeah. 

Mike Crispin: That is a constant, is a constant reverification of a large, crowdsourced, [00:55:00] crowdsourced network of people that basically says you are who you are. And it's constantly living and breathing and updating itself as, as time goes on.

And because we just have no way of verifying people except for if you're in the room together. 

Nate McBride: But the paradox, Mike, is that we, we acknowledge we need to verify humanity. Like we get that part a human being needs to be verified, but we're also building, well, we're starting to build gen AI agents that we want to act autonomously.

We want them, I mean, this is more in the machine learning realm, I suppose, but we want them to act autonomously. So we need to verify humanity for some things. We need to accept AI for others. But here's the real problem, which is how do you build a, a system that can distinguish between the two? So yeah.

Great. I think we'll get, we'll make, we'll make great progress. We'll make progress in verifying humans. We'll make progress in verifying ai, [00:56:00] but how do we make progress into finding something that can, on the fly, distinguish between which is, which that will be, because there's no standards, right? There has, there have to be some standards around that.

Um, that's the part that's uncomfortable. 

Mike Crispin: And I, I think that the very high level question there over time will be a, as results and output from these AI agents and these autonomous agents get better, will we care? Well, if I get, I 

Nate McBride: think I, I, I wanna know, Mike, if I, if I, the message I got came from Mike or came from Mike's AI system in Slack, like, I wanna know.

Mike Crispin: You, you do, you do. I'm, I'm saying today you do, and, and today, and you may always, because you, you have lived through this change. I'm, I'm saying as time goes forward, if I, I mean, there's probably a time where people had calculators and they're like, I'm writing [00:57:00] out that math. I, I don't trust that, that that calculator's gonna get that right every time and I'm writing it up.

So I think that as, as like the, the younger generations, if they're getting good answers from this stuff and it's good and it's not breaking things, 

Nate McBride: and 

Mike Crispin: it may not even completely Right. 

Nate McBride: An answer might be one thing, Mike. But I said like, if I need, Hey, I need to schedule time on Mike's calendar, and your agent responds, great, but hey Mike, I need you to approve this a hundred thousand dollars po.

I, I'm gonna, I'm gonna say no, I just gimme the real mike. Like there Yeah, there will be, there will be lines of things we will not cross, hopefully. And, 

Mike Crispin: and I think, and I'm just playing devil's advocate here as I like to do Yeah. That's why we we're here on, on the podcast is, is there are people who sign very important documents on behalf of other people, uh, administrative assistants and other groups.

Sure. Um, and that just sign off on stuff. Um, and it's not the real person. [00:58:00] And they may find they've delegated the authority, um, because that's their decision to make. So I, I do think I, I, I care as well. I, I just see as there are examples today where we sort of implicitly trust if that signature comes back, then we can move on to the next thing.

Okay. I got what I needed. I can move on to the next thing. And I'm not asking. I, I'm personally asking, did, did, did my boss sign this or did their twins sign this? Um, onto the next thing. So, and I think a lot of people will be that way with some of these responses they're getting from ai. Like, I mean, it's, I I'm not saying it's right, but, you know, I, people using AI to write very personal letters to people, it's not even them, you know?

Um, because people are lazy and it's, but Right. So I mean, that's, that's kinda, if you look at the trajectory of the math of people, I give it, I give it three 

Nate McBride: years. Three years. You'll [00:59:00] have employees who use AI to attend meetings for them. You'll have employees use ai, respond to emails and complete tasks.

Absolutely. And I guess at some point we, our future HR departments will have to ask the question, who are they employing? The human or they employing the ai? And then how do you, again, back to the verification that, like how do you distinguish which one you're talking to at any point in time? 

Mike Crispin: Or, or, or how do you, maybe even more so, how do you distinguish what's been created?

Is, is real, is truth, is, is, is work product that you need and that's, well, your, your 

Nate McBride: chief truth officer will have to hire, um, don't you verifiers model, uh, come on. You know, it. Um, model auditors, model auditors, model auditors, and truth, truth, truthiness quotient, um, developers. I, 

Mike Crispin: I was terribly kidding 

Nate McBride: about the Chief 

Mike Crispin: Truth Officer.

I thought you would love that. [01:00:00] 

Nate McBride: No, I, I do love it because it's, uh, it's you, because you've just said it. You have literally manifested that into existence right now. It's real. It's real. There's a co there's a company that just made a Chief Truth Officer because they're concerned about their Chief Data AI officer who's going off the rails, honey.

Mike Crispin: They're concerned about their, uh, uh, the truth dishonesty officer. No. And they need to have a chief, a chief, chief truth officer to kind of offset 

Nate McBride: what they're concerned about is the fact that the, the company made it that every employee in the company has to have three AI goals by the end of 2026. Yeah.

They didn't put anybody in place to fucking do anything with those AI goals. That's, listen, okay. This is just generic, but if you're a company and you go ahead. You made every single employee have to have AI goals for 2026. You suck. What, [01:01:00] what are you thinking? Back that one down? They're laughing at you.

Um, and they're all gonna learn how to use meta prompting and query refinement to just have the, uh, AI write them for 'em. So you failed and. Back that one down. That, that is 

Mike Crispin: awesome. 

Nate McBride: So we talked about, uh, before we got really started here, um, authority verification. So, okay, yeah. Let's come back to that. Who, who gets to decide, who actually gets to decide why, um, someone is who they claim they are?

How do we do it today? We do it today based on a couple factors, right? Like what's your title? What your title enables you to spend, spend this much money. Your title enables you to contribute this way or that way, um, by virtue of your title, right? So your title comes, comes with esteem and, and data privilege.

[01:02:00] Uh, could be your, your function. So I am in the IT function, so I need to see all the IT stuff that's my permissions. Like that's what I get, right? But then you see just immediately get into all the gray areas. Well, I'm also on this committee. I'm also this and I'm also that. Um, so somebody above me needs to verify and has the authority to do that.

So it's CEO. Okay, but what CEO has the answer to every single authority verification need in the company. Yeah. So I think a way we get around this is all around the company. We've empowered people to have a verification authority, uh, privilege for like, Hey, I need to share this document with you and you work for me, so therefore I have the authority to do that.

It's not written down, right? It's not, not in the job description, it's [01:03:00] not uh, in the, some sort of working document. It's just part of how corporate companies work. Yeah, great. But what about, um, what about this basic question, like, and I talked about this in the book, but, uh, employee, like you request access to a sensitive file and I'm able to verify you're you, but how do I verify that you should also have access 

Mike Crispin: now you need someone else to do that.

Yeah, 

Nate McBride: exactly. Like I, what I would do is I'd probably ping your manager, can Mike see this file? And your manager would say, yes. Yeah, now I can stop. I can stop right there. Or I can ping that manager's manager and ask them if that manager should be able to give you access. Like I have to, I'm not even kidding.

Like this is where you do, you run all the way up the chain. Well, some people might say yes, because that file is so sensitive. Oh, well now we have a, now we have a new [01:04:00] problem. Um, how do I know that it's that sensitive? And you see where this is going, right? So. The authority verification. I think 

there's, 

Mike Crispin: yeah, I was gonna say, I think there's some account there is most of accountability is on that first level manager.

They should be. Look, if they don't know how sensitive it is or what it is, they should say they should escalate it. Not, not, you shouldn't have to do it. Yeah. Um, and we, we, we know, we know that that doesn't happen, right? I mean it's, oh yeah. You know, I'm, I'm, I'm, I'm in a meeting. I had a meeting to hop on in five minutes.

Just go ahead. No, no problem. Yeah, yeah. You know, so I, it's, that's what happens. And I, I think this is in an awareness thing, um, that, so, so, so often I think it's sort of lost in the shuffle is it's just data security and, well, more than that understanding, and hu and hu [01:05:00] just humanizing it a little bit more.

And 

Nate McBride: I agree. But I would say that, sorry, go ahead. I would also add that, well, AAU author, I mean, authority is contextual, right? Um, sure, yeah. I can, I can certainly, I can definitely approve X, but maybe I can approve YI, I don't know. It's contextual every single time it's a new context. Unless it's like extremely black and white, right?

Then you have, yeah, you have authority changes. Mike got promoted. Um, exactly, Mike. Yep. Mike has a role change. He's got new delegation functionality. These, and this goes right in hand hand with that context. Well, Mike's title didn't change, but now he has new responsibility. Yep. So, and then you have, uh, that verification event happens at the wrong time.

So only at login, not at the action itself. Uh, which I think back to my Documentum days and e room days, you would log [01:06:00] in and then when you accessed a file to open it, you verified again. And then when you close that file, you verified a third time. Nobody, I mean, I can't remember if people complain or not, but I feel like, I think I, I, I think if fired people complained a lot, I'd probably remember that.

Um, and then you have, uh, a manager delegates approval authority to an AI agent, and the agent therefore can spend a hundred thousand dollars. Is that, is that, is that valid? And maybe the company somehow is able to get around SEC and, and all the sort of auditing state rules to make that happen, but. 

Mike Crispin: I think the same, uh, just in terms of agents and trusting them, we, we've, you know, for for many, many years [01:07:00] built intricate overly complex integrations that automate some, some of them real time integration that's supposed to move information from one place to another, like bank accounts, um, you know, impor very important things.

And we trust that that is working correctly and that it's been built by the right person, and that the right person who holds the public private keys is responsible. And so there's an an awful level of, an awful lot of trust of deep trust, or assumed trust, I should say, with very crucial information just because someone works in a certain department, so they must know what they're doing.

Um, so I think with, with agents, I mean largely the, the, the additive is that they can [01:08:00] act on their own with their erroneous erroneously actually, and you can't explain why. And I think that's, that's the need for the verification component of a lot of these things. So if some agent is applying, is, uh, approving something, uh, on, on someone else's behalf, we just gotta make sure that.

That agent has the same level of scrutiny that that integration rew wrote 10 years ago. 

Nate McBride: Well, would would you go so far as to say that as much as I would challenge the verification of who Mike is, I should apply the same veracity of challenge against an AI agent. So like, why does the AI agent get a pass?

They shouldn't, right? They should be, they should be also be forced to prove that who they are through their, through their own. I I 

Mike Crispin: think one of the nice, the nice things about this, and, you know, we're both using markdown and, and prompts and other things, is that I think if we, if we write these agents with [01:09:00] the appropriate prompting, detailed prompting, that we can use that prompt to help prove the expertise of the agent.

So if, if we've just got this prompt we've written that's very, very high level and not specific as to what the role of the agent is and what it is supposed to discount or not do. Uh, so we haven't written an intricate, we haven't written a persona for this agent. We've basically written just a, a, a paragraph of crap.

Nate McBride: Yeah. 

Mike Crispin: That's gonna be hard to verify. Uh, if, if we're using one of the, you know, like the prompting that you're teaching Nate, like if a very, we, we have the right compartmentalized prompting that's built into an agent and to an MD file that's linked to that agent. It's all traceable and viewable. That might be better than an integration.

As long as you trust the model, uh, you, you've done the vetting of the model and the, the, the platform you're using, [01:10:00] let's say it's Anthropic or Gemini or, or what or one that you host yourself. I think having that map in the prompt is what's gonna make it easier from an IT perspective to verify that this AI agent is doing the right thing.

Where the risk comes in, I think is when these simpler tools like, uh, the AI studio from Google are merging, which is fantastic and it's very cool. The antigravity stuff as well is that it's making very, very, and actually Workspace, workspace Studio just came out today is making these things so simple that anybody can write an agent and write it in a way that is very general so that it, we shouldn't trust those agents and we need to formalize and build agents in the appropriate way so that they can be verified.

And that, that, I think that's one of the areas where, you know, it can help lead the way is that we're, when we're building these AI agents, that [01:11:00] they're standardized and centralized and managed. So that the prompting is appropriate and can be traced and understood. And anyone who asks about that agent, you can hand them the prompt and say, this is exactly how this agent is work working.

It's, its responsibility, it's ethical code, it's capabilities, it's access control. Everything is right there in the prompt. That's key. But there are gonna be a lot of agents that are just handmade, just like very basic. And companies might trust them with everything. And then they, and that's where I think we get into trouble.

So it's almost like any other tool that we implement, any other integration that we do, if it's documented appropriately, if it's, if it's built in the right fashion with the right concerns by someone who knows how to build them, I think we could be massively powerful. But if we give everyone the keys to build a bunch of AI agents, I think we're gonna be in trouble.

Nate McBride: Do you find [01:12:00] it interesting at all that, um,

I mean any, every company I know and I would, so I would go so far to say, most companies verify identity constantly, you know, every session at least once a day. Verifying identity, sure. But never reverify authority after the initial access grant. So on day one, Mike gets two things. He gets identity, he gets authority.

Yes. Authority is ne never challenged 

Mike Crispin: again. I think it's the, you know, we talk a lot about onboarding and offboarding, but changes get lost. So role changes get lost, project responsibility changes get lost. Internal, Hey, I had access to a project I worked on. Uh, I have still have access to a project I worked on six years ago.

I don't need access to that folder anymore. Or I shouldn't have access to that [01:13:00] folder 'cause someone took over my role and I moved to a different role in the company, but now I still have all this access. Yeah, I think there's still a fundamental it component that I think, you know, it gets lost a lot of times 'cause it's not communicated well.

Changes are difficult, uh, role changes, authority changes, like you said, I think they're not verified because they're often not highlighted or understood until, or people don't speak up and say, Hey, I don't need access to this anymore. Um, well there's, I I think it's gonna be built into your process. Um, I know I can certainly do it better.

Uh, definitely seen it, you know, enter a ticket, you know, and something's changed. You're li you're largely relying on the managers or the, to do this. But to your point, the we, since we own the security of the data, we need to verify that and we need to do it more often. And I think it's, uh, it gets lost in the shuffle 

Nate McBride: unfortunately.

Mike Crispin: I think it's a good 

Nate McBride: point. I mean, [01:14:00] that's, there's two sides to that coin. You have the, um, you know, the, the, so I mentioned the, I, you know, you do verify identity frequently, at least once a day, if not more. But you only ever verify, um, authority one time. We don't do the same thing on the other end either, which is like a a a deck shows up in your Yeah.

In your environment. Um, this is my favorites, right? When a deck shows up, it's completely done in box, right? It's completely done. It says version one. So the person created it outside, they did all their stuff and they brought it in. Okay, no problem. So we have no idea how it was created. We gotcha. But then the question, I mean, they, they brought the document into the environment so they quote unquote are owners of it using the nomenclature of the box.com world or, you know, your normal storage classification, but who wrote it, [01:15:00] who actually wrote the thing?

And then same thing with code, same thing with images. Is this something that you actually took or did you use AI to generate it? And the videos and the data and the decisions. So there's the top down approach. Then there's the stuff that's being created to feed that top down approach. Um, sure. Have you heard about the, uh, have you heard about the C two pa?

No, I don't think so. All right. You should look this up. It's the, it stands for Coalition for Content, Providence and Authenticity. Okay. And it's a TE Accord. It's a TE Accord. Got it. Um, which Adobe and Microsoft, the B, B, C and a few others are all part of, and what it does is it just embeds made metadata, and again, this takes us all the way back to Windows, uh, 2000.

Got it. What, whatever, whatever was in [01:16:00] place during the Documentum EMC days where metadata was enforced and it showed, it took all this information and built it into the document. Now you get chain of custody for a non chain of custody document. So you Sure. You get, you get more than just version history.

Which version history to me has always been, uh, lipstick on a pig anyway. But the C two P is behind that. Then you have, um, Adobe content cred credentials if you're in the Adobe ecosystem. Oh, that's what this is. 

Mike Crispin: Okay. I have seen this before. Yes. Yeah. 'cause Google's using it to verify Yes. Verify images right now, right.

Two. And then you, and 

Nate McBride: you have the cryptographic signatures. Yes. Yeah. I mean, basically if you were to account of, of dating your company today, that has providence to it. I mean, you can, if you simply were to say. Uh, I have Providence because in in box it shows that this person owns this document, then you're probably like, way high [01:17:00] percentage.

But if you were to truly actually try to prove Providence, Nate created this document on his desktop at this time and this is what happened to it, um, that's very hard to do. Yeah. Because metadata, metadata can be removed. You can screenshot something and create it from scratch. You can copy paste from Claude, um, and nothing stops anyone from adding false provenance.

So the chain, the best chain of custody can get very quickly poisoned. Um, yep. And so like, again, it's the ai it's the distinguishment question, right? If, if an employee uses AI to create 90% of their deliverables, then they add in a little sprinkling of human touch, a little, a little human dust, you know, the flaking off of the dead cells on on the stuff.

Do they create it? And do you care if it's, if it's, do you care? As long as it's good. Back to the plumber question. I mean, yeah. This, this is all the [01:18:00] providence part from, okay. Even if we were to solve the authority verification, who gets to decide who sees what? Yeah, we still haven't solved for who gets to create what?

Mike Crispin: Yeah. And how, how do you prove it? I, I complete completely hear you. That's a great, great discussion. And I think in Providence in general just

are, are you an, are you an art? It's where does, where does it, where does art end, you know, can you create art by assimilating a number of things that someone else created and create a new art with it? And I think that that's, that's the one of the questions certainly, um, collaborations and, um, what's the other word I'm trying to think of?

Um, 

Nate McBride: well, if you're mapplethorpe or Banksy, you can do that for sure. But 

Mike Crispin: yeah, I mean, if you're, if, and without, again, without stealing o other people's things, without their consent, [01:19:00] obviously, but I think it's some of the whole dream stream and dream state stuff is that you, you may have a unique idea and you may have a, a unique approach and you use these tools to help make it a reality.

You know, is it, are you still actually creating this idea, this pro? Are you still creating something that's art if you use these tools to help create it? I, I think there's a big debate about that right now. And the question of whether or not you, you, you should be allowed to do that or that you should.

You know, you should be, uh, you should be an, if you're really an artist or you're just kind of a poser. 

Nate McBride: Right. I guess to some extent. Right? No, no. It's if you're a plumber or not a plumber. I mean, we're getting right back to expertise verification again. Ai. Yeah. Well, I think 

Mike Crispin: I, I, I, I think that's, no, I, I'm just, I'm just saying like, there, there are people who you [01:20:00] didn't go to school for it and are experts in it.

I mean, there's people who didn't go to art school and are some of the best artists ever, similar with musicians, and there could be a plumber who's didn't get certified and learn from AI and just has a natural skill to retain that information and apply it. I, I think that's more of the way the world's going.

And, um, it's almost these self-made experts, uh, that, that's, uh, merged. And I'm not in every area, obviously, but it's, it's, is the expertise good enough to solve problems or is it better, is the passion to be successful, stronger with some of these people? Um, well, I mean, it's all over the map. I think it's all over the map is what I'm saying.

I think there's examples of very, very people who may have one year experience, who have picked something up [01:21:00] and are better than people who have 12 years experience. Let's, well, 

Nate McBride: let's give, let's give a shout out to those people that can generate, uh. Expert level content in any domain rapidly. Um, I'm, I'm all for the, for it, right?

Like if you need, sure, if this is gonna make you sound smart for a presentation, then go for it. But I mean, that, that just makes even more meaningless. Any credentials that someone might get. Like if AI can pass most certification exams, what's a credential? And then even experience can be fabricated. I mean, people have been lying on resumes since resumes were invented, but, um, yep.

Yeah. That's, that's 

Mike Crispin: where the, the, the creative and the ability to ideate is, again, thinking, talking about 10 years, 20 years down the line. I feel like that might be all that's left. 

Nate McBride: But let's suppose you hired, you hire a senior developer, 

Mike Crispin: that's thing. 

Nate McBride: You hire a senior developer, Mike, three months in, and you realize that they've been using gen AI for everything [01:22:00] and they don't know fuck all about coding.

Uh mm-hmm. You keep 'em, how could you have verified this? Like, do you keep 'em on board? I mean, seriously, would you keep 'em, I might ask them, 

Mike Crispin: can you do more?

Can you go faster? If it's good, if the work product is good, if they're using cloud to write an application that is very successful and working well, it's like, can I hire three more of these people? Do three times the work? I mean, I, that's, that's kind of, I'm thinking the augmentation angle. You can't just be a.

Uh, you can, people have to have some ability to understand what they're doing, I think. Um, but again, maybe I'm just putting too much faith. 

Nate McBride: Okay. 

Mike Crispin: Okay. But 

Nate McBride: now I'm, so, now I'm gonna back you into a corner. Um, yeah, yeah, sure. If I can hire anyone with the remotest knowledge of it and have them use gen generative AI to do any role in it, then why couldn't I just hire anyone into it?[01:23:00] 

Why? I mean, yes, let's say they have to have like two years of experience about what a computer actually is. But yeah, if they can use generative AI to fake the whole thing, and their primary skill is being good at writing prompts, who cares anymore what mm-hmm. Your experience is, can you type, do you know how use gen ai, you're hired here, you're gonna be the VP of, you know, butt sucking.

I don't care what it is, but you're going to do this because you're apparently break with prompts. We we're gonna, no, we're not going to. I can see a future where we, we no longer care. Um, yeah. I what your expertise is. I think that 

Mike Crispin: I, I, I think it's, I think there's still other soft skills and other areas of interest that people will need to have, but I think it'll be more important for them to be able to leverage those resources in some, in some roles than it will be to have.[01:24:00] 

A, a degree in machine information systems. 

Nate McBride: You know, I 

Mike Crispin: think, 

Nate McBride: let me, let me, let me ask you one question real quick. Sure. When, what year will it be, do you think, when you won't care whether they know how to do it or not, so long as they can make it d make it happen.

What year? Um, how many more years before you would let somebody, you would hire somebody who doesn't know anything about a topic, but is very good at prompting?

I, I don't think in the next five years, 

Mike Crispin: I think I would expect to still be able to interview and have, now they could prompt and get all the right answers. That's fine. But as long as they're able to have the conversation and they're able to speak to it without sitting in front of the computer, there's, uh, value in that to me.

So if, maybe so it does, it 

Nate McBride: does matter that they know, that know the subject or it doesn't matter. [01:25:00] 

Mike Crispin: It, it matters that they know it in the respect that they can recite it and they can speak it. And they can apply it. Okay. So in an in, in an interview scenario, right? Like I would want the more, it's harder to do this, right?

Because we do everything virtually now, right? You see the TikTok videos of people doing this all the time. Um. Is that, I don't want your ability, but I get the point your your, your ability to figure out if someone's being honest is just as time goes on. And even more and more so maybe even now is you just gotta have a radar.

They can recite to you all the things that they know. Maybe they know that, maybe they don't, but you've gotta be able to interview them in a way that you can weed that out of them and ask them questions if you think that it's being recited there. Some people have rote memorization and lie on their resume.

So I, I guess what I'm saying is that I'd rather them be able to apply these skills and [01:26:00] have the front wheel that's so important that we talked about in season one. That's be able to work with people and be able to connect with people and if they need to do their research to get certain things done. I, I think it's, you know, what do you think about, it's also hiring people.

You can grow. They don't know everything. Sometimes they know and, and, uh, many times you're hiring them for a specialty that, that you don't, that we, or I don't know much about, let's say. So I don't know of a, of a year. Uh, but I, I think it's more about a, a team fit and a company fit more and more as time goes on than what they know 

Nate McBride: for it.

So, okay. So. I think what we're both getting at is to point that within the next few years, it is possible you could hire anybody off of the street who has a basic understanding of prompt, prompt writing. Uh, or you could teach to them in just a matter of, you know, two or three hours, but they had a very nice personality and [01:27:00] turn them into any role in your company, conceivably.

I think that's, I think that's a couple of years away. 

Mike Crispin: It's po it's, it's, I don't know about any role, but I think, I think lot of, well my first, 

Nate McBride: my first set of prompts is to go into Google Dini research and to ask it, uh, how a particular role would work and function in a certain role that they've been hired to do.

Understand all that. And then use my playbook for every other prompt that I would write. 

Mike Crispin: That's the key, is understand it. And I think the amount of information that comes out of this type of stuff, if you're, if you're don't have any background or any context, a you're, you're not gonna be able to verify any of that information.

And we're, that's a, that's another whole problem, right, is Yeah. I implicitly trust everything that I get outta Gemini. I go to do my job and it's actually not the, it's not correct. It's not the role. It's, you know, it's that there's nuance to, to your company or your therapeutic area. It doesn't, doesn't line up with whatever they researched and they can't retain it all.[01:28:00] 

They can't re You can't. If they can, they're genius. Like all the information. If you don't have experience as x, y, Z role and all of a sudden you prompted a few things and two days later you're an expert, you're, you're a genius. I you're hired. If you can retain all that stuff and jump right into a role and do it by just reading it on chat g PT after two days, that's, that's amazing.

Well, I mean, if, if they, if they're, if they're just, if they're in a, in a role where they're not in front of people or they're not connecting with people and they don't have to think on their toes or then Yeah, I think it's, there's very good possibility that could be anybody and that there will be roles like that.

So I think there, it's definitely possible that you could hire anyone. There's a culture fit. Let's, let's, here's a good, here's a good way to think of it maybe is, I mean, think about just in, you know, maybe when we were younger in hiring people, you hired people you knew, you hired people who you thought you could grow, or people that, uh, you know, [01:29:00] didn't have any real skill in these areas and they, they learned from working with someone on your team.

Yeah. I mean, can they learn from an agent? Can they learn from ai? As long as they're a team player and they know, you know how to solve problems and they have a passion for the business and the, and the therapeutic area you're in, and the patients, maybe it's the same thing. You know, you bring, now you have a better tool to train them and to bring them up to speed.

And I'm speaking more so in our industry, in our area. Yeah. Yeah.

Nate McBride: Yeah, I mean, there are companies that are, that are rushing to fire everybody in the hopes that they can make this happen and they're kind of doing it in a back ass way and they'll regret their decisions I think much later. I, 

Mike Crispin: I, I agree with you because I think that this thi this, this, this automation and agent stuff.

I mean, let's take Claude for example. I mean, you know how big of a fanboy I am now? Uh, I I love the product. It's great. It's fucking slow. [01:30:00] 

Nate McBride: Yeah. 

Mike Crispin: It takes a long time to do simple shit and it doesn't get it right. And if you don't prompt it correctly, and even if you prompt it correctly, you've wasted an hour for a garbage output that now you've lost it all and you gotta start over.

Like this is stuff that in an agentic scenario, like if that happens, people get discouraged and throw it away. Yeah. And I, I just think that there's, that we're is a little bit of a head, I think in reality that this is gonna take everyone's job away. 'cause it's not automatic. It requires humans to build it into the business logic of your company for it to work it.

Most of these pilots have failed because everyone's using chat GPT and thinks that that's the answer to everything. Um, or did I think that was just like the, the thing you were supposed to do. Um, and there's the, this is just like any other major transformation. [01:31:00] 

Nate McBride: No, just like any other major transformation, there's, there's been no other transformation.

I think that's like this, that brings into delight all of these questions. Like even cloud Oh. Did not bring up all of these questions. 

Mike Crispin: Oh no, I, I I I I'm talking about the amount of human intervention that's gonna be required to make this for everyone to lose their job is, is, is huge. Like right now, oh, I, I don't think everyone's losing their job then.

I, I don't either. But that's what people are saying right now. There's a lot of, a lot of like, oh, it's gonna cut. Uh, they're firing people 'cause they're gonna replace everyone with ai. Like, that's actually that worth they're doing. I think what's gonna happen is if it pro, if it, if it delivers in the promise that people really think, and maybe this is pie in the sky, is that companies can be able to do more with the same amount of people.

And if they don't want to do more, they will reduce their staff. But it's gonna take a long way to get there. But I think more [01:32:00] companies wanna make, they wanna make more money and they're gonna force multiply with these tools. And if someone knows how to use them and how to drive them, they're gonna, they're gonna hire them.

And that's the augmentation theory, I think is over, over time that there's, it's a blend. You're gonna need a human in the loop for a lot of this stuff. And you're just going to, I think if you step aside and go, you know, some people are like, oh, just this isn't gonna work and I'm, I'm gonna be replaced.

Yeah, you're gonna get replaced. Period. I think if, if you're not, like, like you and I and a lot of our peers are up to our eyeballs in this stuff, trying to figure out what happens next. And the same time, even though we may be optimistic or pessimistic about it, we're trying to learn everything we can about it.

And that's what we should be doing. 

Nate McBride: It's, it's understanding your, your opponent, I mean, effectively. Um, you know, it's, it's, it's in, when I think [01:33:00] about my 2027 hires that are in my hiring plan, I do think that, and, and I've done this, but um, way before there was a generative AI zeitgeist effect. I'd hire people that had no it backgrounds, but that had Yeah.

Something extra special that, um, could translate. And I've done this, I don't know how many times, but, and many, it always, always pays off. Yes, yes. I agree. And so, so, and look where they are 

Mike Crispin: now, Nate. Many of them, I mean, we don't know. I know we both know many of them. Uh, and I, I know many of that I've, it, it's fantastic to see where they've gone.

And, 

Nate McBride: but I guess the, the rude question still remains Mike, which is, um, in terms of expertise. And how much emphasis we place on it today. Like using today as the comparator, you know, oh, you're a senior vice president of X, Y, Z, you must be an expert in [01:34:00] X, Y, Z. Okay, we've connected one.to another, we're done.

There's no more dots to connect. Mm-hmm. But I hire, I hire somebody from, from, you know, off the street who is multilingual, who's great at typing and understands how to write prompts. You are now my expert in Python. You're gonna use, use Claude to learn Python, write python, and make Python for me. That's your job.

'cause you're really good at it. And also, you know Yep. Be, be amazing, all these other things. Um, yes. I, the, the, my like the authority and verification landscape just changed. They just, that's immediately changed when I did that because that person's got authority and they have no world, the expertise having like background to have that authority.

But then they just granted by default on the role, which means that needs to be solved at the role issue, uh, which we are not gonna do tonight. And then there's the, um, I'm verifying you now. We're only one [01:35:00] step away when we do this. Mm-hmm. To just simply hiring a bot. Um, well that's who can do the same thing.

Mike Crispin: That's where our buckets come back into play. So if we talk about our dream staters spectrum all the way on one side and our verifiers on the, all the way on the other side. There's, there's still an element of creativity, let's say a creativity id, ad ideation, passion. Yeah. Um, perseverance. Right. That's gonna come on the, on the, sort of, on the dreamer side of the house.

And then there's on the ver fire side, as we get closer to that, which I think is happening in different pockets, it's more statistical. It's it's logic. It's, it's almost like the, um, Myers-Briggs or the, uh, you know, the, the, the, uh, what are the, what are the, what are the colors again? What's, oh, insights.

Insights. Right? It's like there's just [01:36:00] like, you take those, those areas and those buckets and you apply them to the, the sort of this futurism we're talking about. And even if you bring someone in for that role to learn Python, they've gotta be willing to stick with it. They've gotta be interested in Python.

They've got to, they've gotta genuinely be into it. Yeah. And be passionate about it. And I think that if you bring anyone off the street who isn't passionate about it, but has a great personality, they're not gonna succeed. So I think there's still, um. Sort of personality traits or emotional intelligence components that are part of hiring the right person so that they can succeed and learn and grow.

Uh, but now we have these new, like you said, these new tools that can help them along the way. Yeah. To not to not just learn, but actually produce a result very early on that will build their confidence [01:37:00] to be more passionate as they grow forward. And it does change the, the hiring dynamic and the roles and how you set up roles.

Well, we also have to, and it's, it's optimistic and the respect that it still allows for jobs to be created because we're realizing that we still need that until there's androids, I guess, that are just like you and me with feelings. Um, R 2D two. Yeah. Like Nexus seven or next to six from, uh, blade Runner.

Uh, until we get to that, until we get to that, there's still a human element, a human in the loop need. Okay. 

Nate McBride: Okay. So, so hold that thought for a second. I, I wrote down some notes that you just wrote and especially on the verification side. Yeah. Um, so right now we, the companies in trust hr, by virtue of their sole existence to be a verifier, they're given the rights to verify.

And they do, they do a lot of [01:38:00] verification pieces. Yeah. Yeah, absolutely. Absolutely. It, it then is also a verifier. We are the, the two primary verifiers in the company. Yep. Really, everybody else, I would guess I'd put in the class of sub verifier or adjacent verifier. They're, they're agreeing on the VE verification, but they're not the verifier themselves.

So I think those two, so say, 

Mike Crispin: would you say that the would, would you say that the, the hiring managers, sort of the SME verifier though, the subject matter verifier? 

Nate McBride: No, they, they, they can't do verify unless they know this person personally in real life. They can't do, it's going on a cv 

Mike Crispin: that's going on the cv.

Nate McBride: Yeah. I gotcha. Right. So, so we have two groups of verifiers or three, we have IT and HR as the primaries, and then you have everybody else that's involved tangentially as sort of like some sort of sub verifier. But like, what do we have for verification? We have, we have government IDs, [01:39:00] so, um, driver's license or passport.

And you mentioned clear before. I am, I know, I, I have a clear presence. I have an identity, but I'm not, I think clear is a, I don't think that's a government backed function. I think that that is a bipartisan, uh, the government, was that a, is that a Peter Thiel company? Uh, clear. I'm not sure. But then you have, uh.

What's coming, which is, like, for instance, the EU has their new digital standards, which is awesome. The US will probably never get them. Um, we have mobile driver's licenses now be coming out. Uh, we have national digital identity. Uh, like let's talk about the uk I mean, the UK is coming out with a national digital identity.

India has already got it. Yeah. Especially in the, in the province. So the problems of course, is that in the US will, will never have a standard. I mean, we have driver's license. 

Mike Crispin: Don't, 

Nate McBride: don't forget 

Mike Crispin: your favorite, [01:40:00] your favorite identity service that's coming down the line. World coin. Yes. 

Nate McBride: World coin. Um, thanks Mike.

Mike Crispin: Don't you want to get one of those? You need one of those little eyeball things, right? I mean that's, that's the answer to all our problems. World coin. Let's do it. 

Nate McBride: That's right. That's right. Uh, world coin. Is that 

Mike Crispin: Sam, Sam Altman's thing? 

Nate McBride: It was, yeah. He remember he had that, that that expo last year. Yeah. I think it was in India.

And if you did it, you got like a certain amount of Altman coin or something. 

Mike Crispin: You, you scan your eyes into it. Yeah. And the whole thing. Yeah, yeah, yeah. You 

Nate McBride: feed the database. Yeah. Yeah. Fuck that. So, I mean, again, I love it. I don't know if the US in our lifetimes will ever have universal standard because the company's split, the country's split in half.

Um, of course there's privacy, which is, I don't know, pretty much all but gone anymore in the us. But government tracking of your identity usage, um, not everyone has a government id. I was say there's 

Mike Crispin: a topic, [01:41:00] the privacy's dead. You should add that to the list. Privacy is dead episodes. Is it? Pri, is privacy dead?

Nate McBride: Just 

Mike Crispin: kept by that. We can for hours privacy, 

Nate McBride: private fuck privacy as the episode. 

Mike Crispin: I didn't mean to interrupt, but I'm like, I make sure he writes that one down. 

Nate McBride: All right, I made the note. Um, so, so let's back up to the, the HR and IT people. Yeah. Uh, with all these sort of, uh, uh, ident, ident verifying functions, um, what should your company allow and right now, so, so most think most companies like ours does driver's license and social security.

Never, not no. Driver's license and passport. Yeah. Those are our, those are our two. In fact, I think it's just, just passport. You don't even need a driver's license. You just need a passport In our company, no. Is that okay? Yes. Passport. Is that a sufficient digital id or, I mean, I think it is. I'm just, I 

Mike Crispin: mean, yeah.

I, I think it is, it's very, very hard to spoof. I think the more [01:42:00] that they can tie things to the phone, it, it's even better. I mean, it's, I mean, we're verifying some of the most important. Things that are in our lives now by the phone, whether it be bypass key or just password or email, things that are a text message.

Not the greatest thing, but still. Um, and you're less, very much less likely to not have possession of your phone than your passport or know that it's been lost, right. Or know that someone else has it. So I, the more that we can tie things, this is why, I don't know why Massachusetts is so slow on this compared to every other state in the country, but, you know, not being able to put our IDs on our phones.

Um, if you lose your phone, you're screwed. I mean, but at least your phone's locked. You know, your phone is a password or it's encrypted or whatever your passport or your license is not. And it's just like, it seems like a, in this day and age that that would be a slam dunk. But [01:43:00] yes, I think passport. I think passports are fine.

But, um, I can't say phones yet because you and I both know that we probably have worked with people in the past that still can't quote an authenticate our app on their phone. So Yeah, 

Nate McBride: I I know, I know. But basically they might be too 

Mike Crispin: soon. 

Nate McBride: Problem. The problem with doing this is, uh, it's, it's all great and fine until you have the first employee that comes from XUS, which uses like a different, a different ID system.

And now all of a sudden Exactly, exactly. You have to expand your verification with, I hate to reuse the word here, expertise about how to verify verification from another country. That's why World ID man. World coin. World coin. Okay. Yeah. Alright. So, I'm sorry, Nate, I'm sorry, I'm sorry. No, no, no, no. This, this, this episode, by the way, is brought to you by World Coin, uh, and Sam Altman, uh, just wants us to remind you, you can shove it up your ass, uh, because he doesn't give a fuck.

He just put $5 billion behind some startup about increasing his life. So buy [01:44:00] world coin, shove it up your ass. It's an orb, the orb and s up your ass. It's an orb. Multiple si, multiple sizes, uh, depending on, um, you know what you need. Okay. It's a, it's a big silver ball. Yeah, that's right. Big silver ball. So that's our advertisement, that's our, that's our, uh, support for, for today.

So then, uh, so I was, I was cracking up because I was trying to figure out. Um, I'm sorry. That was, that was good. That was good. That's okay. No, I was, so, I was, I was a passenger in the car and then went to Pennsylvania for Thanksgiving. Yeah. And I was trying to, I was doing some research in the back 'cause I was, I had to do work all day in the car.

I was trying to figure out, um, is there actual verification for LinkedIn? Like, can you make an account? That's totally bullshit. And what I, what I discovered both by, um, the lovely world of Reddit and a little bit of self experimentation is that it's mostly an honor system on LinkedIn. It 

Mike Crispin: is. 

Nate McBride: You're right.

You, um, [01:45:00] except to my Google, except to my Google voice as a number. So I could make any Google Voice number. I wanted to verify my account. No need to show a license or anything. I couldn't. But as long as I had a picture, uh, Facebook and Instagram, from what I understand, only give verification for whatever, whatever constitutes a celebrity.

Uh, GitHub only needs your email. Yep. And then, um, most platforms, it's the absolute minimum because they want to get people to actually not run away. Um, and so I guess the que the question I have is, we were talking about badges earlier, but let's suppose that Mike gets the verifier badge for his company, like it's on his little slack avatar, verifier, or.

You know, dream streamed officer. Does, does that badge then give you a sense of trust? If someone's got a badge on their name, [01:46:00] what does that do to you? Any badge? Just any like, like on LinkedIn, my badge says, uh, I think premium or something. Yeah. What does that badge mean to you 

Mike Crispin: on LinkedIn? It means nothing to me.

I, I don't even, I don't even know what they are. I mean, I, I see them on there, but I don't even what they are. But to your point, those badges, I don't know if they would, they would hold much for people that I don't think they would mean much. Yeah. I badges must it became more of a standard thing that there's important and like, I guess like I was saying before, like in terms of rewarding people and incentivizing that the, with tokens, let's say, just that it's, it's actually valuable to them.

It, you can use them for different things. That would mean more to me than a badge. But 

Nate McBride: yeah, again, I'm talking about the, the ver the visual [01:47:00] verification. Mike's name has a badge on it. It says he's the verifier. Do I then immediately stop there in my, in my quest to prove Mike is the verifier because I trust badges and it's just, it's just a weird, you know, I just, I think about this because, uh, it's so weird.

You mentioned badges and I look around and I see badges like on Discord. You know, people have badges and stuff and you're like, oh, that person's got a moderator badge. They must be a so and so. Um, badges are a funny psychological mechanism. I think the, 

Mike Crispin: the benefit to badges, at least as long as they're still verified with a credit card, 'cause you have to pay for them, is a decent verifier, at least in the public arena.

So in the social media arena, I think it's, it's good, it's not perfect, but you had to give them some payment infor information. And unless you stole that, which is [01:48:00] also very possible, um, then it's, it's okay. It's an okay ver it's better than not having any verification that, well what about third party services?

And another verifier is, is just, is the crowd, is is the crowd component again. So like if, if you, if you get a friend request from someone on Facebook, let's say, and you can see that, you know, the 400 people that they're connected with, 400 of the thousand people that they're connected with, you know who they are and you, you can identify their picture.

The chances of those all being fake are very, very small. So the crowdsourcing component works great better for verification if you have that visibility, but that's usually only one. One, um, degree that you get from that person. So if it's anyone outside of that massive network, then it's useless. Like if that, that that methodology is useless.

Okay. So if you're trying to identify someone that doesn't know, if someone's trying to [01:49:00] connect to you, you should think twice that it's someone you don't know. Like the InMails that you get probably Right. That you, you know who they, who they are. They're not associated with anyone you've ever met before.

Um, well, 

Nate McBride: so we, we, we, we, we use Okta for enterprise identity. That's pretty easy. But what about the other identity providers? I mean Yep. You have, you mentioned Clear, there's a persona. There's id me, yeah. Um, Lexi Nexus will do it too, apparently, but, you know, yeah, there's a bunch of these company, the, the downside of course is that they all have massive identity databases.

So I just wanted to make a quick amendment to a previous point. Like I was saying that HR and it are the primary verifiers, but, um, that's actually not entirely true. So I think what HR is, is a proxy to the verifier because what HR will do, at least in my company, is take a, [01:50:00] a photocopy of your passport.

Then they send that, send that to the, um. The, uh, background check agency and usually you might have to file a little form too for the background check agency who then verifies you, tells HR you've been verified, at which point in time they pass on your verification to the next stage of the hiring process.

Um, sure. So I mean, to the extent that you would ever consider centralizing on one of these third party services as your primary verifier and then like not having to go through the, the, the middle people of hr, would you ever do one you think?

Mike Crispin: I think if we feel like it's important that someone is a, is a human number one, yes. Um, and there may be, there may be cases. And there, right now, I think in the immediate future there will be the amount, [01:51:00] the background check, I think is still necessary because criminal history or criminal, yeah, yeah, yeah.

Right. Or, uh, or, or even, even credentials. You know, sort of, you know, degrees other, did they actually 

Nate McBride: graduate from this college, you know, like you said. Yeah, exactly. So 

Mike Crispin: I think we still need that, but I think there really is value, uh, to have a. Identity service that exists, that bridges businesses and personal, and it can be used to verify who you are without you having to share any additional data.

I think there's huge value in that. And if that emerges Yeah. Uh, in a, in a productized form and in a, in a trustable form, uh, in a trustworthy form, I should say, then yes, I think it would be a great thing. I don't think it's there yet, and I don't think anyone has dared go that far, but I do think as, uh, you know, in, in the Web3, in the Web3 realm, something will emerge some at some point that [01:52:00] will cross 

Nate McBride: both planes.

You mentioned one thing that I find interesting, which is, uh, back to our earlier point, if you use a third party agency, and let's suppose that we use third party agency, and part of their role was to verify prior employment expertise, uh, to get a role for my new, um, just do whatever the hell I tell you with outputs.

Exactly. Yep. They're not, there's gonna be no alignment whatsoever unless your, all your previous expertise was, uh, prompt output. Sure. Adventurer at alio, a lot of, 

Mike Crispin: uh, a, a lot of roles that I think are that, that, that are out there. They're, they're, they're just looking to verify your high school diploma and college degree and those type of things.

I think, and criminal history and references are usually used if, if, if used at all, uh, to confirm. By the way, do you use references anymore? I'm just curious. Um, I, I, I, I haven't, I've gotten some, [01:53:00] I've had a, you know, in past lives where I've had a few final candidates where I've wanted to speak to them, uh, just 'cause I couldn't make a decision.

Like there was too, too many good candidates. Um, yeah, it's usually, that's usually when I use them is when I have two really strong or three really strong candidates. Yeah. And I, I'm trying to find some, some qualifier that helps me make a, a decision. Um, 

Nate McBride: but if all the references are like glowing, then you just 

Mike Crispin: Yeah.

Wasted time. Well, sometimes the quality of of of references are, can, can be like, if you speak to someone and you can genuinely tell they're they're passionate or you even with the person you're speaking with, that's the references. You're having a real conversation with them. It's not just someone who's got the checklist in front of 'em and it's just yessing you to death.

It's, that could be a, that could rub off positively on the candidate. Like the, they're they're quality references. They're not [01:54:00] Yeah, I worked with you 10 years ago. Can you just put in 20 years ago, can you just, you know, someone's gonna call you. Can you just put in a good word? When, if they're really quality references.

Sometimes if there's kind of a tiebreaker, it's interesting to hear that someone who's put the time in is probably someone who's a good reference and. If they've connected with someone that's that good, a caliber of, of person, then they're good at building those relationships. And that's, that's a good thing.

Um, I, I mean, that hasn't happened yet, but I could always see like, Hey, that could be something that puts it above. Yeah. I, I think it put, to be honest with you, one of the, one of the times that it wasn't really a tiebreaker, I had called, reached out to a reference and I couldn't get in touch with them.

And, um, you know, the other person I spoke with was highly, you know, highly regarded life sciences background. It's like, oh, this is great. And it helped, helped me make a decision. [01:55:00] So, yeah, I, the answer is not usually, but that's one instance where I, where I, I, I do 

Nate McBride: Okay. Um, 

uh, 

interesting thinking about how we're gonna do this a year or two from now.

Like, I generally don't rely on references. I will go and find out who that person knows and I'm connected to, and then go through those people. Mm-hmm. Uh, and it generally works really quite well for me. Uh, plus I also just know when I'm talking to somebody if they're gonna be an asshole or not, so I can, I don't need someone else to tell me that.

Um, yep. But, uh, you're right. Absolutely. I mean, 

Mike Crispin: that, that, that is true. I mean, you're gonna know. But we get, we get to, you should be able, you gotta be able to read people. Absolutely. Yeah. 

Nate McBride: I think we, what's happening is, I mean, the things that you mentioned, and we're getting into this clear IDs, we're getting into like the surveillance creep problem where we start off with just saying, Hey, listen, we just wanna verify you're human to [01:56:00] one level deeper.

Like we need to verify you are actually you, and then we need to verify you pretty much continuously. And then lastly, we need to verify every single thing you do. Like you can see it escalating, right? And in a very short period of time. I mean, short being relative, maybe three to five years. But we can see like every single time Mike's, if Mike's given the, the capability to edit and create an organization, Mike needs to be verified constantly.

If Mike's just a commenter or a viewer, maybe he needs to be verified a little bit less. Uh, yeah. But yeah, you can see how this is going to grow. And we get the, the only next, the only the very next step is full surveillance. 

Mike Crispin: I think the, the question in the shorter term is what's the probability from a pers personal verification perspective that someone is being, is being, uh, spoofed [01:57:00] or being cloned or whatnot?

For, for nefarious purposes, let's just say the, the probability of that, uh, just in my opinion, is probably. Probably a little bit lower in the respect that we need to continuously verify them, at least in the current state. Yes, there's deep fakes and there's voice duplication, and there's a lot of those components.

Hopefully, again, just trying to weigh the probability and the risk is that there'll be other warnings that will help us to identify those things. And again, that's optimistic. I know. 

Nate McBride: Yeah. Um, 

Mike Crispin: I do think there's more likelihood that work product in the short term will need to be verified. And that's, and that, that, that I think is coming very quickly and an area that there's a, there's a gap and I just don't know how we're gonna do that.

I, I don't know how someone's gonna have to, to figure out sort of 

Nate McBride: getting 

Mike Crispin: Well, you give people, uh, this is you, [01:58:00] people being, people being duplicated, I think is, is already an issue. It's already a problem. But it's, 

Nate McBride: it's where you let them create. I think it's rare. Yeah. It's, yeah. It's where you let them create versus where you let them comment.

And I, we, I know we keep coming back to this idea, but not everyone needs to be a fucking creator. And if that person's really gonna be a superstar at commenting and, and providing commentary, don't give them edit rights, period. It solves this entire problem. One shot. Yeah. Well, not entirely, but it solves a big portion of that problem in one shot.

Or if they need to edit, may only let them edit at a terminal. I don't know, but like, bring it down to a level that's completely verifiable. But that's, we're, I think we're talking about an expertise level that, um, we, we, we don't have everywhere. Like we don't [01:59:00] have, like, my kids wouldn't know how to do that.

And they're smart kids. We have low income individuals, we have elderly people. We have people in areas with poor connectivity. Yep. Disabilities injuries. They, they have poor biometrics. In and on and on and on. Yep. How do, how do we, how do we, you know, how do we and digital literacy allow these 

Mike Crispin: people to, 

Nate McBride: we, we have people that are in their sixties that are gonna be around for 30 more years that don't have a fucking clue what's going on and, and have no way to learn it.

It's moving to transfer that. That's right. That's right. 

Mike Crispin: Yeah. You're, you're absolutely right. That's, it's a it is, it is. Definitely. You mentioned digital literacy, so we we 

Nate McBride: exclude them from the, from the population of our No. Our scope. No. I think, 

Mike Crispin: I, I think they're, they're, they're in included in the scope.

I think it's a question of how realistically they can, they can have access to understand or if they even are interested [02:00:00] in understanding. The amount of effort that it might take to learn, you know, this, I, I mean, I can speak to experiences I have with, with family members and friends of family members, and that they could be compromised one time and the next time it happens again.

And we're not even talking about, we're not even talking about complex attacks. It's deep fakes and all this stuff. We're talking about basic social engineering. And it's, I think it's a, we talked about trust earlier. It's a symptom of default. Trust. Trust everything first, don't verify trust. And it's, we've, I think as generations have gone on, it's become the, the tru, the trust, the trust thing was everything, right?

For our, maybe our parents. Yeah. Right. And I think in our [02:01:00] generation, I think we're both Gen X, right? It's this kind of trust and kind of verify, you know, just thing, something seems weird, so I'm gonna verify it. And then I think the younger generations, they may not just default to not trust, but not verify either.

And now I think that, that the most recent ver uh, generation, it will be verified. Well that's a, that's a friction issue though, 

Nate McBride: mean. Right now, click on, click on Amazon. You, you buy the book within two clicks. It's on its way. There's no friction there. But imagine like a, a friction where you go to Amazon, okay, prove you're human.

Okay, great. Now prove you're who you are by logging back in. Great. Now prove your authority. So open up your two fa on your phone. Great. And then go ahead and press this button. Then do that all again before you have authorize your credit card. Then do it all again before you press this shit button. Go fucking way.

Every one of those clicks is mm-hmm. You're [02:02:00] gonna lose, lose a customer or lose a UA user. 

Mike Crispin: And I think that's, and that's, you know, going back to what we were saying earlier, that means that the, the, the, the human element of work is going to be verifiers. I mean, the job, the work that's being done is largely a verification.

Can be creative, it can be new, it can be interesting, but it's essentially creatively proving things. And there's a whole eco, there's a whole economy that's gonna be built on that. We haven't figured out how to do it yet, but that, that's hold on you. 

Nate McBride: But you're proving, so you're saying there's no verification required and everything must flow through them, or there's some level verification required.

Mike Crispin: Look, look at, look at Amazon today. So you, you just mentioned Amazon. I buy a product. I don't like it. I return it, I got my money back. I buy the product again. Same product. I can return it. I use it for two months, I return it. [02:03:00] No big deal. I can do that. I buy it again. I return it. What I'm trying to, the example I'm trying to give is that the most successful companies have reactive mechanisms.

Yeah. And the user experience is not about protecting them upfront and saying, oh, let's make sure that they, you know, they don't, they don't, they, they verify and they do all this stuff ahead of time. They reactively verify and they reactively win. And that's, that's the future to me. I think that the jobs that are created are gonna largely be verification after the fact, after the damage potentially is be done.

So we can fix the damage. And there'll be tons of jobs that are created to retroactively put things right. It doesn't sound efficient, but that's why there's gonna be a need for tons of humans to do it. And that's the way it'll work so that people have the best experience, they can move the fastest, they get the most done, and ask forgiveness afterwards and clean up as much of the mess retroactively, because there'll be [02:04:00] thousands, hundreds of millions of ais to help with that too, in some respects.

But we need tons of humans to help do that, that work. And they'll make it creative and they'll make it so people feel like they're doing something rewarding. But it won't be about protecting, they'll be about responding. And that's, that's just philosophically what I think will happen and why it'll still be very necessary in the equation, is that we.

We will have to respond to errors and mistakes. And like I said, a lot of these shops and a lot of these, these companies like DoorDash, another example. Yeah, my burgers screwed up. I'll just, I'm just gonna give you the money back. I, I'm not gonna ask you a question about it. I'm not gonna, you can take a picture of it, but it's, who cares?

It's at least gives us a record. But my burger's messed up. They just, oh, we'll credit you six bucks and you'll get it back. You'll get the money back, no questions asked. Like this reactive model. And I think we're kind of, [02:05:00] I think we're moving towards that, where we're gonna need a lot of human intervention and human in the loop thinking to still be able to move fast, but save ourselves from destroying ourselves.

I think it's all gonna be responsive and reactive and not proactive and not built on governance and regulation. It'll be built on, oh shit, we gotta fix that. And just human nature, it'll be all reactive and we'll have enough humans who are looking for things to do, to fill those roles and to help make, put things right and solve problems.

It's just on the other side instead of the it happening before we get to the product and the result, it'll be post result that we're working. And that's just just my philosophical belief. I think that if, if we keep going the way we're going, and we may not. We may not, we may find that this is all a big illusion and it's only, it's 20% efficient for us.

We're ki [02:06:00] we're spending so much money and energy and burning up all this stuff and damaging the environment. We may just come full stop at some point and say, maybe we just need a little bit of this. But if we keep going at the pace that we're going, I think we're largely be a verification economy that's more about reactive response and asking for forgiveness, like I said before, instead of asking for permission.

Okay.

I'm sorry, you probably got your, you got your No I response. Sorry. Um, that's okay. I know you're, 

Nate McBride: you're getting 

Mike Crispin: pinged. 

Nate McBride: Gimme one second. 

Mike Crispin: It's definitely, um, yeah, I know for the listeners, I think it's this, this is just an area in which I think a lot of roles will emerge and right now it might sound like, oh my God, that sounds horrible.

You know, we're, we're just gonna be checking boxes and, you know, checking things out and trying to back trace [02:07:00] things that we're not even sure we know how it work. And I think the new jobs will be created by the people who can build those proofs and can find innovative ways. Yeah. To explain results that have been done, that have been, that have been concocted by machines and.

It will take expertise, it will take people who understand how to code. It will take, uh, people who can think, uh, logically and pragmatically. It will take people who have education and have a background. Um, it will take people who, who understand algorithms. So I think it's still there. And then in terms of subject matter expertise in which it may not be as close to computing or to digital comp, compute, digital work.

There, there are people who, you know, may have at least passed down more and more from the trades perspective, interpersonal relationship perspective. There'll still be reasons to learn [02:08:00] skills and learn how to do things. But it won't be to, to build, it will be to verify. 

Nate McBride: Yeah. I, I, I realize we're getting a little long in the tooth on this episode, but I still have so many, so many, so many parts of this.

And one of the things that, yeah. Let's do another 

Mike Crispin: one that's, we can, we can continue next week or just table it too. 'cause I know we're getting, we're, we're, we're getting along. Oh, 

Nate McBride: we're, we can, why don't, uh, what I wanted to talk about was our role. I mean the, yeah, sure. The original point of this podcast was, you know, it leadership and, and all, you know, all the things that come along with that.

But we haven't really touched on what you and I need to do in all this. Mm-hmm. Uh, and I think I wanna talk about that and then. Sure. Like there's some, there's some still big questions on the table. Like how would you define the right balance between friction and security? Uh, and, uh, even though it's about the, we, we talked about [02:09:00] tonight and we, it's in the book, we should verification be continuous and things like this.

Like, I, I wanna get into these, so maybe what we'll do Yeah. Is let's have a second episode next week. Okay. Yeah, sure. Sounds good. I think this is a, 

Mike Crispin: Nate, in fairness, I think this is a massive topic. I think they're there and it's a new one too. I'm not sure our peers are thinking about this maybe as deeply as we are.

Um, there's so many other things happening. I just think this is one of those areas. And, and you're right, we haven't talked a lot about our role in all of this, but I think it is important for us to set the stage just for what you and I are thinking around the, the, the impact and why it's important that we under, we understand it and how, like you said, we will have to address it.

Okay. I do think it's a gradual thing. I don't think it's something tomorrow we gotta set up a strategy for. I, no, but we can talk about that. 

Nate McBride: All [02:10:00] right. So, so next week. Uh, come back and join us for part two of the Verification Economy. We'll talk about sort of the, one of the big topics we haven't gotten to yet, which is it's role in all this nonsense.

Um, Mike, and I'll pick that apart. And then we have some other open questions we still haven't answered yet. We'll get to those next week. And then, um, we will, I mean, we have episode, episode fives, uh, kind of a big deal coming up. And then we have, um, I mean, just looking at the rest we have, episode five is who Speaks for It anymore.

Yeah. And then episode six, I cannot wait for is Slack as the operating system. I mean, obviously it's meant to be discussion about all the possible things that could be, um, that could be an operating system these days, but Sure. Uh, that would be a good one. So anyway, join us next week. Do we have any, do we have any special guests coming up or anything?

I think for episode [02:11:00] five, um, Pablo might be joining us, but uh, I have to confirm with him. 

Mike Crispin: Cool. 

Nate McBride: Cool. Um, and then potentially some, some other folks down the line, but, so for next week, part two of episode four on the Verification Economy, um, uh, be nice to people. Uh, have your pet spade or neutered. Um, happy holidays, both, uh, retroactively and to those that celebrate holidays in December.

And, um, Mike, and I'll be back on the Wednesday of next week, which is probably the 10th. I would have to if I did the math. Yes. And don't forget to give us all the stars on the systems and don't forget to, um, buy our merchandise. It's in our show links, uh, bias beers where you could use them. I'm down to Jagermeister again, so I have to go back and make another beer run.[02:12:00] 

I don't mind Jagermeister. Actually, I quite like it, um, as you can tell by the banner behind me. But, uh, it's usually how I, it's usually a starter for the night, not an ender.

And I'm gonna go solve the world. Well, I solve the world's problems here. Here's to all the future, future, uh, model auditors to you, model auditors out there, and you chief truth officers, wherever you may be. And however you may do your nefarious tactics, you know, tasers, uh, electrodes, uh, whatever it is that you do to get the truth.

Here's here, see you. I knew you would love that one. Just, just as soon as you said Chief Truth Officer, you just lost. The reason that went in my mind was electrodes. Like just sitting in a, sitting in a chair naked, a metal chair naked with electrodes and someone's got like a, oh [02:13:00] goodness, a hose. You're lying.

Sir, what is your name? It's Nathan McBride Zap. No, you don't, you don't have a name. You're the guy with no name. That's your new name. Okay. A couple of, I don't know, A couple of wavy lines. A couple of wavy lines. 

Mike Crispin: This isn't your lucky day.

Nate McBride: Is it a star? Why? Yes. It it's a star. It's a star. 

Mike Crispin: I didn't know Dr. Benjamin. They're just coming to me.

Nate McBride: Well, all right. Uh, December 10th, join us semi-live. Uh, part two, episode four. Rock and roll. All right, dude. Good night 

Mike Crispin: everyone. 

Nate McBride: Thank you. All right. Be good, man. Talk to ya. You too, man. Thank you. Have a 

Mike Crispin: good night. All right, you too. Later. Bye-bye.[02:14:00] 

Trance Bot: The calculus of it,

season three,

verifying this identity.

Sometimes 

Speaker 8: you just have to take it.

Sometimes you just have to take it.

Trance Bot: It's season three divided autonomy,

verifying

the calculus of [02:15:00] it.