Tech on Trial Apple OpenAI and the AI Accountability Debate | TSG Ep. 1025

Hey, everyone. Happy Monday. Hey, I'm sorry we're a little bit late.

I'd like to blame it on the snow, but we don't have any snow down here in Florida. Who am I kidding? We had a little, we had a little production, a little production blip, though.

Some of us, I guess out here today are in the snow. Mm-hmm. Uh, let me introduce you to some of us.

Anyway, as long as I mentioned them, we've got a great, a great gang line up for you today joining us from the true north where, I don't know, is it snowing up, up by you, Garima? Yeah. Always snowing here.

Always snowing here. No big deal. I love that attitude.

Garima, Baal. Uh, also where it snows quite a bit though, everyone I know who's going skiing is telling me there's not enough snow there. Mitch Ashley in Colorado, yes, we have little, very little snow in Denver, but in the mountains there's snow.

That's where it needs to be. Very cool. And then we've got Mr.

Lake Effect, Steven Foskett. I imagine it's snowing pretty good there. Yeah, I actually had to shovel twice this morning, so yes, it's definitely snowing here.

And, and Mike, I I've already spoken to three people from New York this morning doing interviews all in Manhattan as it turned out, but I heard there's 16 inches of snow in grand in Grand Central, not in Grand Central in Central Park. Yeah. And it's closer to two feet up here in Westchester.

And a and, and an interesting phenomenon happens, and the kids are home from school, and suddenly I'm fighting for bandwidth, right? Yeah. Yeah.

That, that is true. That's why you need that fiber. Um, I'm not talking about the fiber you eat, I'm talking about fiber optics, but it all, you need fiber you eat too.

Anyway, guys, it another busy week, last week, A crazy weekend. Lots of new news coming out. We're going to talk about this week.

Um, Mike, you know, questions of culpability and, you know, are we blaming the right people? What, what's this one about? Well, This conversation has been going on for some time, but now the state of West Virginia wants to actually take it to court and see what the courts have to say about it.

But they're saying that Apple bears some responsibility for moderating and blocking child exploitation material across its services. Basically, they're saying that Apple is in, uh, some like legal jeopardy for distributing what amounts to child pornography. Interestingly enough, this comes at the same time when OpenAI is saying that, well, it actually did see that the fellow who was involved in a mass murder in Canada, they knew about that as well, or at least they knew about his mindset and, but they felt it didn't quite, uh, come up to their standards for reporting.

Now, that's the question is, well, what is that standard and what should it be? So, Alan, I know we've kind of been kicking this around for a while, but it never seemed to get to court. So are we finally gonna settle this one way or another?

No. Move on. Oh, no.

I mean, the short answer is no. The short answer is no. You know, let, let, let's look at both of these cases a little bit, and I'll try to do a brief about it.

In the Apple case, you know, we, we, the, I'm not making Apple to be the only bad guy, but all of these hyperscaler, social media, tech bro giants, they hide behind things like five 20 and, and these other, uh, uh, statutes. And then what's going on? Saying, well, we're not responsible for all the data that goes out on our network.

We're not responsible. We can't watch everything. And, and, you know, we're don't, don't shoot us, go after the people who are putting it there.

You know what, I used to believe that in a big way. I, and I, I, you know, I felt that doing anything less would be stifling innovation. It would be stifling freedom of speech, freedom of press.

But I, I've, you know, what did President Obama say? I've evolved, I've evolved in this, in that, um, I, I do think these, the apples, the mets, the Googles, the exes of the world bear some responsibility for cleaning up the sewer that has become the internet, right? I know, Steven, it offends your libertarian, libertarian morals, but if not them, who, if not them, who Now let me go to the open AI situation.

'cause that, that's my hammer. First of all, they pay attention to everything you're searching and writing and saying in ai, their algorithms. And that's really the problem.

We have become slaves to the algorithm here in this world. Their algorithms picked up that this person who did this horrific, horrific shooting in, in, in a small town in Canada, was dangerous, was, was a prime suspect to do something violent like this. Their own, their algorithm showed it to them, so much so that they canceled that person's account.

They canceled that person's account because they were exhibiting behaviors of violence, uh, you know, violent behavior, potential for violence. However, their own internal threshold, before they would report that to, in this case, the Royal Canadian Mounted Police was not met. So they shut the person's account.

They didn't do anything else about it. And a few months later, eight people are dead. Eight people are dead.

This is why I say, and this, it is not an isolated, not an is. Oh, and, and by the way, after the fact open, AI says, oh, Mr. Mounty, Royal Canadian Mounted Police, we have all the evidence here, let us give it to you.

That's great. Tell that to the relatives of the eight people who are dead. We gotta do something we otherwise forget about authoritarianism and totalitarianism and all this, it's algorithm.

The algorithms are gonna rule us, and we, we need, we need to do something. Steven, I'll see the floor to you. Okay, well, I am not gonna argue that, that on the OpenAI case, I think that that is absolutely ridiculous that they, they knew about it, they talked about it internally, and they decided to do nothing.

And when I say they, I mean real people who could have done something. So I'm certainly not gonna argue that the Apple situation is a little different. Apple is not a maker of social media or web software or really anything.

Apple is a maker of individual devices that are privately used and privately shared. And this West Virginia lawsuit that we're hearing where they sued Apple accusing it, of not doing more to, um, combat CSAM, well, again, apple could have perhaps done some things and they have developed, uh, software that would allow them to flag this stuff. But I actually am on Apple's side on this one because any attempt to break, uh, end-to-end encryption is essentially doomed to failure.

As those of us in the cybersecurity industry know any backdoor is, or any, you know, extra keys available to law enforcement or anything, it's going to be exploited. It's going to weaken the platform, and it's going to cause havoc. You know, uh, somebody asked me the other day, what's the safest way to send a copy of my ID to my accountant?

Should I send it by email? Should I, you know, send it by Dropbox? Should I send it by iMessage?

And I said, hold on, they've got iMessage. Send it by iMessage because it's end-to-end encrypted and nobody can intercept it. That's how secure this stuff is.

And, and the fact that these, uh, West Virginia lawmakers are trying to weaken encryption, it reminds me of what's happening, what has happened in the past in the US and what's happened in, in, um, the UK and other places where they're trying to bust a hole in Apple's end-to-end encryption. I don't want that to happen because I like the fact that my communications are encrypted. Do I like ceasing?

No, but this is not necessarily the, I I think that there's bigger fish to fry, and I think that breaking and end encryption is not the right answer. I'm not clear that you have the right to end-to-end encryption. And I think the current laws would suggest that if it's being transmitted over anything that looks like a telecom network, that government has some right to intercede and look at that data, and they don't even need a, uh, a formal subpoena.

They can go to one of these handy dandy little courts that they set up to get that request. And so, you know, I understand what Apple's trying to say. I'm just not sure whether there's any legal doctrine to support the position.

Well, that's not true. During, you know, that then all this grows outta nine 11, right? Where in the, in the throes of nine 11, not knowing what was coming next, if anything, we needed to get into encrypted phones and encrypted said communication.

Apple said, no, some Israeli company sold it to us anyway. Right? Said, well, they hack that, and then Apple developed means to circumvent that and protect your phones.

Uh, we do have a right to privacy. I believe. I, I, I believe, look, I believe your right to privacy extends right to the tip of my nose.

But if my life is in danger because of your right to privacy, well then we have what we call a balancing of rights and, and, and responsibilities. And if your right to privacy is endangering my life or my loved one's life, and, and Steven, you know me, I'm not, I'm not from the, you know, from, from that side of the house. So I, I do believe in a right to privacy and freedoms and all of that, but I do think that we, we need to carve out something for the public good.

And, and that's the thing Apple did. They developed software that would allow them to detect this material, even in end-to-end encrypted conversations and to report that to the authorities. And that was rejected because what it didn't do was provide the authorities, the keys to your house, the keys to your phone, the keys to your Communication, right?

They go in, and I'm gonna tell you something else is just to show you, I'm not an anti privacy person or freedom. I don't know if I'm not a big fan of child pornography, obviously, right? I'm a father of children, but I don't know if that's kind of stuff that rises to that level for me.

I needed, like in terms of weapons of mess, disruption, destruction, life and limb type of things, not just garden variety pornography or financial crimes even. I think you're skating on skin ice, my friend, when you, 'cause child pornography isn't gonna be considered garden variety, and it's not gonna be, well, It's not garden variety. It's terrible, but it's not, it's not a weapon of mass destruction or wide public harm.

But if, Well, let me jump in. I do. I am sensitive to that and, and people are absolutely harmed for life by this material.

But Abso I'm not saying no, I'm not saying no, but it's different than find, trying to figure out where's the next plane crash and into someone's building. You know, I, I'd follow in the absolute yes, I, we have a right to privacy. We don't have a right to privacy to conduct criminal activities.

We don't have that right period, whatever it is. So there has to be some mechanism where there's the FIA accord for international or something equivalent to be able to get to data and to communications when there is, you know, strong enough probable cause if they need to investigate. And frankly, if the, if the authorities aren't gonna accept a Apple's offer, at least it's a start.

You know, well, shame on them. That's not good either. Mm-hmm.

You know what's interesting? Everyone agrees that open AI was wrong not to go to the authorities. Garima, we haven't heard from you.

I think, uh, in both the cases, my opinion is that we have to act carefully. Because if you don't scan these systems or you don't have a mechanism to scan, I mean, being a woman, being a mother of like a kid, I feel very unsafe with all these kind of, you know, privacy issues, which people have because it's God forbidden. I think we have a, uh, the, the law, the government for a reason, right?

And, uh, it's, I'm not new to all this because I work in critical infrastructure space. I work for telco, right? So telco is not new to the server.

Agilance monitoring, behavior monitoring is the core, right? I mean, we have to kind of obli we are obligated to do that. And in case of, like open ai for example, it's a case of reporting, right?

They didn't report because, or they, they, uh, they saw that this was a privacy kind of a, uh, issue. But I think in Apple's case, I think it's also similar. And I, I do believe that in both the cases, and the answer is not very, um, straightforward to be honest, because it requires a lot of diligence on how do you build governance and how do you treat these applications like critical infrastructure.

Uh, if these kind of, uh, uh, activities of behavioral, uh, scanning is, uh, happening at open AI level, why didn't they report? And, uh, they have to be obligated to do so. Of course, there is one way to get around all this, and that's hold the phone up to the person's face, and now you got access.

You could do that, or we could you that their finger off and you're No, but, but here's, here's, here's where I wrap this all up into a point, which is I don't believe any one company, one organization, one government even should be the, the, the, the, the purveyor, the, the, the per the, the entity that yy this. I think we need to develop public, private, worldwide, even this a this is what things like the United Nations should be working on, right? Worldwide kinds of things where, hey, maybe child pornography should rise to that level, you know?

But these are the things that rise to the level where we, we, you know, the, the, the, the public health public wellbeing outweighs the rights of the private, of the, of the private individual. But The thing we call that, isn't it a form of exigent circumstances? And you could argue that if we had something like that, we wouldn't be talking about, you know, Jeffrey Epstein and all this crap.

But remember, I think there are other reasons why we're still talking about Jeffrey Epstein. I think after seeing who's involved, it's obvious what was shave Gone a decade ago, Release the whole files. Well, if we all lived in a panopticon, nobody would ever be able to do anything wrong, right?

Right. Thi this is true, Steven, that I, I'm not saying it's not, I I'm just thinking it's wrong to blame just Apple or just open AI or just Google. I think this is the kind of situation where you need standards, yes, standards that are agreed on a, in some sort of worldwide body type of thing.

And I recognize that what what may be acceptable in the EU may not be acceptable in the us what's in Canada, may not be in China, but we, we need, we need standards around this. We, it's been kicking around now, nine 11 was 25 years ago, and we haven't really developed those standards, But the standards have to reflect the reality of technology. Yes, it does.

And, and, and this whole idea that there sh could be some master golden backdoor key that the government can use in the cases of blah, blah, blah. It's simply not technically feasible. No.

But I think, Right, but where it starts from, right? I mean, every standard it evolves. And if you incentivize the right behavior, it starts there.

So if, you know, these problems exist, these are the cases there has to, this is an opportunity, right? And if we want to take this opportunity to build a standard for behavioral monitoring and in incentivize companies who are the first, uh, you know, first in the line, that would be great. Mm-hmm.

All right. But I wouldn't trust necessarily the United States Department of Homeland Security in its current form with that capability. Right.

Amen. That's a good segue into our next segment, Mike. Yes, it's yes.

What do we got here? All right, well, you all remember, if you don't, if you're old enough, radio Free America, and I guess we're gonna have now radio free internet, and the United States government wants to set up a portal where they're gonna put content that, uh, other countries, particularly in Europe, have banned. And so the idea is that this is the ultimate form of free speech, and that you can access it through, I guess they're gonna set up VPNs for this purpose.

And it's gonna be kind of interesting. Of course, you know, where it gets really interesting is, well, is there content that was banned in the United States that somebody else is gonna set up? Steven, how crazy can all this get?

What a freaking joke this is. Speaking of, speaking of what Alan was just saying about this, wouldn't it be great if people could get together locally and decide together on community standards and, and make rules that they would all agree to behave on? We could call them governments, and then we could create, you know, I don't know, geographical boundaries where people lived and call them countries.

And then we could, you know, have, I don't know, international rules that we abide by. gov. So essentially what the United States is gonna do is they're going to allow, uh, presumably right wing nonsense to, uh, go to countries where it would be otherwise illegal.

It so, so basically the United States is going to violate international law and international local standards by distributing things. What a great idea. Um, you know, it, it, as pointed out in some of these articles, it's, it's very different, um, from what Radio Free Europe was doing and Radio Free America was doing, um, in around the world during the Cold War.

This is essentially set up to violate, you know, the, the local rules and regulations of the internet in various places. And maybe that's a good idea in authoritarian regimes that don't allow people to have like the us you know, their own. Um, I would personally like that here in the us.

Um, but I, I don't think this is a really great idea internationally. And frankly, I don't think it's at all technically feasible. gov and call it a day behind the great firewall?

I mean, it, I just don't see how this is gonna work. That's What I was gonna say, is, by the way, somebody needs to tell 'em that you can block VPNs too. And yes, they can Add more servers and domain, but it's open source, Mitch, it's open source.

So look, okay, let me, let me just blow my horn here. Stop show. Okay.

gov is. And then there's my article. Go read my article.

Okay. 'cause let me, so this is not a site, I Can't get to it through this portal, I'm sorry. Well, no, well, the, the US may have blocked it.

Yeah. The US thinks that this is left wing, right? So we, we can block that.

But see, you wanna know the truth, it's not even about right wing. This is not a right wing left winger. What this is, is they've set up an app and a website to be a VPN.

They're storing no content. It's just a VPN that lets you go wherever you wanna go, wherever you feel, right? What they're blocking is, it could be right wing, it could be crazy left wing.

What they're blocking is Europe. Europe had a war just like we did and the rest of the world did about 75 years ago. And they're trying not to let the forces that were unleashed 75 years ago come back in and start killing 6 million, 10 million and however million people.

And they realized that a lot of that war was brought on by propaganda and hate speech and, and, and speech that ignited people's emotions with lies and, and made them do inhuman things. And so they passed laws about speech, and they've had it before. There was an internet you weren't allowed in Germany since World War II to stand up in spout Nazi nonsense, right?

Now they do it on the internet and Europe said, no, they can't do that. But here's where it gets really screwed up. Here's where it gets, and I pointed out in my article, it depends who you, where you get your news from.

gov is here to, to free the oppressed people of Iran and China and let them have a free internet experience. Let me get clue you in something on my, my friends, the folks in Teran and the folks in Beijing, they, they know a thing or two about evading the firewalls and using VPNs. This isn't helping them.

No, this is not gonna help 'em. They already do that. Interestingly, Fox News doesn't mention about opening Russia up to the internet, but we don't, we do have our sacred cows.

We can't talk about Russia, right? Instead, we're gonna liberate the oppressed masses of Rome in Paris and London living behind that iron curtain that's descended on them to, to keep Nazi speech away, because JD Vance went to Munich last year and, and made a admonished them for trying to clamp down right wing propaganda or what have you. And here's the real kicker.

Why is this under the Department of Homeland Security on my dime and your dime, I thought homeland Security was about the homeland. Hence the word homeland. What the f goes on here?

And why, why, like, what I, I, this is ludic, Steven. I, I agree with you, ludicrous, frigging ludicrous. I'm done.

I, I would wa I want to move a focus on, uh, from ethical point of view to commercial as well as technical point of view, right? So ethics, I mean, we, uh, we can discuss this for long hours, but from a technology perspective, also, it's not safe, right? It's a single breach kind of a, you know, adoption, which they're propagating for.

And the second point is the commercial viability of this. Like what is, like how you can erode 20 to 30% of, you know, the commercials, which is at stake with internet and open, like whatever standards we have, interoperability and all that. So if we do, we do go it that way.

Ha. Have we done a commercial assessment on this? Like what implications it would have?

So these are the two things which needs more deeper insight as well. I mean, I wouldn't comment on that ethics side because I think, Well, they, they open open source. Anyone could go look at Steven.

You're Are people who, these are people who don't understand. They don't understand how tariffs work. Uh, so, uh, clearly they haven't done even a, an elementary assessment of the costs, uh, benefit analysis of this.

I'm gonna, I'm gonna take this a step further and I'm gonna set up a portal in Boston and start sharing Yankee's content that has not been censored. Don't hear that the MLB will come after you, But you, you know, and, and just comparing this, 'cause I, I, I put it up on LinkedIn this morning, my post, and someone were asked me a question, the difference between this and radio free America, radio free Europe, radio free Marty, which was beamed to like Cuba, Venezuela, all those years, um, there we were actually beaming content, right? To, we, we, it was our content to give our point of view, the American point of view to people who were otherwise prohibited from getting the American point of view.

This is not what this is. This is a VPN that lets you go into the sewers of the internet. And if mucking around in sewers is what you're about, you, you could do it without this.

Believe me. See, I think we're looking at this all wrong, Alan. What this really is, this is a commercial opportunity.

I'm gonna set up a instant response company that when this thing gets hacked and it will get hacked, right? So go in and say, okay, let's remediate this. Green is laughing at me.

I'm here. I mean, it's that ridiculous. Well, and the other thing too is, you know, in talking to my friends, uh, around the world, um, let me just say that there's not a great appetite among the people of the world to trust the US government on this point of free speech.

Well, How do you know it's not censoring people? No, no, but this, this was another point in my article about glass houses and stones, the, I'm not going to use that word. The stones of a government who is checking people's social media when they come back into the country, their own citizen social media to see whether they should be detained or not, is gonna preach to others about how open we should be.

You know, what we call that? We call that hypocrites. We, well, kindly we use the word hypocrites, where I come from.

And the whole thing ultimately is just, it's, it's exactly, you know, they, they like to rail against virtue signaling. Well, this is right wing virtue signaling. This is, you know, nonsense.

It's not gonna work. It, nobody's gonna use it. It's gonna be a complete flop in fiasco.

Frankly, it'll probably never even happen. But they Don't insult right wing. You're Saying they're gonna do it.

It's not right wing. It's maga virtual signaling. 'cause I, you know, I, I'm, I'm gonna, I'm going to do a favor and say, I think I have right wing friends who don't buy into the MAGA thing, and, and they, they realize this for what it is as well.

So, well, the Europeans use this to kind of figure out who is actually consuming this content. Maybe they'll figure out the IP address of those people, and Why Not? They're gonna need to ask Apple ry, but Why not?

Hey, why not? I, I feel bad for all those folks in Brussels and Marsai and Madrid not being able to, You know, Monaco, you know, all those oppressed areas. Yeah, last time I checked they were pretty good at distributing this content without any help for the United States.

Absolutely. It's here in America where you go to the good sites and they say, I'm very sorry you don't have proof that you're an adult or whatever. But hey, let's, let's move on.

Let's move on. I think we've all said our piece here. All right.

Uh, Moving on to something probably just as controversial, but, um, a lot of folks are publishing reports saying that all this investment in AI has not moved the economic needle. ai. I think it's the National Economic Relations Bureau or somebody like that.

But Goldman Sachs is saying the same thing. And so are all the other Wall Street firms as well. And of course, some of this may be what we call a trailing indicator, but whatever it is, it seems like the first wave of hype around AI is not living up to the promise.

Garima, what's your take here? Yes, I think, uh, I was waiting for this moment, right? In 2026, just that, you know, people are starting to say that, uh, the AI investment at the Hive has come to a stage where, uh, we have to reflect as leaders, the growing gap between expectation and the measurable outcome is not a surprise to any one of us here in the panel, as well as a lot of people out there, right?

So we should be able to do a due diligence. Is it the j curve of the technology? You know, are we seeing this because we are over-investing and we are hyped in, uh, you know, the AI tools, the data centers, and we are yet to see that kind of productivity curve, or we are just thinking about productivity in a microwave, right?

We are not looking at the macroeconomics of, you know, how AI would benefit in the larger, uh, you know, sense to our, you know, ecosystem enterprises. I like, I'm witness to all this. I live and breathe this every day that there are task level efficiency improvements being discussed at the board level, the executives are, um, personally feeling they're more productive, right?

And then, you know, when it comes to how to break that, a AI mirage, nobody has an answer, right? So probably, you know, there is this time where this is the time for reflection. Like why are we in the trap of macro, micro paradox?

You know, we are looking at things which probably would have mattered when we were experimenting. You know, there's a zillions of pilots in every enterprise at this STA point in time. There's over investment, uh, for AI tools and applications.

Have we done our due diligence on how we will ensure it or of investment? Is this, is this a large scale kind of, um, adoption problem? Is it a culture problem or is it just over hype?

So my question now is, are we measuring the right things? If those, if those trials, Mike or those, uh, you know, first AI implementations were wildly successful, let's assume they were, they were just fantastically over, over exceeded expectations, would it, would they have moved the needle? If it wouldn't, then we're not measuring the right stuff.

It kind of goes back, you know, you know this Alan from Gold Rat's Theory of constraints of if you aren't measuring the Right things, the whole book, the goal is about figuring out, well, what's the right thing we should be measuring? Right? Businesses are ultimately here for what?

To make money and to be profitable. So are you measuring the right thing? I think that's a lot of what we have is let's look for the problems that we think might be interesting to solve.

They may or may not be the ones that are gonna move the needle. I, I think the problem is we're in such a fast moving environment right now, that by the time, you know, it, I used to always make fun of analysts back when I was in startups that, you know, when Gartner came out with their magic quadrant, that was a great snapshot of what things looked like two years ago. Um, but the world changed in two years today, that, that timeframe is not two years, it's two months or two weeks in some cases.

I, my my feeling, you know, I, I spent the morning doing interviews with some of the winners of the Quantum Security 25 list we published last week. Spoke to some, you know, a lot of them were involved in AI as well. They, people I've spoken to over the last week to two weeks, I am hearing almost universally a consensus that sometime in November or December, the light bulb went on, whether we're talking about the quality of code that we're seen, you know, being generated by ai, the how it's writing, how it's performing, its usefulness, its productivity, if you will, has really changed in the last, let's call it two to three months.

And so I'm glad we have these, you know, great recalibrations and, and look, no one's saying there's not over hype here. We've all lived through enough cycles to know you can never live up to the hype. This hype is like nothing I've ever seen hype.

But, you know, but to say that it, it, um, it's stagnant. Stagnant from a productivity point of view that it hasn't generated any revenue. Um, I, I, I, I'm in Mitchell's case in Mitchell's camp, I think you gotta connect new dots.

I think there's two Phases to this thing. I think the first phase was these copilots, and they did not live up to any expectation or promise. The next wave is these AI agents, and to your point, the underlying tech is getting better and smarter.

It's also getting more expensive to run. So you gotta be smarter about what it is you're gonna use that for over time. And, and, and you have to also distinguish to what re was talking about.

It's one thing for, you know, me to like be able to write an email better and faster, but that doesn't necessarily drive the bottom line or the top line. So, and also, you know, and business people will tell you, it's like they'll look at you and they'll go, well, it's great. I can do all these tasks faster, but the, the x number of people that can buy the why thing that I made has not increased.

So therefore I'm not really gonna push this kind of business bottom line benefit that you're expecting with ai, because, well, you're a bunch of tech geeks and you don't know squad about what the real business is. Steven, what do you hear? That's, that's A real problem.

And, and so I, I agree with Alan to the extent that I feel like this stuff has gotten a lot better. So let me just point out, I'm not a, uh, a Luddite here saying this, this is garbage. Now, as a, uh, uh, person with a, uh, humanities background, I will say that it's writing is, uh, uh, still writing in air quotes, not actual good writing.

But that being said, it's maybe good enough. I mean, basically we've built a machine and its whole purpose is to fool us into thinking that it's doing a good job. And so it's doing, you know, it looks like it's doing a good job.

The problem is, as Mike's pointing out, a lot of the, the stuff that's happening is honestly a lot of, um, a lot more heat than light, I guess is the right thing. Um, I was talking to a friend of mine who's deeply into agentic ai. He is a very, you know, active claw, uh, open claw user.

Um, and he was telling me about some of the incredible things that he is doing, and it really does sound incredible, except that none of it was actually productive. It was, it was doing amazing things. And, and he's so scattered.

It's like talking to somebody, you know, I mean, obviously the guy's A-A-D-H-D anyway, so it's like talking to somebody with A DHD. But, you know, essentially, um, you know, he was telling me, you know, over the weekend I worked on four different things and I had my agents doing this, and then I went to the grocery store, and when I came home, it had built a thing and it was pretty cool. And so I had my agents, but I had another idea.

And so I had my agents working on this other idea. And then when I came home and I saw what it had done, it was really cool. But then I thought of another thing.

And so I had my agents worked on this other thing, and none of The, but Steven, no one ever said this was adder roll, Hang on, none of us value. And all of them drained his wallet. And that's the thing, at the end of the day, the guy's like, man, this is gonna bankrupt me because it's so cool and it's so productive and it's doing so many cool things.

I used up all of my tokens. Can I have some more tokens please? But none of it's actually, you know, making anything, Anything.

So I think we need to clean a new term. Steven performative ai, that's what that is. It's not really valuable.

I mean, you know, because like, for example, he came up, he made an app, he had an agent make an app, and then he said, but I'm not gonna bother putting it in the app store because yeah, whatever. I'm, I'm onto something else. And so what was the point?

Yeah, the question is not what we can do with ai. The question is what we should do with ai, and this is a mindset shift. You know, how do you bring AI as a strategic collaborator?

How do you, you know, finally curate your use cases where it creates value without bringing tool fatigue? You know, right now what I am seeing that, you know, there are zillions of pilots going on in like enterprise grade level, you know, systems, and does it do something? It is creating more noise than it is bringing value.

And the question is not how fast we build. The question is how we build sustainably. I think you're right.

It's what we build to that point, Karima, it's, it's, this is why I have a problem with AI is for automation. It can be used for automation. But if it auto, let me take you back, Alan, to a time, you'll fill, you'll, you'll get PTSD we created in the early two thousands of product called van.

That was, oh, taking all outputs of nesting it in into a workflow. And all it did is create a whole list of, of, uh, unplanned work for people who didn't have time to do any of it, right? So we created, we, we solved a problem of let's surface this and give you a tool that you can go work it, but we didn't give you the problem, didn't give you the answer of, well, here's how you solve the problem.

Let's get rid of that problem altogether. And I think that's the, we're at this phase where we're gonna see a tipping point where vendors start to focus on, let me start solving the problem for you. You don't need any more alerts.

You don't need any more 500 vulnerabilities that we found in open source code. Who gives a crap if I don't have time to fix it? All you're doing is creating a backlog, making my backlog worse.

Fix those things. Same with observability, same with governance. It isn't about just reporting anymore.

We have to take action. And that's, that's what a is trying, AI's trying to figure out how do we build the credibility and trust with within end users, control, planes, government governance, accountability, things like that are really now being emphasized much greater. Yeah.

But I, I also feel that it's a great moment to recalibrate based on the reports, based on the findings. What we get is that if we are under investor pressure to invest in AI and do not have a strategic landscape, a long short for what we want to do with AI in next five to 10 years, we will continue to lose talent in this industry. And then there is no going back.

I think, you know, a lot of business people are looking at their IT people and seeing Bullwinkle, right? It's another one of these, Hey, Rocky, watch me pull a rabbit outta my hat this time. For sure.

Gee, Bullwinkle, I used to love Rocky, but Boris and Natasha, Boris and Natasha, come, we go. Um, look, just because she can doesn't mean you should, right? That, that's a lesson that we always need to learn to learn.

And, uh, I think we're gonna need to learn it over again multiple times. Performative ai. Ai, that's what it is.

Yep. Hey, but we're outta time today, guys. What a, what a, uh, emotionally charged, packed episode this was.

I feel like I, I cleed, I need to go like take a rest or something. Yeah. Venting, a venting session.

It was a bit of a venting session. It was a bit of a venting session. It's what a weird world.

Anyway, Garima, Steven, Mitch, Mike, thank you. Thank you for watching. Hey, we've got repeat of text on tv.

Uh, immediately following this, Stephen, when's the next field day? I don't know. com.

Now we're, we're gonna be, uh, we're gonna be, uh, doing some, uh, some more field day events, but not until March. So for now, uh, we are focused on things like this. Well, Steven, it is the last week of February, my friend.

Good point. Just I had to be thinking about March, right? Yeah.

For All of, you're derailing me. Okay. March 11th, cloud Field Day.

There you go. There we go. That's what I was looking for.

All right. Garima, you have something coming up? No.

Yeah. Limit. It's a month, so there's a lot of action coming up.

Uh, I've also released, uh, one of my new books, so that is another thing which, uh, we can talk about in the next episode. Beautiful. All right.

Hey, until next time, I hope you've enjoyed this texture on gang. Come back and see us tomorrow. Same time, same station.

Until then, this Alan Humma we're out.