Who Oversees AI Safety, Music Fraud and Surveillance?

As my mo- Hey everyone, happy Wednesday and welcome to the Wednesday edition of Techstrong Gang. I'm Alan Schimmel. I'm coming at you today from Denver, where I'm still out at Splunk Conf, doing my Splunk thing here and listening and learning about so much going on with agentics, obviously.

We've got a bit of a globe-trotting- Mm ... gang panel today. Let me introduce you to them.

Coming at us, you can tell she's not in front of the brick wall anymore. She's out in Iceland. Our own Teri Robinson.

Welcome, Terry. Hi. I don't know how to say hello in Icelandic.

I know. I just realized I should've learned that before I got here. Yeah.

It's a hard language though. It's a hard language. It is.

And then coming at us from his new home, I guess it's, what? Around dinnertime. Well, no, they don't eat dinner till 8:00, 9:00 there.

Yeah. That's true. It's siesta time almost.

Our own Andi Mann. Andy, welcome. My friends, it's tapas time here.

Welcome. Hola. Hola.

Very cool. And then holding down the fort in her New York bastion, our own Liz Saffer. Yeah.

Liz, how are you? I'm great. How are you?

Good. All right. Guys, let's jump right into today's gang show.

We've got some interesting stuff to talk about, and obviously a lot of it has to do with AI. There's so much... The news cycle has now been consumed with this AI extinction, AI's bad, AI's going to do this, we've got to get a hold of it.

And taking a page out of Dr. Goebbels- Oh, no. the bigger the lie and the more you say it...

And I'm not saying AI, this whole thing, is a lie, but I'm just beginning- ... to feel like we're being herded. Right?

That there's somewhat of a concerted effort that... Because you know how our media works. Our media is a reflection of ourselves, and when the media sees high ratings and high engagement on a particular subject, they give you more of it.

And unfortunately, I think it's led to a lot of the problems we see in the world in general. And clearly, the media has decided that this AI is bad mantra has legs, so we're getting more and more of it. Our first block kind of plays into that sentiment and it talks about some of the leaders of the AI movement, and it's always the American companies, right?

Because I don't think China's really bought into this whole thing. But the American AI companies, and the American AI industry, has been meeting to form a blue ribbon panel on safety, and what we can do. Liz, what do you think about this one?

Well, I think a lot. I think it's a very multifaceted thing. But in some ways, I'm for it as a frame of a first step because I feel that the AI CEOs have been talking security theater for a long time and asking the federal government to step in, well knowing that they won't, right?

So I think that they're finally taking some action to frame some sort of safety methodology for themselves, shows that they have some walk to their talk. However, I do also believe that they are not the ones that should be sitting and enforcing the rules. There needs to be some sort of third-party oversight and checks and balances, and that's actually the thing for all our blocks today is checks and balances.

I do think that being inside the labs and seeing... I think that they do see stuff that we don't see about how automated the recursive self-improvement is. I think that they probably do have a good perspective on the dangers, but I don't think they necessarily have a unbiased take on the controls and enforcement.

So I think that it's nice to see them actually walking their talk a little bit, but the devil here is in the details. Yeah. Andy, first of all, what they're proposing is...

Well, so the Secretary of the Treasury, Scott Bessent- Ugh ... jumps in here and says, yeah, they want to set up a FINRA type of organization, but it would be under the auspices of the SEC. The SEC handles public companies and stock and so forth.

Is that the right organization to handle this? And two of them are currently private companies. Sorry.

Yeah. Well, look, obviously self-regulation has really worked quite well in pharmaceuticals, for example. Also in banking.

And airlines. There's no negative impacts there, right? So obviously self-regulation works quite well with firearms.

Look, I'm kidding, of course. The big slash S on my statement should be obvious to everyone. I don't think this is the right way to go at all.

At least there is some government oversight. But of course, the president has come out very strongly and said there should be no regulation, no watchdog, no legislation. And so we're going to go into a self-governing cycle.

Well, what is that really going to look like? As soon as you put the foxes in charge of the henhouse, which is essentially what self-regulation always is, regardless of industry, you're going to have the profit motive overriding the ethical motive. It's going to depend on individuals within these bodies, within this safety watchdog, taking ownership and being ethical and responsible.

I'm trying hard not to laugh here just using these words. What mechanisms are actually going to be in place? What enforcement, what policy?

And I will say, one of the things that really worries me greatly is the organizations that are not part of this process, and especially the organization... It's one thing to have the SEC running over it, or some organization that's government run-ish over it, but what about overseas and foreign organizations? You mentioned China, Alan.

Is China going to pay any attention whatsoever to the policies and methodologies and mechanisms of this safety watchdog? Is it actually an incentive to not join this group so that you can go rogue and dominate whatever ethics happen to be established here? Look, I think it's great that they're thinking about doing something, because goodness knows our federal government is not.

Right. Well, that's a concept of a plan of an idea of- Yeah. it's like what Trump said, right?

First step, yeah. Yeah. And let's be clear, I don't think Trump has anything to do with this, nor will he get behind it.

But this has been my problem all along, the shortsightedness of the media, the shortsightedness of our politicians to think that somehow down in the algae-ridden reflecting pool of Washington, DC- It doesn't even have water in it, Alan. Let's not even pretend ... Yeah.

We could somehow issue edicts like we're Zeus on top of Olympus. " Really, what you're doing is tying your one hand behind, or at least the US companies, one hand behind their back when... And I wrote about this earlier this week, and my friend Daniel Newman mentioned it as well, is I think what needs to be done here is recognize we're not going to be able to stop, slow it down globally.

And the thing to do is then put your eggs into the what do I do about it basket. If I can't slow it down, how do I make it as safe as possible? And look, having a industry watchdog, industry peer review kind of thing.

And we do, you mentioned pharmaceuticals, but the fact of the matter is you do have real trials of drugs, right? And they get sent to peer-reviewed or publications, and these findings have some weight of authority. Maybe something like that, and we could do that on an international level.

It is self-control and self-regulation. And are these guys going to go against their own self-interest? No, of course not.

Yeah. Okay, so I know everybody's poo-pooing this, but I just want to say, I think something's needed, and you know how all our AI leaders are talking about how society is going to change, and we need to imagine what an AI future is. So why not reimagine what oversight and transparency would look like in this new model?

I'm not saying we need the federal, especially this federal government, to put in oversight, but there's 50 states, and there's a couple of states that are really trying to innovate about AI legislation. So maybe we think about what would work. Like you're saying, if we can't slow it down, if they're not going to slow down, maybe we could slow down the process a little bit.

And so I think that there's some there there. " But we need to do something, if anything, just to change the conversation around doom or utopia. Because right now it's very black and white, and as we know, the reality is much more nuanced.

Yeah. I'll just mention two other things. Number one, Bessent again analogized this to FINRA.

I mean, what could go wrong with that? It almost burnt down our economy with mortgage ratings in 2008 and '09. Right?

So that's not a good model. Secondly, though, and this comes right out of my book, "The Indispensability Trap," Emodi volunteered this up in exchange for anti-monopoly protection. Yeah.

Right? I caught that. Mm-hmm.

Which I'm going to tell you is exactly what AT&T, a guy named Samuel Insull was electric... I'm blanking on the person's name now. At the time he was president of AT&T.

He went to the government and said, "We will-" Charlie Brown? No, no, no. Right.

He was two-time president, one time when it was Bell, and one time when it was AT&T. But anyway, he asked the government for recognition of a monopoly, and in exchange, submitted to government control. It wound up with AT&T giving up the transistor patent, which gave rise to the computer industry.

But this is an indispensability trap thing. The same thing happened in electricity, right? The utilities became regulated as monopolies, but were given monopoly status, but the real money was made by GE, who made all the appliances that plugged in on top of them.

And so this does fit my book's kind of hypothesis of things that happened here. Tell, I think, was the guy's name from- Yeah ... from the AT&T.

Anyway, but this is history repeating itself to a certain degree, but it still doesn't answer the international question. Right? The US cannot alone do this.

I don't care how big we are, we're not that far out ahead of the rest of the world. Right. And another thing, Alan, just on top of that, you know that I care about my practitioner people, right?

My devs, my ops, I love my practitioners. Stuff like this is always going to flow to them as well. If there's not black and white rules, if there's not things they can read and know, yes, I should do this, no, I should not.

A product manager's going to put a dodgy requirement into a spec, and they're going to go, "You know what? " Yeah. And again, these decisions are going to devolve to the practitioners who really shouldn't have to make them.

No. And talking about politicians versus practitioners, I would feel better with the practitioners than the politicians. Yeah.

So maybe that's part of it, Emo, right? Yeah. Because I think the whole cartel thing is sort of an out for the labs to, it's sort of an invisible foot on their throats, because if they try and do this on their own, then the federal government can come in and shut them down under that premise.

But something, I think... I don't know, how do you guys feel about the whole nuclear treaty analogy, right, that was able to bring the world together? So I mentioned in my article, and I actually did a little Techstrong special report on this, and I used the anti-proliferation.

So two things there, right? Trust but verify, right? If you're going to do that, that only works if you could actually, number one, have trust in what the regulations are, let alone the players who signed up to it.

And number two, can you verify that it is in fact what it says it is, and they're doing what they in fact say they're going to do? If you don't have trust and verification, you don't have the basis- Right ... for that.

So does that verification part lead us right back to needing some sort of body that is independent, that does our due diligence and protects and balances? You probably do need some independent body. But the other thing is you got to look at...

So I'm out here at Spark. I had a good talk with this fellow, Sunil, who's now heading up all of security observability, and go-to-market, and agentics for Cisco, and he talked about this concept of recursive AI and then recursive security, and sort of a flywheel, right? And that's what we're really afraid of, of runaway AI, right?

Once AI gets to the point where it could improve itself, it could rewrite its own next edition, which in some ways is almost lifelike, right? Mm-hmm. Then how do you put brakes on it, right?

Because it's just going to accelerate its own evolution and capability at AI scale that we may not be able to keep up with. Right. Now, that's- That's real ...

all the panels in the world, that isn't going to help you with that. And we talked about recursive security. I have an article coming up on that where we could use observability to feed a flywheel where security becomes self-improving, right?

And maybe that is the new model, using AI. Everything becomes self-improving, and we improve it. But that leads to this runaway where it's certainly not humans at the loop, and it's not even humans at the helm.

It's AI driving it all. Right. It's the tail wagging the dog.

Mm-hmm. And that's scary. And I don't think anybody wants a loss of control event.

And when I think of AI doom, I don't think about the paperclip issue. I think about humans behaving badly and humans not acting when they should, and that's the- So that's the other thing, right? Humans using AI for ill-gotten gains versus AI.

So I wrote a paper on this too, a special report, about how could AI cause the extinction event people talk about. And they break into those two camps. One is humans using it, and the other is AI itself deciding, something out of a Star Trek movie, right?

Carbon-based life forms are no longer- Right ... required. Right.

Bad for the environment. Yes. We want to change AI.

Yeah, exactly. But these are real things, and I think it's things we're talking about every day, and I'm not into doom and gloom, but I do think we need to figure out how to keep humans at the helm. With that, though, we're a little over time.

I wanted to jump to our block two here. And we discussed this a couple weeks back. 5 billion to a gazillion different authors, and each one- Including my dad ...

did not- The estate of my dad, yes ... they got up to $3,000. " I wonder when tech media publishers' turn is going to come.

We'll get to that. What do you make of it? Yeah.

I find this amusing, of course, because Anthropic did pay a lot of authors, but not me. I might be a little bit salty about that just by the way, because they are on the record as having stolen my work, one of my books. But anyway, look, they are all thieves.

AI doesn't create anything. We know this. It's all theft.

It's all someone else's property. It's all just spicy auto-complete, right? AI has not created a single thing by itself from its own knowledge.

Of course, it has scraped knowledge from everything. So whether it's visual, film, sometimes it's authorized, and we see that quite a lot in things like, I think it was Bruce Willis came up with the rights for his likeness to be recreated by AI because he's got a degenerative illness and is unable to do that. And so he's given up his rights to do that.

Sometimes it's not quite as good, like when the estate of a dead actor gives up their rights to do the same sort of thing. But a lot of the time, most of the time, it is just theft. Let's face it.

I'd certainly question how committed these streaming services are to it. Because it seems to be driven primarily, as the great article, by the way, says, primarily by profit motive here. It's significantly because of, as you mentioned earlier at the top of the show, Alan, the backlash against AI.

They're looking at survey data that says that millennials and Gen Z are rejecting AI-created art, and I get that. And so they're looking at this as, well, we're not going to make money if we don't do something about it. So they think they're going to do something about it.

But I would also personally question whether they're also a little bit contributing to the problem. I'll tell you what, I listen to Spotify and I love their generated playlists. Things like they've got a release radar, new releases that sort of align with my tastes.

New Music Weekly, which is much more eclectic, and it's great. They also have an AI DJ, which they tried for a while, and wow, that did not get things right for a long time. Diving into this too.

And the algorithms they own also surely have some accountability. You think about Google nerfing its own algorithm so people couldn't game it with SEO. For the longest time, it was one of the biggest challenges in media and advertising and so forth was, how do I do SEO when it's unpredictable?

Surely the streaming services have some role to play themselves in this. Look, I think it's great that they're paying attention because I do get a lot of AI-created crap, I tell you, I guarantee it. I'm not happy with it.

I'm like, it's not creative. It's repetitive and dull. It's slop.

Slop. It's slop. 100% slop.

Well, I think the problem is how can you definitively say something is totally AI generated? Because that's the dividing line here. This new agency, rating agencies or thing they want to set up is they call it fully AI generated.

What exactly does that mean? What does that mean? Yeah.

Yeah. I'm going to tell you something. I just registered my book on Amazon and IngramSpark, which is all booksellers besides Amazon, basically.

" Right. " And it gave like a little, a lot, a little bit or a lot, a lot a bit or a little. Yeah.

Confused the crap out of me. I hit the question to explain it, and they said, "Well, did AI write it and a human heavily edited it? Did a human write it and AI heavily edited it?

" And another fourth. I didn't know what to pick. " What did we do here?

What's the right answer? And it said, "Alan, you wrote your book. We helped you, but these are all your ideas, your thoughts, your words.

You edited it. You hired human editors to edit it on top of that. " No.

But I got to tell you the truth, the fact that I did use AI to help me, I put down that I did, and I put down that I used Perplexity and Chat and Claude, but it was heavily human written and heavily human- Right ... edited by multiple humans. And if they want to ding me on that, so be it.

But I know I wrote my book, but I'll put the same thing out with music. And I think there's some interesting opportunities here as well. You see, for example, Claude is putting in watermarks even into plain text.

Yes. I don't know if you saw it. It's really interesting.

Watermarking plain text just by changing up their algorithmic choices into non-organic choices that can then be subsequently detected. You don't have to change the quality of the music to make an effective watermark. And I think that could be very positive because this does come back to creatives.

This does come back to recognition and actually the ability for creatives to pay their bills. Look, it's not like music pays that much for that many people. I'll tell you a little story.

" And he loaded it up around... It was actually last year, sorry. He loaded it up around Christmas time onto Spotify, and he got so many hits, he was making a living out of it because people would search for Mariah Carey, and they wouldn't search for her, they'd search for the title of the song.

His would come up first, and he gets the playwrights. Now, I'm not exactly going to cry in a bucket for Mariah Carey's income, but real struggling creators who have got one or two songs. I've got a good mate who's a digital musician.

He gets peanuts out of Spotify, and with this sort of thing, he'll get even less, and his music is entirely analog created. No AI at all, Alan. Not even helping him.

No, so here's the thing. I do think we're going to develop markets for totally analog, if you want to call it that, analog created content. Whether it's music or words or art or what have you, there will be an audience who, much like you could buy a Swiss-made watch made by the finest Swiss watchmakers.

You pay tens of thousands of dollars. Or you could get the J... This goes back to I'm dating myself, the Japanese.

Yes. Who made mass market, mass producing fantastic watches, which are now not very cheap either because you could get much cheaper ones in other places. Well, or you could wear a sundial on your wrist, right?

There you go. Or a digital watch. There you go.

Back to the original. Yep. Again, we're at a crossroads.

I think, Liz, you said the word crossroads. We're at this crossroads where I think new markets are forming. There will be opportunities up and down here, some bigger than others.

Conf where I am in terms of how are we going to do security? How are we going to do observability, Andy? Something very near and dear to you, I know.

Right? How much should we let the agents do? How much should the human be involved?

Right now, I think what we've seen is a knee-jerk reaction to, hey, all this AI's just going too fast for us. We haven't wrapped our heads around this, and we got to think about it. And I don't know how realistic that is.

I just don't. You got to build the plane while you're flying it. That's what the problem is.

Well, that's the thing, and we're going to have to get comfortable if someone else is flying it. Yeah. So anyway, with that, let's move over to point three.

Okay. This is a feel-good story? Yeah.

Yeah, this is going to make everybody happy. I don't pick the segments. If it's up to me, I would have...

But there was a lot of backlash around these flock cameras. Yeah. Going back a couple of weeks ago.

Terry, I know this is a story you wrote on. Yes. There were demonstrations near it, around it.

I guess if you're not doing an anti-data center demonstration, you could do a flock camera one. But people are fed up that this, again, we're being watched by... It's one thing to be watched by Big Brother.

Now it's AI Big Brother. Yeah. But what about- Well, this is the thing.

People are alarmed, I guess, and thinking we're turning into some sort of surveillance state. I believe that ship has already sailed- Yeah ... anyway.

But yeah, so Flock faced a whole lot of criticism in recent months that law enforcement was abusing the data collected from these Yeah. I know, and Andy- I can see that ... I see your shocked expression.

Yeah. I was equally as shocked when I started doing this. It was a real end of innocence for me to write this story.

But yeah, so there's a big hue and cry. Flock says, "All right. We're going to update our platform.

We're going to make some updates. We're going to do some things that should satisfy the privacy advocates and the people who are concerned," blah, blah, blah. And that's all well and good, and I'll tell you a few of the things that they did, but the ACLU stepped right in and said, "Hold it just a minute.

We don't believe you to begin with, and we don't believe most of this stuff is anything new. " Now, the one thing that they were keen on is the shortened data retention period. The seven-day thing.

Mm-hmm. Yeah. Well, so it shortens it, not as much as the ACLU want.

I think the ACLU wanted two days, 48 hours or something like that, and this is I believe seven. I can't remember what I wrote, but anyway, so it's shortening. " And then it can stay on ad infinitum.

It can stay on at the local levels and whatever. And then- It goes on your permanent record card. It goes on your permanent record.

There it is. Until AI erases it, yeah. That permanent record that your mom wrote- From elementary school.

Yeah ... there we go. So there's some other things.

They want police officers to have to input a case number or a code so the searches are legitimate, it's tied to a legitimate case. It's not just some officer wanting to mess around with whatever or going on a fishing expedition. I don't think the ACLU thinks that's going to be a glad to see it, but they don't think that's going to necessarily save the day here either.

And so I think that it's interesting. The ACLU has accused Flock of basically treating all of this like a big PR ploy, which, duh, again, shocker. Andy, where's your shocked face?

Yeah. I need that. I need an emoji for that.

Yeah. So evidence mode, it depends on how that evidence mode is going to operate. That evidence mode is when it's going to be used as evidence in a criminal investigation going forward.

And there's a bigger problem here, too, about how we collect and use data. It comes down to that, and our comfort level. We'll go back to comfort levels.

Our comfort level with our data being collected and used, but then our comfort level with actually using collected data. So there's two sides to that, but... I don't know if you put this genie back in the bottle.

I don't think you really can. I think your better traction is to go after the abusers personally. Right.

Upstream, exactly, and that was my big takeaway from reading this one was why is Flock... Terms of use is just a check mark on a screen. It means nothing based on enforcement, and Flock isn't the ones that should be enforcing it.

How about police stations? How about mandatory firing pending a review of... It's our enforcement mechanisms, the checks and balances, which I think is a theme across all these three stories- Yeah ...

needs to evolve for the age of AI, and it can't just be with the AI vendor or the AI labs, and I think or the existing political structures as they currently are right now. And that's why I think all this public pressure is so interesting because we, the people, kind of have an opportunity to reshape this by when the labs and businesses hit hard limits. So I just encourage us to think about what would we want?

What would put some checks and balances in this system so that police officers are not incentivized to where the risk of getting caught outweighs the temptation? Well, that's right. So some punishment here that fits the crime.

I think you're right about it shouldn't fall on Flock's shoulders, but they're not off the hook because there are things that need to be there. Entirely on Flock's shoulders. Yes.

Yeah. No, entirely on their shoulders. And yeah, so I think there are things, and one of the things that they're doing is you'll get banned from using Flock if you're caught doing something suspicious.

Pending a board review, whatever the board is. So a user will be blocked. The user will be blocked, but that's something that I think they needed to include or make sure it was solid in their offering.

That you could do this, that you can literally just ban people when they hit certain markers or they have certain characteristics of a search or whatever. I think those are effective. I've been banned from OpenTable, but who cares?

But I think that- Road rage ... kind of things, those fall on those. Yeah, so there's always Resy now.

But I think that those kinds of things do, or those capabilities fall on Flock, and that they have to take it more seriously. They're out to make bucks, so I don't think they really particularly care about- I think one of the problems, though and it's almost a uniquely American problem. I was talking to someone about regulation.

This was for AI, but it lends itself to this just as well. If you allow, as we do in America, every jurisdiction to put in their own rule, you're only allowed to keep 20 days. You're only allowed to keep 10 days here, 30 days here, 15 days here.

There's 50 states, but there's literally thousands of jurisdictions. There's counties, there's towns, there's cities throughout the US, and that's just the US. And it's a huge burden to a company, even like a Flock or one to say, "Well, okay, in New York City, I could do 30 days.

But if I go to Nassau County right over the line or Suffolk County after that, one's 20 days and one's 10 days. And then if I go up to Westchester County, it's 45 days. " Yeah.

Even with AI, that's a lot to wrap around. Well, it is. We need some sort of standards around this.

We've been so bad at doing that in the past, whether it's through regulation or legislation or whatever it is. Especially legislation, we don't seem to be able to come up with those standards ever that make sense and also are doable and evenly distributed. That's the gap.

That makes sense and are doable. Yeah. That's the chasm.

And so- I think you're right, that's what we need, but I don't know how quickly we can get it. I certainly don't think we can get it with our current state. And Terri, I just want to highlight in your article, great article, great quotes.

There's a killer quote in there, and I want to dig into it just a little bit. It's from John Bamford- Bamford. Yeah.

I love him ... president of R Street. " And I want to dig into that specifically.

I'm a bit of a semantic guy. Judicial oversight, because the legislation, you're absolutely right, Terri. The legislation across the board, across all sorts of kinds of industries and problematic problems, has not been super effective, in America specifically, and I'll tell you why.

Because, at least in large part, the same people are writing the corporate rules as are writing the legislation. The corporations and lobbyists are literally writing the legislation to put in the loopholes to make sure they're not over-regulated. Whereas that's not quite as obviously happening in maybe some other jurisdictions like Europe.

So that's where judicial oversight is coming into play a lot more because the legislative oversight is, even there, insufficient. Yeah. And he does point to the case law surrounding cellphone location data.

And I think that was part of his point, that that case law existed in that judicial effort, and that's what he'd like to see here. Yeah, he's a sharp guy, Ian. Always has an interesting perspective.

But yeah. I thought that was very interesting, and- He is ... spot on.

But, yeah. So I don't know. People have to take to the streets.

People have to demand. Well, but let me play devil's advocate a little bit too. No.

I'm even shy, but I'm going to do it. No, I think- I think that there is... Look, we just passed 9/11's 25th anniversary.

Mm-hmm. There is a reason why we have gone to a constant surveillance state. Oh, yeah.

And it's not just the US, it's in Europe, it's in China. Much more in China. But it's all over the world.

Singapore. There's a reason, and there's good reason for it. The problem is you can't blame the tool.

Don't blame the game, blame the players. Mm-hmm. The abuse of it is what the issue is.

Well, it's the abusers. Again, and it goes back to that. Right.

That's probably the place. But we still need some standards. We still need some standards.

And I think the judicial part of it plays into, we need teeth when we're dealing with people who abuse these things, and their privilege, I guess you call it privilege. But that's a very slow process that needs to be set up. It's a hugely slow process, and we are very slow in this country.

And it takes political will. I don't know if you guys saw yesterday, the clown who runs the FBI now was testifying before Congress. Oh my gosh.

What a sight. " And he wouldn't answer it. He was too busy defending people who have been convicted of bestiality working for the FBI.

Defenders of bestiality. God, this is the greatest America. But this is the world, the state that we live in right now.

Yeah. But if you don't have the political will, we could talk about this till the cows come home. Well, maybe we should.

Well, but I think it also goes back to what makes sense and what's actually doable. Like Terri, when you were talking, I was thinking, well, what if Flock, just instead of blocking the user, what if they block the whole customer, the whole police department? They're never going to do that.

They're not incentivized to do that. So there's controls that might be effective that won't be implemented because there's no incentive to do that, which all goes back to, what do checks and balances in the age of AI look like? Well, I will say this, and I think it's because of the public outcry and because of the clear abuses.

And this isn't Flock doing it to them, I understand that. But a lot of police departments and jurisdictions canceled their contracts or didn't move forward with contracts that they were considering with Flock. But see, I always think that stuff is short-lived.

It's not anything that's going to really last because for a lot of reasons. And don't forget, they're also not necessarily canceling their surveillance program- No ... they're just using a different company to do it.

Using another company. They're going mobile. Which is why anything meaningful on the part of the company is unlikely to work.

Spot on. Yep. Excellent.

I'm depressed now. Yeah, I know. I've got to have a drink.

It's noon, I've got to go have a beer. It's- It's almost five o'clock here, I guess I can. I was talking to Corey.

Well, yeah, you guys in Iceland and Spain, it's late enough, how that is. Yeah. But it's only 10:43 here in Denver.

But here's the thing, I was talking to Corey Mitten about it, as I mentioned earlier. I think this is all part of the fact that we are living in a historical era right now. Agreed.

I lived through the dot com when the internet became real, and it was amazing. But the stuff now is far beyond that. This is beyond just civilization.

There's so much promise, but so much peril, is the word I'd think of. But also, it keeps me very silly. I can't wait to get up in the morning and find out what the hell happened next.

First I survey if I'm still alive or if AI has killed me. Yeah, you're just an AI recreation. Am I a simulation or am I real?

Yeah. But we're going to continue. We'll continue to do this.

And that's where we are. Anyway, hey, we're about out of time. Liz, Terri, Andy, what an international panel today.

Thank you. I know. Such awesome guys.

Thank you for coming on. Thank you for watching. As usual, we do this live every Monday to Friday at noon Eastern Time.

tv or our OTT app on Android and iPhone and Apple, Roku, Amazon, or on the YouTube, Techstrong TV YouTube channel. You can check it out there. We'll be back tomorrow with more.

com. We've got Patrick Debois, kind of the godfather of DevOps, who's going to do a live AMA, Ask Me Anything, about DevOps there. I am very close to having someone from Anthropic who wrote the playbook for using Claude in CI/CD do a talk there.

I'm hoping to make that official. I'll be presenting. Dan Newmann, you'll know Daniel Newmann.

Mitch Ashley. A lot of good people. So check that out.

I mentioned my book. That's coming October 6th. If I wasn't so far away from my backpack, I'd show you my first proof here that I've got.

But do check that out. We'll be talking more about that. And I'll see you all tomorrow on the gang.

But for now, we're out. The agentic AI race is on. Which approach will move your team forward?

Join DevOps Experience 2026 for practical insights into AI and DevOps. September 24th, free and virtual. com.