Google Called a “Parasite” as $1.5B Anthropic Settlement Reshapes AI Copyright Law

Hey everyone, it's Wednesday and it's time for Techstrong gang. Thanks for joining us. We've got a great show today.

I think I say that every day, but every show is a great show. But anyway, what makes it even better though is we've got some great people to talk about our stories for today. Let me introduce you to our gang members.

First of all, joining us, I guess down Florida way, my friend JP Morganthal. JP, how are you? I'm doing great, thanks.

Hi, everyone. Good to have you. And then, out west in Silicon Valley, our man in the valley, John Stortz.

Hello and hi, everybody. Hello, John. And then joining us, Terry, what did you do to that beautiful brick wall?

It's there, but I think because I hung a piece of art, it's blurring me out. Yes, it is. Yeah.

We don't- Maybe I'll remove it, so you can have the... I'm afraid you can't see. No, look, I'm a fan of the arts.

" So. Yeah. Oh, very.

I wish I could see it too, actually, but- Yeah, I don't think you can see anything but a big red- No, it's blurring it out. Well, all right. Well, gang, welcome.

Mike Fazard is still MIA. I have heard he's crossed over the border into Canada. I don't know what he's a fugitive of or anything, but I guess we'll hear all about it at some point.

If they let him back over. If they let him back in. He might be on the no-fly list.

But, anyway, let's jump in. It's a busy Wednesday as usual. I wanted to kick things off today with an article that I actually wrote.

ai. I put up a separate LinkedIn article that links to it, and I'll be talking about it on "Shimmy Says" tomorrow at 2:30. I've been on a binge lately writing about the changing relationship in the tech media.

Not only tech media, I think all media, but specifically tech. Between PR, and agencies, and vendors, and publishers, and writers. And I kept calling that we need a new basis, a new covenant, a new way of how we all work together.

" And we could say it nicely, but I'll say it the way I wrote it. Google's a parasite. Google has become a parasite.

The entire open web was based on a covenant of publishers, content providers, allowed Google access to their websites so that Google could index those websites. And when people searched for information, Google would serve up those sites as information, and people were free to click on them and explore the answers to their searches. And for publishers like Techstrong, that represented 80% or more of our traffic, what we called organic search, of which 99% came from Google.

That enabled a lot of things because that traffic, I always tell people, those eyeballs make the world go round. They allow you to get advertising dollars. They make sponsors want to come to your site because you're getting those eyeballs.

They allow you to do a lot of earned media interviews and maybe take in some stories that, look, you know you wouldn't have taken in a different economic climate. But something happened along the way in the last three years, two years. Google, and under pressure from the AI LLMs, decided to change the covenant, to break the covenant.

They went to a no-click policy, basically, where when you do a search in Google, they're still taking your information, my information, Techstrong's information, the information you're writing, and they're serving it up as if it's using it as for their own purpose on their own site, without the link out back to your original content. So they are taking your content and monetizing it without giving you the click back, which is what they always did. So this changed from a symbiotic relationship where you gave them information, they sent you traffic, to the classic definition of a parasitic relationship.

They are taking your information, using it for their benefit, and giving you nothing in return. Mm-hmm. And just weakening you.

That's a parasite. Now, the LLMs do it, right? They train their whole model- Yeah ...

on, in essence, stolen information. We're going to talk about some of that in block B. Right?

We'll get a little slap on the wrist for a couple billion dollars or whatever. But the fact of the matter is Google has broken the covenant, and it's not just about publishers and Google. It's about the PR agencies, the marketing agencies, the earned media placements, the vendors and sponsors who want to get their messages out.

It's about you, the end users. Why more of the information you want to have access to may be behind reg walls or paywalls and subscription? Because the economic model has been broken by Google here.

John, you've been in this business. Terry, I know you've been in the media business far longer than I have. Yeah.

I'm interested in your take. Do you want to go ahead, Terry? Because I have this long history of dispute with, not personally, but dispute with what Google has done to the media over the years.

I can be faster and you can go ahead and say whatever you want. Go ahead. I'm fine.

Okay. Sure. If you want to expound, it's good.

I'm not- Okay, sure. So, I'm going back to Google and Facebook siphoning all the advertising away from the old traditional media, which never got it, never understood what was going on because the old traditional media is living in the past and was used to a semi-monopoly and didn't know how to adapt to technology. I can attest to that at every place I ever worked.

But this one, I think is particularly heinous what Google's doing because in a sense, this AI overviews, they're synthesizing original reporting into these overviews, and they're capturing the commercial value while eliminating the user's need to click through to the source. So this creates this parasitic dynamic, or I think, we call it Google zero threat, where publishers bear the full financial burden of producing journalism and research, and Google is starving them out of the traffic needed to sustain it. And it goes across traditional digital publishing model, where publishers can no longer rely on anonymous rented audiences and must build direct relationships through registrations and memberships.

You mentioned PR, Alan. PR agencies and tech vendors are going to have to accept that unlimited free earned media is unsustainable without commercial support. And the readers must recognize that maintaining an independent, high-quality content requires providing their identity, engagement, or direct financial support.

So I think everybody loses except for Google, of course. Oh, I think Google notices it too. I don't think so.

" And keep in mind, the search business is still what floats Google. Right. Right?

All the ads and everything else. But, yeah. But it's, again, so lopsided, right?

That's the problem here. Because we've all been through, we've been around a long time, as you said, Alan, we've seen models shift and companies, as you said, John, try to sort of struggle with the new models and how to present. " Sometimes your name wasn't even clearly on it.

Right. Not attributed. Yep.

Right. Yeah. And so we've been through this before.

This seems somehow more heinous, possibly because of the way Google is such a large entity, and they've got such control, and we talk about them being everywhere with their tentacles in everything. And it seems to me, and maybe I'm missing something, but it feels like we have lost what used to be, as you said, the symbiotic relationship with this. And how do we get that?

I don't know what the answer might be to this because what kind of leverage do publishing companies, do writers do whatever? What do we have to... I'm not saying we're completely defenseless because I don't ever believe that about myself or most of you guys anyway.

But what- Everywhere I worked, I never had control over any of my content. Yeah. That's the thing.

I started out as probably some of you did at print, right? Where you just, you wrote it, it belonged to them. They printed it in their newspaper, magazine, whatever it was, and that was kind of it.

And I guess maybe if you worked for some publications, you got, depending on what kind of writing you did, you did some sort of royalty agreement, but that wasn't for all writers. That was a small fraction of people. But what do we do?

What's- Well, I have some ideas, but I wanted to ask JP because of the four of us, JP is not necessarily a content provider, right? JP is a CTO. JP is a man on the ground with technology helping companies adopt technology.

So he represents sort of the consumer, if you will. JP, what do you think? For years, I lived by content as well.

I was producing a lot of articles for magazines and publishing books. So I look at it from both angles. And you're right to say that what you're seeing looks like X from Google.

Can't deny that their business model is affected. I think part of the issue is that they live in both worlds, and they have to adapt or die Right? They created the largest, most popular search engine in the world, and it was all based off of HTTP.

And they built the system to go and index and crawl and catalog that data and then expose it behind an advertising scheme. Brilliant. Times change.

They are also one of the largest and biggest contributors to the AI movement. So you can't have a foot in the AI movement and try to protect your turf in the HTTP search world. The two are going to conflict.

They are by nature, one is cannibalizing the other. Now, in business, I've been in so many businesses where we talk over the years, cloud initially, and now AI. If you don't cannibalize your own business, somebody else will.

So while you're not incorrect in your description of what you're seeing happening to content providers because of Google, looking at it from their perspective as a business, they have no choice but to cannibalize their existing business or die. Now, their future is predicated on being an AI solution, systems, inference, whatever engine provider. That's where the future is.

And so they're now in that space where they're growing that and they're looking at the impact to their existing advertising world. Now, interestingly enough, they are not completely moving from one to the other. What you wrote about is the summarization, right?

It's like I do a query. Here's all your links. Here's all the sources.

You can go directly to them. All I'm going to do is I'm going to show you what my AI engine evaluates to be the summary of that content. Now, in the future, is that slowly becoming the master for what everything is going to be?

Where it goes out, it does the searches, it brings back the summarization. You don't see the pictures. You don't see the full content.

You don't get the full context. You get this summary. And I think as humans, the question is, are we just going to absorb the summary and not look at the detail?

I think we can say based upon our political stature in the United States at least, if not worldwide, and what we've seen, people tend to read a blip and it becomes their reality. I think that's the awful downside of what could happen, is that people read these summaries, they don't have the full context, and they're not taking the time to go look at the sources which are provided to evaluate for themselves whether that summary is accurate. So there's a tremendous danger there.

But they are at least giving the sources and allowing you to delve deeper to say, "Oh, this came from that. Let me look into the detail here. " Hmm.

Oh, Alan. Oh, sorry. It was interesting you distilled it really well at the beginning, near the top of the story.

You talked about how Google used to pursue the path to a destination. Now they want to be the destination, period. Right.

I think that's really it. We don't have a lot of time more on this particular segment, but as I said, this affects the relationship not just between Google and the publisher, but it affects, John, your relationship is you're both an editor, writer here. Terry writes here.

It affects your relationship as the writer. It really affects the relationship and business of the PR agencies, because these PR agencies go to these clients and say, "Hey, if you pay us $15,000, $20,000, $25,000, $30,000 a month, we're going to get you mentioned. com or Security Boulevard or what have you.

And now they come to these publications and we can't afford to do their earned media unless it's something really, really newsworthy because we've got to do things that are keeping the lights on here. So, you started asking for what we've done with our Techstrong Council and some of the other programs we've put in place, and these PR agencies say, "Oh, no, we don't want to do any paid. " Those days are over, too, because of this, right?

The repercussions go right through the whole food chain. I heard a lot of complaints about that, actually, at RSA- Yep ... when I was out there, and the whole idea of the paid versus earned and this pushback against paid, but also with some, an acceptance even if it came with reluctance, like a resignation that it's- Well, here's the fact.

It costs Techstrong about $75,000 to go to RSA. Between the broadcast alley booths and the flights and the hotels and the meals and the union- Right ... shadows and all that, it's a 75 to $80,000 proposition.

Right And I should maybe qualify too, that's a super valid point, obviously, but they weren't actually railing against any of us on Techstrong and in particular, I'll do it- No, but the Techstrong's no different. Look ... Yeah, but there were some outrageous, or what a lot of people considered outrageous pay-to-plays.

Yes, there were. And, that's I think what... And I mostly just didn't get in the conversation.

People would bring it up to me, and I heard it from so many people, I was like, "Wow, this is becoming a thing, and people are really thinking about it. " But- Someone has to pay for those fancy cameras and lights over there. No kidding.

But I know who they're talking about. But look, we try to do our best, but we're over time. I've got to move to the next segment.

Mm-hmm. But guys, look, this is not a story that ends here. I think it's a story that is, it made it all the way to The Times.

The New York Times had an article this week, How Google's AI Search Is Imperiling the Open Web. It goes exact, categorically against as recently, I think, as 2004. Four, yeah.

Right? When Larry Page wrote to pre-IPO shareholders. Right?

About who Google is and the do no evils and all that. Right. The other important thing too is these companies are created to be the anti-companies they're running against.

So when Google was created, "We're not going to be like Microsoft," and they always end up being just like the companies they thought they'd- And that's exactly where they end up. Yeah ... never be like.

Yeah. And so there'll be somebody who will challenge- Ain't it the truth ... Google, and be the next Google.

Ain't it the truth. All right. Let's move on to our second story today, though, and this is along in the same vein, right?

Right. Google's not sending clicks, and the LLMs built their whole foundation on stolen data and stolen content. That's right.

Now, some people did sue, and Anthropic, who prides themselves on being the good guys, it looks like they've done a settlement. What have we got here, Terry? Okay.

Well, there you go. It's the authors and the publishers versus Anthropic in the courts, and the authors and the publishers, at least this time around, seem to have won. 5 billion class action copyright settlement, but I'm going to tell you, as we were discussing sort of before we got started here, that is a pittance when you think of it.

It's even less than a rounding error, I think, for a company like Anthropic, right? And for the people affected, the people that brought the suit, that's like $3,000 per book. Right?

Per copyright. So, depending on who you are, that's not a lot of money. You're not retiring off of it.

No, you're certainly not going to retire on that. But yeah. But let me tell you, Terry, who did make the money.

The lawyers? As usual. The lawyers?

The lawyers. $101 million. That's retiring money.

Now there, you could retire on that, for sure. The other thing, though, that's interesting about this is this lump sum settlement is primarily, I think, is motivated as much as anything by Anthropic trying to clear their massive liability risks off the board ahead of their IPO. So they want to get rid of this.

This was a settlement rather than a binding appellate ruling, so there's still this whole overarching argument about legal boundaries between fair use and generative AI is still unresolved, and there are pending lawsuits against OpenAI, Meta, and Google going forward. So yes, it is a pittance, but I think it was kind of a IPO-related strategy, and yeah, 3,000 books. $3,000 a book sounds like a publishing deal for most authors.

It's just nothing. Yeah, we'll give you $3,000 for- Yeah. I can attest to that ...

Yeah. We'll give you $3,000 for all your blood, sweat, and tears on- Exactly. Yeah ...

this thing. Yeah. Another way of looking at this, though, is a billion and a half dollars, that's an awful lot of money, even by today's standards, right?

I don't know if-- It wouldn't buy you a data center or anything like that. But it's still a billion and a half. And I think it's the largest copyright recovery- And it's the largest what.

Yeah ... in US history. Yeah, it's the largest ever.

But it just gives you... Who was it? It gives you insight into just how large the theft here was.

I don't know another way of saying it, right? Yeah. When you've got a billion and a half dollars, but the class of people involved, that they each get three grand, $3,000 each, it gives you an idea of just how big- Half ...

of a- Yeah, half a million books ... theft. Yeah ...

basically. And what's interesting, though, is I think that we're going to see other types of lawsuits, and we actually just saw one yesterday where the University of Tennessee Research Foundation filed a patent infringement lawsuit against Anthropic. I think it's believed to be the first patent-related legal challenge against the company, and it claims they incorporated patented neural network technology into their commercial AI systems, including Claude.

So maybe there's some more lawsuits coming, not just in publishing, but in terms of actual properties. Yeah. It's interesting.

Interesting. Totally. Yeah.

JP, you wrote some books. Are you in on this one? No.

I missed it completely. I wish I had. Yeah.

I assume that there is going to be some basis for determination about whether the content is viable or for your participation in the suit, I assume that they are looking for some attribution or content that is in the corpus right now in order to- No. I believe, actually, Anthropic, as part of discovery, had to produce a list of material that they ingested. Yeah.

You didn't necessarily have to go do a prompt and search for it. I think Anthropic had to produce the list of the books they used. Did you get a little card in the mail, though, that said, "If you have ever written a book" ?

Have you written a book? Right. That I wish I had.

Usually, that's what you wait for, whether you're part of the class. Yeah. Right.

Yeah. If I had, I'm sure those lawyers would've hunted me down. So there are all these issues, but I also want to talk about maybe one other thing, too, here, and sort of in my area of cyber.

Do you think, too, that AI training data is going to have to be treated like supply chain now? I'm just kind of curious how it's going to-- By the enterprise security team, I'm talking about. Right.

Well, it's a logical argument that my AI is going to spit out based upon what it's trained on. Look at the food it eats to see what it is and- Hopefully not lettuce. Well, or from Taylor Farms.

No, it's not Taylor Farms. He's a big supporter. Anyway.

So yes, I could see it be sort of a supply chain issue. Where, what was this LLM trained on? Now, for the big ones, these frontier models and some of the bigger open weights, it's harder, it's more abstract.

But as we continue to move towards a world where you have smaller language models that you're running specific tasks on, what those models ingested, what they were trained on, I think becomes more and more important. Yeah. And there is movements all around that of using open data.

Yeah. So that there's a traceability back to the sources for what information and what data was used for training. There's a lot of activity around these particular actions within the various frontier model providers.

So are companies also going to have to maybe put in some filters here, too, so their users don't use copyrighted data, or so the models don't spit that out? Like if that's a- Supposedly there are filters But that doesn't seem to be the case. It would be content generated based on what it was trained on versus sourced data that is quoted in a particular generated data or text.

Right. So I figured that quoted-- But I'm just curious what that means in practice within an organization, and what kind of boundaries- You're right. " So last night I talked to this USC graduate class on digital transformation and digital publishing, and they were asking questions.

We talked for about 30 minutes. They were asking questions about AI and its impact on them. " Some of these folks want to create films, or they want to write books or write screenplays.

And their greatest concern was, how protected am I? And how do I protect myself? And this decision has just come out, and I think that's-- They're concerned about getting a job first and foremost, but they're also concerned about what they create.

Well, does it belong to them? Well, I thought- Great question It's just a matter of it getting credited for it. Yeah, exactly.

That was the thing that they were saying. "Okay, we understand we have to work with AI. We're going to adapt to it.

We're going to align with it. " Well, what is the thing you used to do? What was it for-- You'd have to tag-- Oh my gosh, now it's completely eluding me.

Oh, if you were an influencer or something like that, maybe before they were called influencers. Mm-hmm. People worked on all these little tags within articles or whatever that would come up and would say that's something you were affiliated with, like that you might be talking about.

So you like Aquafina water. If you happen to mention it in your whatever, to keep you honest, it doesn't-- You wouldn't say- Analysts ... analysts.

Financial analysts are the- Analysts ... the only ones that were forced to- Yeah ... adhere to that rule legally.

I just wonder-- Yeah, it was financial analyst. Well, I think it was others, but definitely- It was others ... financial analysts, yeah.

And so I just wondered if there would be, for what you're talking about, John, something- Yeah ... like that, too, that you wouldn't rely on anybody else to-- It would be automatically tagged or whatever. But I don't know.

Anyway, it's an interesting ... thing. 5 billion is not considered.

It's a big enough number to indicate a big crime, but it's not a lot of money in your pocket when you get down to it, and it's certainly not a lot of money for these guys to cough up. It's a lot of money, but it's not. You're right.

I mean- I think the headline for this story is pittance, and that's probably right. Yeah. Yeah.

In the bigger picture of things. But look, if Anthropic puts a billion and a half in, and OpenAI throws in a billion and a half, two billion, Google throws in, before you know it, you're talking about real money. Yeah.

On that note, let's jump to block C here today, talking about spending and money, a big tech spending surge, at least according to Forrester. John, this is an article you did. What do we got?

Yeah. So we have some conflicting viewpoints, too, between what Forrester thinks and what the Futurum Group thinks. So basically, Forrester talked to 2,600 folks, and I'll just give you a quick summary.

More than 80% of the business and technology leaders anticipate budget increases for the 2027 fiscal cycle. They say it's propelled by this unpredictable operational cost rather than pure economic expansion. And their reasoning is the transition from predictable per-seat software licensing to usage-based tokens and credit consumption is going to wreak havoc to a certain extent.

Now, Futurum's read is different. They say AI pricing is an accelerant, but it's not the trigger. The real driver, we say, or our parent company says, is the pilot to production maturation curve.

AI's moving from a line item to an embedded cost across the entire stack, which not because it came from the Futurum Group, but I think that makes more sense to me. Mike and I always talk about these reports and we nitpick them. " And in this case, I think I'm going to side with the Futurum conclusion more so than Forrester.

Although Forrester, I'll give them credit, it was an interesting study. But yeah, no, I think it worked here. So regardless of who's right or wrong in terms of what's driving it, I think the thing they all agree on is the significant spending going up.

Mm-hmm. Right? Where it's driving up.

And it's interesting because at the same time you hear that marketing is probably one of the hardest hit departments in terms of budget cuts and layoffs. Not budget cuts, but headcount, layoffs, and stuff like that because they think you could do so much of your marketing using AI and so forth. So this is a real problem, right?

That's I think the crux. Whether it is purely AI or is it moving AI from lab to production, one could go back and forth. But the irony of so many of my friends in marketing who are out looking for jobs right now, while it seems the budgets are going up here.

Mm-hmm. Yeah, judging by what I'm seeing on LinkedIn, right, with people- Mm-hmm ... looking for jobs and then being out of work for substantial amounts of time, too.

Yeah. It's not just, "Oh, it's taken me a month or so," and whatever. There's not a lot to be had.

I wonder if there'll be a backlash, though, and that more human touch will be craved again at some point. Well, you see that. Every day in my feed in LinkedIn, I see people saying, "If I see...

" Not the spine. "Here's the load-bearing wedge," or whatever, right? We've moved beyond the EM dash as the AI tell, right?

Yeah. There's certain phrases that are clearly AI tells, and we see them, and people are revolting against them. I do an interesting experiment with some of my posts on LinkedIn.

Some of them will include the AI-generated image. Some I put up without an image. Hmm.

And see does it work better without the image, or does the AI image- Yeah. Did you come to a conclusion of how effective it was or- I've only done it like three or four without the images. And they've done well, but I need a wider body of information before I think I could tell you for sure.

That's an interesting little experiment there, though. Yeah. Well, I have nothing better to do.

I don't know. These are the things I think about. Is this it?

Is that... I don't know. No.

I think it's interesting. Dealing with- You hit on something there, Terry, that I wrote about this a while ago, but this whole idea of looking for jobs, and we did a segment on this a few weeks ago, and it's mainly more about who you know. Mm-hmm.

Human-to-human interaction when it comes to finding a job because some of these ghost jobs that are listed haven't been filled for two years. Yeah. I can now mention three or four of them off the top of my head that I keep getting poked about, and I just delete them because I don't think they exist.

Yeah, that's right. They don't exist. Or they're either there to lure you to something else, but most of the time I think they don't exist.

Sometimes this is an old journalism trick, right? " You don't get the job, but they use your ideas or they use your writing sample. Oh, that used to...

Yeah, I can remember- And that's happening with these jobs in marketing. They're asking for marketing plans in some of these companies I think years ago, it wasn't that long ago, it was a few years, somebody asked me for some ideas and they said, "Can you work one of them up? You don't have to go whole hog on it.

Just give us-" No. " Right. I'm not giving you- Good for you ...

my ideas, nor am I, even worse, writing up half of a feature for you so you can run with it if I don't get the work. It's kind of audacious. When you think about it.

JP, you were recently on the job market for a while, though. You've run into this stuff, right? What's your take?

Well, I don't know how we got from here to this topic. I'd rather talk more about tech spending, but look, we've had these discussions. Andy Mann and I have spent hours talking about this.

Andy is another guest on the Techstrong gang. Yeah, another gang member. The system's broken.

Why? Nobody really knows. It is broken in the worst possible way in that there is demand, there is supply, and yet demand can't find and talk to the supply for whatever reason.

It is a broken supply chain. The two don't meet. It is interesting.

So I have just one quick example. A guy that I know has been looking for a little bit, not exceedingly long time, a couple of months, whatever. And he was talking about on LinkedIn, "I see that a lot of my former coworkers went to such and such company, and that's great.

" And he said, kind of was saying he wished he had gone. And then what pops up? I get the little ting, that there's this job that actually matches his qualifications almost down to the thing.

And- Just coincidence ... Yeah. But I sent it to him and I said, "That's so weird.

If you've put in some sort of application here, whatever, you guys are like ships passing in the night. You're not connecting. " But he said, "I didn't get that.

" I said, "Well, Michael, that's what-" So going- Part of the distrust factor, it seems like the only way to get any momentum is to have somebody that is working at the company or knows the hiring manager personally- Mm-hmm ... to make a recommendation. Exactly.

That's the way it happens. It has to. Going back to the Forrester, though, Forrester report, if these IT managers or executives expect a bump in spending, does that run counter to this fear that AI is going to be costing jobs in the short term?

I wonder can the two coexist in a certain way? I don't know. Honestly, I don't know.

Something's got to give. Yeah. Yeah.

Yeah. No, something does have to give. Look, here's the things that I think we got to look out for.

I think if we're really going to go to usage-based AI pricing, AI has to become economical. If it's not economical, people aren't going to chase it with budgets for things that humans can do just because it's AI. And so I think that's really the linchpin that breaks this whole impasse where we are, where we've got to allocate more money for AI.

We got to allocate more money for AI when AI's getting real expensive. That only goes so far because the promise of AI, there's this paradox, Jevons Paradox, that the cheaper something gets, the more easily accessible it is, the more use it's going to get. Right?

The more the barriers break down, the more use it gets, and then the more use it gets, the cheaper it gets still. And that dynamic, that paradox is at play here, right? As more companies adopt AI, as we use AI more, the thought is or the gist is it's got to get cheaper.

Mm. If it doesn't get cheaper, this whole model starts breaking up. This whole thing that we're in, this whole cycle is going to break down.

Today's Wednesday. I think it was Tuesday, Gemini or Google announced three new versions of Gemini, which are based on price- Yeah ... undercutting the prices- Scared, right?

of their competitors. Right. Yeah.

Well, it's a whole open- It's going that way ... open weighted models we've seen this week from China. They're going to exert downward pressure on pricing because that's the way these cycles go.

" I think that could be probably recycled and used for a number of things. Oh, yeah. We'll use it later.

We'll steal it, yeah. That's good. Here's the thing, too.

We talk about marketing within companies. I was mentioning to somebody, the frustrating thing to me and to the general market is the companies that tell us that the AI story do such a bad job of conveying what the narrative's going to be like because they don't know. And they say, "Give us two or five years.

" But what are they? " So I think you could go back in a year to any of these trade shows, and you can see that most of it's going to disappear or be re-massaged into something else. It's not going to be there.

And the same people who told you story A are now telling you story B. We do need people who understand AI intrinsically and can help businesses to properly implement and leverage the capabilities. And that is an ongoing role because these systems, these engines will change over time.

The technology's going to move very quickly, and companies will need people who can help them keep up with the changes and how to not get taken out by them. Like we've seen. 6 million bill in a month for inference.

It's how do I control that? I talk to customers all the time. I have great algorithms and great methodologies for ways that they can manage that better and reduce costs significantly.

They don't know that they exist or they don't know how to implement them. " And we're looking at the level of, in my opinion, Great Depression-level job loss because the market can't absorb the removal quickly enough. People aren't going to be skilled.

They have to be retrained. The new jobs have to emerge. That gap, that change in that timeframe between when AI begins to eliminate more jobs versus when jobs appear for people based on the new AI economy is going to be awful.

I guess we're going to have to see, but we're out of time today, guys. Hey, I hope you've enjoyed this at home or wherever you're watching it from. Terry, John, JP, thank you for joining in on "The Gang" today.

As always, thank everyone for watching. As usual, "The Gang" is available Monday to Friday at noon on Techstrong TV and our Techstrong TV network, including our YouTube channel, Techstrong TV YouTube channel, and our Techstrong TV OTT app, which is on iOS and Android and Roku and Apple TV. And if you see us quoted in the summary on Google, click the source.

Click the thing. Do me a favor. Support your publishers.

We'll be back tomorrow with more. Thank you everyone. We're out.

Bye.