Latest / AI Fire Daily / 🎙️ EP 294: US Government Demands AI Equity & Data Shows Public Trust is Plummeting
Transcript
- 0:00Half the country is using AI daily, yet nearly
- 0:03everyone is convinced it's going to ruin society.
- 0:06Welcome to the paradox of 2026. It's absolutely
- 0:10wild. We're outsourcing our daily emails to models
- 0:13we secretly fear. Welcome to the deep dive. Today
- 0:17we're unpacking the G7 summit's shift toward
- 0:19government AI equity grabs. Right, and we'll
- 0:22run through a rapid -fire update of the newest
- 0:24models hitting the market. Plus, we'll explore
- 0:26that massive disconnect between AI adoption and
- 0:29public trust. Now, neutrally speaking, there
- 0:32are several different political proposals right
- 0:34now. The Treasury Secretary wants to use AI equity
- 0:38to fund Trump accounts. Right. But the Commerce
- 0:42Secretary prefers routing equity into a sovereign
- 0:45wealth fund. And Senator Bernie Sanders proposed
- 0:47a totally different law. He wants to give Americans
- 0:50direct ownership stakes in top firms. So the
- 0:54source facts show different paths to the exact
- 0:56same idea. The public sector wants financial
- 0:59value directly from private labs. Exactly. The
- 1:02specific political flavor just determines where
- 1:05the money actually goes. The timing of this is
- 1:07incredibly suspect to me, though. Oh, totally.
- 1:09These talks happened right before the fable and
- 1:12mythos restrictions hit. You know, that massive
- 1:14export control clash. with Anthropic. Yeah, Anthropic's
- 1:17Dario Amadei and Google DeepMind's Demis Hassabis
- 1:20were pushing hard. They want a robust U .S.-led
- 1:23coalition on AI rules and chips. Right. They're
- 1:26trying to shape the cage before they get locked
- 1:28inside. Exactly. The industry's best case scenario
- 1:31is getting Fable 5 back online. And they want
- 1:33to do it without aggressive power grabs. I mean,
- 1:36there's also talk of naturally slowing down the
- 1:37release cadence. This whole new model every month
- 1:40schedule is just fundamentally broken. It exhausts
- 1:43the developers entirely. And it severely confuses
- 1:47everyday users trying to keep up. I completely
- 1:49agree with that exhaustion. I mean, I still wrestle
- 1:52with prompt drift myself. Oh, yeah. We all feel
- 1:55that pain. You know the exact feeling. Your go
- 1:57-to instructions, just quietly stop working under
- 2:00the hood. Yeah, it breaks the user experience
- 2:02completely. But zooming out, we're talking about
- 2:04frontier models here. Let's define that real
- 2:07quick. Frontier models are the most advanced,
- 2:09cutting -edge AI systems currently available.
- 2:12Right. And the worst -case scenario from the
- 2:14sources is pretty dark. Imagine frontier models
- 2:17locked entire... behind massive government gates.
- 2:20Oh, wow. Yeah. Open source development gets banned
- 2:23completely for national security reasons. So
- 2:25the market just faces zero actual competition.
- 2:28Exactly. We end up paying whatever the big labs
- 2:30demand. Or we settle for vastly subpar intelligence
- 2:35in our daily tools. Are we heading toward a world
- 2:37where AI is treated as a public utility or just
- 2:41a government sanctioned monopoly? It looks like
- 2:44they want the broad structure of a public utility,
- 2:47but the control remains tightly centralized among
- 2:50a few key political players. Public utility vibes,
- 2:53but entirely controlled by government gatekeepers.
- 2:57Beat. Precisely. And that central control is
- 3:01what labs are fighting. The government clearly
- 3:03wants to slow things down right now. Right. But
- 3:06the tech labs are sprinting in the absolute opposite
- 3:08direction. They really are moving faster than
- 3:10ever before. GPT 5 .6 and the pro version might
- 3:14actually arrive next week. Yeah, and early testing
- 3:16shows their logical understanding has improved
- 3:18massively. But the coding limitations are still
- 3:21a stubborn bottleneck. Yeah, front -end and web
- 3:24dev are definitely not fully solved yet. No,
- 3:25they still struggle with complex architectural
- 3:27planning. But Anthropix's Chris Ciari is very
- 3:31confident about CloudFable 5. He believes the
- 3:33models will return to the market very soon. And
- 3:36the tool updates are just as aggressive. Perplexity
- 3:38Computer just received a massive brain upgrade
- 3:41for max users. It literally makes you want to
- 3:44provide more content. text every day. It learns
- 3:45your specific workflows perfectly. Google's pushing
- 3:48hard into the enterprise ad space too, right?
- 3:50Yeah. Google's new Ask Ad Manager agent can troubleshoot
- 3:53complex marketing campaigns. But we're also seeing
- 3:56some fascinating cultural shifts around this
- 3:59technology. Match Group recently released some
- 4:01wild data about Tinder. Oh, right. 47 % of U
- 4:06.S. singles feel negative about AI in dating.
- 4:09Which isn't surprising. You don't want a robot
- 4:10writing your romance. Right, but look at the
- 4:12massive contradiction in their behavior. Many
- 4:15of those same singles still want help with their
- 4:18profiles. They hate the concept, but they love
- 4:21the practical utility. Exactly. They still want
- 4:23AI to help navigate awkward intro chats. On the
- 4:27corporate side, companies are getting very creative
- 4:29with structure. Snap is spinning off its internal
- 4:32AI video team as Dotmo. Which just sounds like
- 4:36basic corporate cost cutting on the surface.
- 4:39But Snap is keeping a large equity bet in the
- 4:41spinoff. Right. They want the upside. Without
- 4:44the massive operational burn rate. Right. And
- 4:46that burn rate is driven entirely by the hardware
- 4:49bottleneck. Absolutely. Amazon is apparently
- 4:51considering selling AI chips to outside companies.
- 4:55A challenge NVIDIA. Yeah, they're trying to carve
- 4:57out their own massive turf. AWS CEO Andy Jassy
- 5:01estimates a $50 billion chip run rate. $50 billion?
- 5:06Beat. That's a staggering amount of hardware.
- 5:08It's almost impossible to comprehend that scale.
- 5:11Whoa, imagine scaling to a billion queries. Two
- 5:14secs silence. It requires a fundamental rewiring
- 5:17of our global energy grid. And that compute power
- 5:20fuels incredibly ambitious new startups today.
- 5:23Like Odyssey, right? They just raised $310 million.
- 5:26Yeah, at a $1 .45 billion valuation. And they're
- 5:29building world models. Basically, AI that simulates
- 5:33real world physics and complex human behavior.
- 5:35Exactly. It's rendering a universe, not just
- 5:37writing an essay. So with world models perfectly
- 5:40simulating physical reality, is the hardware
- 5:42demand ever going to plateau? The physical world
- 5:45has infinite resolution and infinite complexity.
- 5:48Simulating it perfectly requires exponentially
- 5:51growing processing power. It definitely won't
- 5:54stop anytime soon. Not until the physical and
- 5:57digital worlds completely merge. That's the ultimate
- 6:00endgame for these massive frontier labs. But
- 6:03let's bring this back down to the human scale.
- 6:05Yeah, let's do that. That massive hardware bottleneck
- 6:08is exactly why everyday users feel squeezed.
- 6:11Which brings us to the microtools patching specific
- 6:14daily gaps. These microtools are solving specific
- 6:17friction points for regular workers. Seaton's
- 6:202 .0 Mini is a fantastic example of this trend.
- 6:22Yeah, it generates high -quality AI videos in
- 6:25under a single minute. Faster, cheaper, and they
- 6:28offer 30 free credits. It completely democratizes
- 6:31video generation for the average creator. And
- 6:33then there's Upstream, completely changing the
- 6:35daily administrative grind. It sorts your messages
- 6:38and drafts your replies automatically. It makes
- 6:40checking your email feel light, fast, and fun
- 6:42again. Honestly is another tool providing massive
- 6:45value to modern digital marketers. It pulls sentiment
- 6:48insights directly from Reddit. TikTok and X.
- 6:51It also scrapes data from YouTube, Instagram,
- 6:54and Facebook automatically. You get unfiltered
- 6:57feedback without spending hours reading angry
- 6:59comments manually. I want to highlight Tabstack
- 7:02specifically for the developers listening today.
- 7:04Tabstack is a highly specialized web data and
- 7:07automation API. You pass a URL and it spits back
- 7:10pristine JSON. Which is just a simple text format
- 7:13for storing and moving digital data. It delivers
- 7:16highly reliable structured output from a single
- 7:19URL. It completely eliminates the parsing errors
- 7:22that plague legacy web scrapers. Are these micro
- 7:24tools permanently replacing whole software suites
- 7:27or just bridging temporary gaps? Right now, they
- 7:30act as brilliant little patches. But as they
- 7:33interlock seamlessly, they will obsolete massive
- 7:35legacy platforms entirely. They're bridging small
- 7:38gaps today, replacing entire suites tomorrow.
- 7:41Beat. Yeah. So let's recenter on the actual human
- 7:45beings using these platforms. Yeah, that's arguably
- 7:47the most important question we face today. We've
- 7:50covered massive models and the incredible utility
- 7:52of microtools. But how do we actually feel about
- 7:55integrating this technology? Well, Pew Research
- 7:57just dropped its massive 2026 data report. They
- 8:01surveyed over 5000 U .S. adults. And the findings
- 8:05highlight a very stark and deeply troubling contrast.
- 8:09Yeah. Adoption rates are climbing rapidly everywhere.
- 8:12Roughly half of all U .S. adults are now using
- 8:14AI chat bots. And a solid 25 percent log in every
- 8:18single day. But public pessimism is heavily outpacing
- 8:22the optimism right now. Nearly 40 percent expect
- 8:25AI to make. our society fundamentally worse.
- 8:27Exactly. They believe it will degrade society
- 8:29over the next two decades. Only 16 % actually
- 8:32believe it will make things better. We're rapidly
- 8:34adopting a technology we fundamentally fear right
- 8:36now. And there's a fascinating demographic quirk
- 8:39hidden inside this survey data. We really need
- 8:42to point out the massive under 30 divide here.
- 8:44Oh, right. Younger users lean on AI the absolute
- 8:48hardest. But they're also the absolute biggest
- 8:50skeptics in the entire survey. Just 14 % of them
- 8:53see a positive long -term payoff. It feels so
- 8:56counterintuitive at first glance. They grew up
- 8:58entirely immersed in modern digital technology.
- 9:01They see the potential for severe societal disruption
- 9:05very clearly. Let's look at the actual market
- 9:07share reality check here. ChatGPT is still the
- 9:11undisputed heavyweight at 44 % usage. Yeah, and
- 9:14Google's Gemini currently holds about 24 % of
- 9:17the market. But Anthropic's Claude sits way down
- 9:20at just 6%. Despite being constantly discussed
- 9:23in tech circles, Claude barely registers. It
- 9:26absolutely pales in comparison to the massive
- 9:29footprint of chat GPT. Why do we keep using these
- 9:32tools daily if we genuinely believe they'll eventually
- 9:35ruin society? The immediate benefit to your personal
- 9:38productivity is highly visible right now. That
- 9:41tangible benefit completely overrides abstract
- 9:43fears about long -term decay. Because personal
- 9:46immediate convenience almost always overrides
- 9:49long -term societal anxiety. Exactly. And if
- 9:52we look at the core narrative thread of this
- 9:53deep dive, we're trapped in a massive three -way
- 9:56tug of war. Tech labs are moving at breakneck
- 9:58speed to simulate reality perfectly. Right. And
- 10:02the government is frantically scrambling to lock
- 10:04down control. They literally want the deed to
- 10:07the technological foundation. And the broader
- 10:09public is caught squarely in the middle. We're
- 10:12relying heavily on these tools to survive the
- 10:15modern workday. But we fundamentally distrust
- 10:18where this technological revolution is actually
- 10:21heading. I want to leave you with a final provocative
- 10:24thought today. Something for you to deeply mull
- 10:27over on your own time. Let's hear it. We discussed
- 10:30the government securing equity in these massive
- 10:32AI labs. Imagine they eventually distribute those
- 10:35financial dividends directly to regular citizens.
- 10:38Like a universal basic income. Exactly. If that
- 10:42happens, will that pay out finally by the public's
- 10:44trust? Or will it just deepen our skepticism
- 10:46of the machines running our lives? Thank you
- 10:48for taking this deep dive with us. Otero Music.