My Two Cents on the AI Debate || Sept 17, 2026

You are brainwashed about Bernie Sanders and Obama and plus Ataturk. You confuse your feelings with reality sometime around those people. Bernie is capitalist as you are, he is more inclined Scandinavian capitalism (he is not even fully there) and it doesn't make him a freak at all. Obama did the impossible just by elected, and he couldn't move forward for some progressive policies for obvious reasons and Trump was the answer for that anger. If you are looking for freaks it is half of the Trump voters. Stay respectful.

IF YOU'RE CONCERNED ABOUT AI, ROBOTS, AND A POST-LABOR ECONOMY... READ THIS: SCENARIO-1: Automation, AI, and the Transition Everything in this universe is based on cooperation there's nothing that does not cooperate with something else and everything else, includingincluding AI and I Scientia Non Domus Habet, (Knowledge Has No Home) antiguajohn a Post-Labor Society Nearly 70% of GDP in modern economies is driven by consumer spending. This creates a fundamental question for the future: What happens if automation and AI eliminate large numbers of jobs, and consumers no longer have the income needed to participate in the economy? If consumers lose purchasing power, businesses lose customers. If businesses lose customers, companies fail. Eventually, even the wealthy are affected because the entire economic system begins to contract. One possible solution is not to resist automation, but to adapt civilization to it intelligently. Tax the Robots: A Transitional Economic Model Employees are expensive. From a business pective, robots and AI systems make economic sense because they can dramatically reduce costs while increasing efficiency and productivity. However, there is a systemic contradiction: The same automation that increases corporate profits may simultaneously reduce the purchasing power of consumers. A modern economy cannot function if productivity rises while consumer income collapses. One possible solution is a robotic taxation system designed not to punish innovation, but to stabilize society during the transition toward a post-labor economy. If a portion of the enormous productivity gains from automation were redistributed back into society, several things could happen simultaneously: Companies would still benefit from massive efficiency gains. Citizens could maintain purchasing power. The cost of living could be reduced. Consumer Using AI robotics 3D printing and other Technologies a small company could actually produce products competitive with major corporations without having to spend hundreds of millions or billions of dollars on a factory if this was kept within the Dunbar number of 150 including employees families and friends and not have to do extensive advertising not having to have all the executive departments they could actually end up producing products at a lower cost than major corporations cuz that would make manufacturing distributed widely and instead of everything are all the money going to the top it would be distributed amongst thousands or hundreds of thousands of small manufacturers remember there's an old saying that says money is like manure you pile up two months it just stinks to high heaven and it would remain stable. Economies could avoid large-scale instability caused by technological unemployment. In the long term, both companies and workers benefit. This is not anti-technology. It is pro-stability, pro-human, and ultimately pro-civilization. Humans Did Not Evolve for Industrial Labor Systems Another issue rarely discussed is whether modern labor systems are psychologically compatible with human beings. Humans did not evolve for endless industrial schedules, long commutes, or extreme work cultures such as the Chinese “996” system: 9 AM to 9 PM, 6 days a week. Such systems create chronic stress, anxiety, burnout, and long-term damage to mental and physical health. Anthropological studies of hunter-gatherer societies — social groups much closer to the conditions under which humans evolved — suggest a very different pattern of life. Within communities roughly aligned with the Dunbar number of approximately 150 individuals, hunter-gatherers often spent the equivalent of about 40 hours weekly securing food and necessities. However, this was spread over seven days, averaging closer to 5.7 hours per day. More importantly, these activities were frequently social, cooperative, and meaningful. Hunting, fishing, gathering, storytelling, and tool-making were integrated into daily life rather than separated into something psychologically experienced as “work.” The modern division between “living” and “working” may itself be historically unnatural. Automation Could Improve Human Life Automation and AI also have the potential to dramatically improve quality of life beyond economics alone. For example, autonomous vehicles could: reduce traffic deaths, lower insurance and healthcare costs, reduce commuting stress, save enormous amounts of time, and improve overall public safety. Combined with robotics and AI-assisted productivity, humanity may eventually move toward a civilization where survival no longer depends upon exhausting labor simply to meet basic needs. The real challenge is not whether this transformation is coming. The real challenge is whether humanity managems the transition wisely — or attempts to force industrial-age economic systems to operate in a world where intelligent machines can increasingly replace human labor. There are many additional implications and considerations, but it is important to address one issue at a time to avoid losing sight of the central problem. Remember, Everything in this universe is based on cooperation there's nothing that does not cooperate with something else and everything else, including AI and I Scientia Non Domus Habet (Knowledge Has No Home) antiguajohn

This guy sounds so convincing but when you scratch the surface there is so much hyperbole and exaggeration in the content. Zero nuance, and quite a lot of his facts wrong…

How about Marc Benioff’s comment on his AI Leadership panel regarding AI regulation that “China has it right. Those companies behave very responsibly”. So is he willing to sell Salesforce to the government to be run as efficiently as the VA or Social Security?

You said AI will probably slow down for 4 to 6 years. Is it possible development and manufacturing continues, but only for the rich - who use it to control the population?

The AI evolution conversation on public safety has a lot of echos with the nuclear evolution conversation. After many people died, there were safety measures put in place that lasted until that generation died out.

I forget how the chip supply chains are supposed to break because of de globalization? Also is the tech community hyping up China, I was off the belief that China will not see the end of the decade.. Interesting plays happening everywhere

Thanks Peter. That was indeed ....uh2centrrr ... 😐 ... Catchuh later ... 😺

Just fyi, CPU/GPU chips don't "burn themselves out". They are traditionally replaced by something better (faster and/or less power hungry per calculation) but years later the older chips still run just as well as they ran when new.

A couple of points/questions here....1) $20 to you Peter if you can show me a picture of Bernie Sanders holdings hands with Steve Bannon. Note, it can't be an AI simulation. 2) Is Peter Zeihan real or an AI simulation created by Peter Thiel to take over the world? 3) Am I real or an AI simulation...created by Peter Thiel to take over the world?

I agree with Peter, everyone has a point. Plus, that discussion about the electronic womb back in the day, that explains a lot, I’m sure! However, I’m interested in haring Peter’s take on Canada joining EU countries in an alliance that does not include the US. As someone who firmly believes the US is currently being led by a group of bozo’s, this issue is different. If this is all real, we can’t really allow an advisory on our northern border. Canada must know that. Other perspectives?

Natasha Kaneva just published a note arguing that, for the first time since the Iran conflict began, JPM no longer has a clear baseline for how the conflict ends, as several of the economic “red lines” previously seen as constraints have already been crossed (including $100 oil, a 5-handle on the 10-year Treasury yield, and record-high fuel prices). Kaneva notes that the market has absorbed the shock very differently from JPM’s original expectation: global crude and product inventories have drawn by only ~555 million barrels, roughly one-third of the ~1.6 billion barrels initially anticipated, while demand has done far more of the adjustment, running ~4.4 mbd below last year’s level since March. With additional inventory capacity still available—particularly in China, Europe, Japan and South Korea—JPM sees room to cushion a prolonged disruption before reaching its ~7.65 billion barrel stress threshold. If Middle East flows remain around current levels, however, Kaneva estimates Brent could average around $87/bbl in 4Q26 and reach around $86/bbl in December, versus JPM’s current forecasts of $80/bbl and $78/bbl, respectively

Don’t be mean to Bernie.

There are two competing factions divided by the personal property boundary. One are the effective altruism people and the other is what I think is best described as the trump account strategy. Both camps recognize that hyperdeflationary debt destruction threatens civilization at the social architectural level, so deficit spending will persist but this also causes what is described as the K economy and left to its math becomes the L economy. All welfare gets spent into the financial architecture and flows up to those holding the assets. So if Trump can scale up billionaire gift redistribution on the Michael Dell model a viable redistribution of wealth to the youngest generation can occur and personal property is preserved. The effective altruists are the Dario people.

Anyone remember CRISPR technology and gain of function research ? The same debate on risk lead to a temporary moratorium on this science under the Obama administration. Big review occurred, safety procedures modified by the CDC, WHO, and international experts……. Thumbs up given to proceed………. 18 months later , the first case of Covid 19 ……. And YES, it came from a Wuhan lab biosecurity leak……..

There is no AI, you morons. It is not AI, we are calling the thing something it is not. This is all dumb. We’re just wasting resources.

All those things. But also, while I don't think they're going for regulatory capture I think it's simpler than that-- they've reached a plateau in the magic and they don't know how to commercialise it beyond chump change. They're stalling. Three major competitors do not all agree to chill at the same time if any of them think they have an edge. But also everything you said is true as well.

Always love your take, but hard no on "wait until the 2040's and we'll see what happens". There will be a TON of turmoil between now and then with jobs, China competition (haven't we learned our lesson - "they'll never catch up" statements?!), and bad actors with the tech that already exists.

Sorry Peter you're COMPLETELY wrong about the software side for China, they're kicking our ass...I have a model on my workstation right NOW running that is free open source at 125b parameters (Qwen 3.8 Flash Next) that OUTSCORES Anthropic's Opus 4.6 model on max effort (2.4T parameter model!!!) across the board, and their team stated it's just an early development PREVIEW of their new engine and it's architectural breakthroughs. For reference, Opus 4.6 was released February of this year as state of the art. Isn't it funny how every person that has left their respective companies leave to run a non-profit (funded or sponsored by those companies), and never disclose what their shareholdings are, considering most employees were paid in stock options? Geee, it's almost as if there's a collective incentive for everyone to scare legislators into passing regulations for review committees run by those same people to decide who can be allowed to compete at all. 🤔 Also, openai and anthropic are desperate to become nationalized so they are too big to fail and aren't on the hook to pay their debiters, because while previously compute and training opex was matching revenue linearly, hardware and the infrastructure costs skyrocketing is now outpacing revenue by rapidly increasing margins. First quarter 2027, have your dry powder ready is all I have to say.

Having started my career in molecular biology in the 1980’s, I remember the sometimes apoplectic concerns about the accidental release of genetically modified bacteria into the environment. Not to diminish the concerns about possible COVID issues, it turned out to be a way-over-the-top reaction. So, yeah, we have been here before.

The Chinese are NOT 36 months behind. Take a look at their some of their robotic applications and tell me we’re still ahead. Obviously that differs from raw compute but still they are near-comparable in this realm.

The most useful thing you said there is that we're in uncharted territory. The problem I foresee is that by the time that analysts, like yourself, and policy makers recognize there is a problem, it will already be too late to do anything meaningful about it.

(RJ) The anti-hyperscale datacenter thing is not monolithic... so far I have seen groups who hate AI in general, those who want to preserve farmland (and are equally anti-solar farms), those who distrust ANYTHING corporate, a bunch with real noise, power and water shortage concerns (at various levels of documentation) and a large, angry group tired of tax-benefit giveaways to richkids and property developers. I haven't seen much by way of complaint against brownfield sites like old power stations or stuff way out in the boonies, or actually in gasfields. (Could not figure out how to reliably invest with the research and moved on.) INAL, but the antitrust fears are most likely real. I seem to recall the financial services industry (back when I used to work for a living) rearing up on its hindmost legs and refusing to do government-requested "joint work on blocking AML and anti-terrorist funding" unless they were granted specific immunity against "actions that might lessen competition." USA PATRIOT Act 314b, IIRC, but I can't find the slide.

Except that the Russians and Ukraine have already crossed the threshold in warfare with drones with autonomous AI selecting and hitting targets including individual soldiers.

That leaves my idea of abandoned data centers similar to the mass graveyards of EVs, ruined. 😢

At the very top - money talks. That’s it. Security concerns - software buggy or can’t deliver up to trillion dollar evaluation.

Bernie Sanders a freak? Lost a bit of my respect there Pete.

I suppose it is nice to have the conversation, now. However, I am not confident that will help, in the 2040’s much, other than to give a bunch of “Legislators Emeritus” an “I told you so” moment. The Cincinnati, Ohio, City Council reacted with self righteous rage, when the festival seating plan used for a 1979 “The Who” concert resulted in a stampede at the doors to the venue, causing eleven deaths. Then the newspapers pointed out the topic had already been debated in council, in ‘72, about a “Yes” concert, (injuries), in ‘76, about an Elton John concert, in ‘77, referencing a Led Zepplin concert (injuries, both in Cincinatti, and Tampa, Fl). All forgotten, all un-acted upon in less than a decade. Until 11 people died. I am hopeful, but not optimistic, about our chances today’s discussions about AI, will somehow ameliorate the potential for an AIpocalypse. Just sayin’…

Technology itself is not dangerous - what it does is potentiate the well-known tragedy of the commons. AI runs on data that was generated by the public so where is public equity? Maybe a good way to start is to rethink AI as a part of critical infrastructure

There is only a consensus among these companies because they want regulatory capture of the market. Kind of like what European dinosaur firms did hand in hand with the EU commissions, they just want to make sure there is no new competition coming their way by innovative startups. With AI regulation, you get the European recipe, so not much startups at all, available funding going all to these “concerned” large companies instead of the new players, chance at a disruptive innovation is limited. That’s why their economy couldn’t renew itself and did not participate in the digital technological revolution. These AI companies fear the scenario, of a new startup coming up with a similarly or more capable models but much cheaper to train and to run, maybe on other hardware like FPGAs or something, as this industry is new, breakthroughs definitely can happens. So they try to copy the recipe from regulation lover innovation hater continent to kill anything new, aka the future. They will also suggest soon to fund these commissions and regulatory bodies, all so conveniently, as if their lobbyists weren’t enough, trying to block the path, they themselves have followed. I hope they fail.

I don’t remember any movies about time traveling looms trying to kill everybody

bu far and away his worst take

Since you're not worried, do you think the AI Doomerism is more about regulatory capture, keeping Chinese models out, or pumping their upcoming IPOs?

What about the dangers of AI Bears? Or Bears using AI?