"The social system of private property and limited government is the only system that tends to debarbarize all those who have the innate capacity to acquire personal culture." ~Ludwig von Mises
There are only two ways to organize society: coercively or voluntarily. In order to sell coercion, its promoters must slander voluntarism as exploitative and evil, in the process justifying their theft and thuggery as necessary and righteous.
~ Kevin Duffy, X post, January 28, 2026
I foresee disappointment. We talk to dozens of companies monthly and large consulting firms focused on the AI rollout. There is almost no evidence of profit-margin improvement. People use AI to summarize email and do ChatGPT-type searches. Ninety percent of ChatGPT use is outside the U.S.
An important MIT study came out over the summer that found that 95% of AI projects found no real use. In our view, Microsoft Copilot has been a total failure… GPU pricing is in free fall. Based on our research, Nvidia’s Blackwell chips can be rented for $4.00 an hour. Blackwell chips are already being rented at a discount versus just a few months ago. If demand was so strong, why have Blackwell rentals been collapsing?
There are 200 neocloud companies today. Depreciation rates are out of line with reality. H200 chips, for example, were selling at 50%-plus off within nine months of release. A few large enterprises mentioned to us that real economic life when used for training is barely 18 to 24 months.
Most of the growth at the cloud hyperscalers is coming from start-ups, and who is funding these start-ups? More than half of the $150 billion that start-ups invested in the cloud was funded by the large tech companies. This is a giant, circular trade. I agree with Henry [Ellenbogen] that a lot of smaller companies will see productivity gains from AI, but margins are falling at the cloud providers. There is an incentive for semiconductor companies and other tech companies to talk up the benefits of AI, but there is little evidence of profit-margin improvement almost anywhere because of AI, except at companies selling shovels to the gold diggers.
~ Rajiv Jain, "The Roundtable: Part 1," Barron's, January 10, 2026
You can’t understate what is going to happen to productivity in this country. Based on where AI is now, the marginal cost of intelligence in white-collar work is going to zero over the next several years. I’ll give you some examples. MercadoLibre is doing seven times the number of transactions it did several years ago, even as the number of customer service specialists has fallen to 7,000 from 10,000. Rocket Mortgage has said that it can serve 50% more clients per loan officer. In the physical world, trucking companies are using AI to leverage better data systems...
Knowledge workers are using AI – in our case to do investment research – or look for novel drug targets, or improve productivity in coding. Enterprises will continue to pay a premium for better intelligence powered by frontier models, which continue to advance at a rapid rate.
Those who say capex is going to implode may be looking at consumer-oriented AI technology, which has become commoditized. They aren’t taking into account advances in intelligence that will drive advanced applications and scientific breakthroughs. Over the next 10 years, we may have 50 years of advances in medical science. Companies that climb the productivity curve first will be able to cut costs, reinvest in their customers, drive revenue growth, and reinvest capital to create enduring competitive moats.
Companies don’t all climb the curve at the same pace. I foresee an unequal playing field. The Mag Seven will do well, and a lot of smaller companies born in this era will climb the curve and unlock productivity gains.
~ Henry Ellenbogen, "The Roundtable: Part 1," Barron's, January 10, 2026
Take a step back, what's the biggest price anomaly?... I think a massive price anomaly is the share price of Nvidia and the share price of Facebook [Meta Platforms] and the share price of all these guys that are today incinerating cash on the hope and promise that this cash is going to lead to outsize returns, but it won’t. It won’t because, unlike the smartphone revolution, unlike previous booms, this time around China is in the AI room. And what I’ve learned out of 30 years of being in China is that when China enters the room, profits walk out. And if you’re competing with China, you’ve taken a wrong turn in your life.
~ Louis-Vincent Gave, "China Just Ended 100 Years of American Dominance," RiskReversal Podcast, 42:45 mark, December 15, 2026
I think that [AI] in the U.S. is a huge misallocation of resources. In the U.S. the business model essentially is, “Oh, we’re going to get 500 million customers that pay us $20 each a month.”… I’m very skeptical that anybody will do it when China is giving it for free… In China, the development of AI has a lot more to do with, “How do we use this for our factories… essentially getting rid of industrial workers?”… If you think there are three goals: One is to make factories more efficient. I can see how that’s going to happen. In fact… it’s already happening. If the goal is, “We’re going to get rid of low-end service workers – the paralegals, banking analysts, etc.,” that might very well happen. I think it is starting to happen. You see it in the rise of youth unemployment here, etc., but that also comes with big social consequences. And if the end goal is AGI [artificial general intelligence], I’m actually super dubious on that… [T]he whole AGI thing is being sold to us as investors: “Do this, give us billions and we’re going to deliver AGI on the other side. I think it’s a stock promote,… literally the Vancouver guy trying to sell you a gold mine standing next to a hole.
~ Louis-Vincent Gave, "China Just Ended 100 Years of American Dominance," RiskReversal Podcast, 45.35 mark, December 15, 2026
Here's a novel idea: stop making enemies. Stop expanding NATO on Russia's doorstep. Stop vilifying China for being economically successful. Stop bullying everyone with tariffs. Stop economic sanctions and embargoes on countries you don't like, e.g. Iran, Cuba, Venezuela. Stop aiding and abetting the genocide in Gaza.
Why not try this approach? I guarantee, the world would be a more prosperous, peaceful place. The big loser in such a world, of course, would be the military industrial complex, whose budget would be easily cut to under $500B (which is precisely why they are fighting so desperately to prevent such a cataclysm).
No, instead we're going to double down on our current path. These people are now pining for a $1.5T annual military budget!! Meanwhile, total debt clocks in at $39T and counting, our growing list of enemies is selling US treasuries, and the dollar is imploding vs. gold (having lost 75% of its value since Trump first stepped into office in January, 2017).
As WWI critic Randolph Bourne famously said, "War is the health of the state." But ultimately, the war machine brings about a destruction of the currency and bankruptcy of the state. We are now accelerating towards that day.
~ Kevin Duffy, Twitter/X post, January 21, 2026
Oracle is a leveraged, one-way bet on OpenAI, a high-cost, money-losing producer of a lagging artificial-intelligence model. If [Sam] Altman’s company is unable to recapture the top billing in AI or raise the billions in cash it needs this year to keep the lights on, Oracle will be stuck with empty data centers and $248 billion in lease obligations.
It might just prove the ultimate bagholder in the coming AI bust.
~ Evan Lorenz, "Not so oracular," Grant's Interest Rate Observer, January 14, 2026
