Join institutional investors, analysts and strategists from the world's largest banks: Subscribe today
Already have an account? log in
The Numbers Behind America’s AI Dominance
The "US exceptionalism" theme's ubiquitous to the point of being a cliché by now.
Universal buy-in (and you can take "buy-in" quite literally given the extent to which US-focused funds dominated equity inflows late last year) almost makes the America "TINA" narrative feel like a contrarian fade. "Almost" because... well, because have you seen the ostensible alternatives?
The choice between investing in US stocks and plowing money into RoW equities is about like the choice between living in Am
AI is spectacular. It is the future and its a mirage and a back-buster. I have lived it since I took a summer course in 1963 to learn PL/1 (yes that is programming language one). I did my grad work doing empirical politics building models of politics. Naive, yes, wildly successful too. I worked at 3 banks made them a ton of money and finally started an FX / FI / macro / political hedge fund. The logic is all Blockchain, AI, and crypto oriented but this is a totally probabilistic trend-following game, often at its stupidest basics. If it went up yesterday, it will go up tomorrow. This whole business will prove to be very volatile and crash oriented. Blockchain is physically based and is best for that reason. Crypto is not anything at all, even if Blockchain tells you that the prices are not lies. Crypto is the old commodity joke: throw shit on the wall. If it sticks we trade it. With shit your delivery is nasty stuff, but with crypto there is no delivery just another fool to buy it. It is the ultimate of the greater fool theory. It takes more brains to make money here than buying a leveraged piece of nothing.
It might be the slew of doomsday AI articles I read this weekend alongside Zuck’s comments about replacing mid-level developers this year, but I expect some AI driven chaos this year. AI might not be ready for primetime yet in areas that require high levels of accuracy, but there are plenty of avenues for AI to be used to sow confusion and create flash points that could explode into real world catastrophes.
We already know our president and his administration are susceptible to misinformation and disinformation. Who’s going to apply the brakes when Trump sees provocative AI-generated content on Twitter and starts launching missiles, both literal and figurative?
Also, clearly this take isn’t novel, but the acceleration of inequality and wealth concentration both in the market and at an individual level will continue unabated. Zuck expects one of his biggest cost inputs to go down (gotta eliminate as many high paying jobs as possible so capital can continue to maximize profit) so I would not be surprised to see the Mag 7 continue to rally and leave the 493 behind. Plenty of catalysts for recession in the real economy, so what happens if the Fed does have to start cutting rates and the long duration aspect of the Mag 7 comes back into play?
The other aspect of AI that I think is still underappreciated is the way AI will create even more tech addiction. If TikTok is bad now, AI will just add gas to the fire when it comes to creating addicting content. I’m sure there are plenty of untapped sources of rage or new memes that AI can unlock.
We are headed to the dystopian future of Wall-E. The only thing Wall-E got wrong was that everyone would be fat. We need the equivalent of Ozempic for tech addiction, but it would likely be banned under the auspice of “free speech.”
Actually, Chinese companies have already achieved a lot, including LLM and inference models that require far less computer power than our hoggish models. Functional AI which doesn’t require nuckear reactors to power. It’s yet another unintended consequence of our tech export embargoes.
Derek – The Chinese and everyone out of the 7 and either Wall Street and Silicon Valley know that they are way over deveoping and making things more expensive. Typically overdone.
Well, according to yjis newletter from Malwarebytes AI is aleady proving helpful to one industry:
https://www.malwarebytes.com/blog/news/2025/01/ai-supported-spear-phishing-fools-more-than-50-of-targets?utm_source=iterable&utm_medium=email&utm_campaign=b2c_pro_oth_20250113_januaryweeklynewsletter_v2_173644541424&utm_content=AI-supported
H-Man, in the AI race, the topic of ethical applications appears to have been banned to the back of the bus.