Director, Global Macro, Eurasia Group. Former RAND China Research Center, Bloomberg Economics, CSIS, DNIO for Economic Issues, and CIA. Views my own.
Washington, DC
Joined March 2011
- Tweets11K
- Following1.4K
- Followers14.7K
- Likes30.2K
Pinned Tweet
🚨🇨🇳🏭📈🚨
New @RANDCorporation report on China's techno-industrial policies under Xi! With @JonathonPSine and Benjamin Lenain. We detail the evolution, goals, and instruments. Lots of charts and summary tables! Please enjoy.
rand.org/pubs/research_repor…
China releases Q3 GDP tonight. In nominal terms, AI is a big part of China’s export boom. Chips and computers account for more than 40% of the recent export increase. But on a net basis, AI trade is weighing on China’s trade surplus. The surplus growth comes from vehicles and other machinery.
The chip export boom is also mostly about prices. In September, China’s chip exports rose 136% in value from a year earlier, but the number of chips rose only 1%. Most of that is memory chips, where the global shortage has driven up prices. So AI trade isn’t adding much to real export growth. If net exports boosted real GDP growth in Q3, it wasn’t because of AI.
Those advocating for Chinese FDI into the US on the grounds that it will allow more tech transfer into the US should pay closer attention to China's export controls. Beijing won't let that happen for important technologies, even if Washington were interested.
Gerard DiPippo retweeted
There are about 200 Uber/Lyft drivers who live in DC. How many DC residents should die to keep Waymos out?
Gerard DiPippo retweeted
My strongest “vibes not evidence” belief is that remote work is a cancer that’s making (most) people miserable even though it seems better in the short term.
Just finished this. Recommend. It's a clearly-written, readable overview of how we got here and the key personalities. I nervously await the sequel...
Gerard DiPippo retweeted
We are going to need forward-looking policies that address the real economic disruption that AI will cause, but this is the wrong way to do it.
Study after study has shown that Waymo’s crash far less often than human drivers, and are associated with far fewer injuries involving humans, pedestrians, and cyclists. Waymo claims that per mile driven compared to human drivers, Waymo’s have 95% fewer serious crashes, 93% fewer pedestrian accidents, and 86% fewer cyclist accidents.
And AI driving is only getting better. I can personally attest to this; I only learned to drive a few years ago (the New Yorker in me held out for as long as possible), and Tesla’s Supervised Full Self-Driving, while imperfect, has improved dramatically in recent years and is a better driver than me 99% of the time.
This year, 29 people have died in car accidents in DC. 25 died last year, and 52 in 2024. Replacing human drivers with Waymo’s will save lives. We will need comprehensive and disruptive policy responses to the economic challenges that AI poses, but reactive ones like this don’t address the underlying issues but will come at a real human cost.
Gerard DiPippo retweeted
Why is our next mayor putting Maryland and Virginia residents ahead of the safety of Washingtonians?
200 District residents are full-time ride share drivers, not “countless.”
More in Capital Commonsense. capitalcommonsense.com/p/our…
Gerard DiPippo retweeted
I do not think that we should accept more crashes, pedestrian injuries, and passenger assaults in order to protect largely part-time gig work, while also sending a signal that DC isn't friendly to new tech.
Gerard DiPippo retweeted
Chris Wood cites token prices for Chinese models and OpenRouter usage data to predict "massive capital destruction". Big claim based on some weak, narrative-driven analysis
- there are many US models on the pareto frontier, with large gross margins
- OpenRouter data captures a truly tiny slice of proprietary model usage
- there is a tonne of uncertainty about how AI develops: how long will Chinese models remain open for? RSI, if it ever arrives, could mean anything less than the frontier is not worth having
The most likely long-term outcome of the AI boom will be massive capital destruction in the US, with market share going to cheaper open-source Chinese models, according to Jefferies. bloomberg.com/news/articles/…
Gerard DiPippo retweeted
Anthropic's leaders need to stop commenting on the emergence of RSI like a local news anchor reporting on traffic on the 101, and acknowledge that they are the traffic, and they're not going to build something that they don't want to exist.
They should move beyond "pacing" language, commit to specific policies and red lines, explain why those lines are red, and then use that costly signal to build an international consensus around norms.
It's not that I trust the CCP, or that I think global lab monitoring and verification is imminent. But the status quo is absurd.
You've got US lab leaders writing (or favorably quote-tweeting) essays about slowing down while simultaneously saying "things are moving faster and faster" and "btw, our internal models finished mathematics on Tuesday, next up: physics."
The effect is the equivalent of Usain Bolt winning race after race while using his press conferences to bemoan the dangerous speed of modern sprinting on one's legs so would everybody please try a little less hard next race, and then ... breaking his 100m record again the following month. Honestly ask yourself: If you were China, why would you take this pausing/pacing stuff seriously?
Gerard DiPippo retweeted
For better or worse the people who’ve been prescient about AI are the “doomers” who founded OpenAI and Anthropic.
The accelerationists are the people who thought NFTs and/or the Metaverse were going to be the next big thing.
This is a funny tweet. Because do people actually associate @pmarca and Thiel with having been prescient about AI? The former I associate with b2b SaaS. The latter with kind of Silicon Valley woo ideas. With Elon it’s more complex, but Brian has a substantive point to make.
Gerard DiPippo retweeted
The King of Prussia Mall is a true land of abundance, just gotta pop some apartment towers on top of these parking structures.
Gerard DiPippo retweeted
Just wild divergence among friends/acquaintances right now in AI use, maybe 25% are like "this is making me massively more productive" and 75% are only using it at the margins. Very little middle ground. (I'm sure the 25% is very high relative to the general population.)
Gerard DiPippo retweeted
Starting to think that building something smarter than humans has implications beyond B2B SaaS
Gerard DiPippo retweeted
Two charts showing extremely important, widely held AI assumptions are outdated:
Assumption 1: Chinese models are cheaper & more efficient. In fact, at almost every level of intelligence @ArtificialAnlys finds US models cheaper per task. Chinese models might be cheaper per token, but US models tend to be MORE efficient with tokens.
Everyone seizes on papers from Chinese labs with architectural improvements for efficiency and assumes US labs are undisciplined. But ask yourself what OpenAI/Anthropic/Google have behind the scenes that would cut down on their #1 cost. And then add that they have more efficient chips. Should not be surprising that US labs can undercut!
You can't undercut free, though, as anyone can download CN open weight models and use w/o paying the labs anything. BUT unless you have your own compute (few firms do), the days of firms running Chinese models for the cost of compute are numbered or over because Chinese AI labs are reportedly asking for 30% cut from cloud providers serving their models. This is likely main way these models are accessed, and the costs are going to be passed on.
Assumption 2) US company spending on AI is spiraling out of control.
@arakharazian finds that companies are getting better at picking the right cost/capabilities tradeoffs and getting better deals from competition between Anthropic and OpenAI. Use continues to rise, but at a cost that is leveling out. Controlling cost will put less pressure on firms to move to Chinese models, esp if they aren't cheaper anymore!
People understandably cite this UNIDO chart, claiming China will account for 45% of the world’s manufacturing value added by 2030, but it’s wrong and shouldn’t be cited.
It comes from a 2024 UNIDO paper, which @JKynge cites in his new @ChathamHouse report on China and European industry. (His report is worth reading. My issue is only with the UNIDO figure!) UNIDO says it took each country’s average manufacturing growth over 2010–19 and projected it forward from 2024. But if you do that with UNIDO’s own data, China reaches 38% in 2030. The published numbers only come out if you use 2000–19 growth instead, which assumes China’s 13% annual growth from the post-WTO boom suddenly resumes in the coming years (it won’t). This appears to be an error in the UNIDO report. The UNIDO chart also says “industrial production,” but its numbers match manufacturing value added.
China’s manufacturing growth has since slowed to around 5%. If we apply each economy’s 2019–25 manufacturing growth rate, China reaches 35% by 2030. In current dollars, China’s share peaked at 30.5% in 2021 and was 28% in 2024 because of deflation and a weaker yuan.
China’s manufacturing sector is the biggest in the world, and it’s still growing. But it won’t reach 45% of global manufacturing in 2030. That would require China’s manufacturing output to grow about 14% a year for five years. Even if the rest of the world’s manufacturing didn’t grow at all, China would need 11% a year. It hasn’t grown that fast in any year since 2011.
h/t @JonathonPSine, who pointed out some of the UNIDO issues last year.
China is going to be nearly half of global industrial production in 2030
The US only 11%
Edifying read from @JKynge
chathamhouse.org/2026/09/chi…
Gerard DiPippo retweeted
I wrote about the systemic failure by the world's most valuable company and powerful government to stop smuggling of AI chips to China
US officials say Nvidia is missing obvious signs. Nvidia says "companies honor US laws as they are, not as America’s critics wish they were"
1/
