Here were a few of our favorites from Stratechery Plus this week.
- Making Sense of DeepSeek. DeepSeek’s remarkable progress in AI has been a big story in tech since the holidays, but it wasn’t until this week that the mainstream—and Wall Street—began to reckon with the implications of a Chinese firm releasing open source AI models that rival those of industry leaders in the U.S. Beginning over the weekend and continuing into a wild Monday in the markets, DeepSeek’s R1 model became one of the biggest stories in the world, prompting all kinds of inquiries from friends of mine who work outside of tech and know nothing about AI. To those people, Ben’s DeepSeek FAQ was a perfect primer on all the relevant breakthroughs DeepSeek achieved, and what that might mean for U.S. tech companies and American foreign policy. Get caught up if you’re unfamiliar with any of those vectors, or, if you understand the issues yourself but are still fielding confused text messages from family members, save yourself some time and send them Monday’s article. — Andrew Sharp
- Why DeepSeek is Not a Sputnik Moment. Several called DeepSeek’s models a “Sputnik moment” in the ongoing competition between the U.S. and China, and there are indeed echoes of that cultural moment, and all of this is probably a healthy corrective to U.S. arrogance and complacency in tech. There’s also, however, an obvious distinction from 1957. DeepSeek’s work is mostly open sourced; in other words, all of this is like Sputnik if the Russians had made it to space and then explained in great detail how they got there. Ben’s Daily Update on Tuesday explored this angle in the context of infrastructure optimization, Google’s S-1 20 years ago, and an awakening for US tech companies moving forward. It was a great, clear-eyed addition to the conversation this week. — AS
- More DeepSeek Answers, and Big Geopolitical Questions. After several days parsing DeepSeek reactions, Ben and I had a great time reconvening on Sharp Tech to catch our breath and answer a whole bunch of emails from listeners. Here, Ben provided answers on the bullish possibilities for American AI firms, the wait for new AI products, frustrations with OpenAI, and the future of compute demand. At the end, we hit on a few unresolved questions around export controls and the inconvenient realities underlying debates about the future of U.S. foreign policy generally. This episode was unlocked and free to everyone, so if you’re not already subscribed, check it out: — AS
Stratechery Articles and Updates
- DeepSeek FAQ — DeepSeek has completely upended people’s expectations for AI and competition with China. What is it, and why does it matter?
- The OpenAI Critique, Comparative Advantage and Infrastructure, Aggregation Theory and Cheap AI— Further grappling with DeepSeek implications, including the lack of focus in U.S. tech companies, the problem with comparative advantage, and the possibility of zero marginal cost AI.
- Asianometry on Japan’s Chemical Dominance, Stratechery Updates, and Chinese New Year Schedule Changes — Chinese New Year wishes, and a scheduling adjustment.
- Microsoft Earnings, Meta Earnings — Microsoft and Meta both emphasized their optionality as models become a commodity.
Dithering with Ben Thompson and Daring Fireball’s John Gruber
Asianometry with Jon Yu
Sharp China with Andrew Sharp and Sinocism’s Bill Bishop
Greatest of All Talk with Andrew Sharp and WaPo’s Ben Golliver
- The Celtics Get Rocked in Los Angeles, Good Players Who Make Bad Decisions, All-Star Starters and Assorted Nonsense
- Choose Your California Adventure, Pick Protections and Related Questions, Kyle Kuzma and Educating the Youth
Sharp Tech with Andrew Sharp and Ben Thompson
- 72 Hours of DeepSeek Hysteria, What DeepSeek Means for Big Tech, Lessons on the Efficacy of Chip Controls
- The End of OpenAI and Microsoft, Risks and Rationale of the Stargate Project, DeepSeek-R1 and Bitter Lessons for the Future
The latest Sharp Tech video is about AI’s impact on software jobs.