In partnership with
THE OT ALGORITHM
Gif by josephinamusic on Giphy
Hi friends,
Happy last week of October! How did that happen?
This week's headline is about who gets to build the AI we use, and who gets to decide which AI is safe enough to use at all. The US government just ordered a popular Chinese AI model out of its civilian agencies, along with anything built on top of it.
It's been a busy week in AI news. Anthropic also made its IPO filing public, and I'll walk through what it reveals about who's paying whom. Then we'll work through a clinical scenario about a cheap documentation tool and the question most of us forget to ask: what model is actually running underneath it?
Let's get into it.
Note to readers: The ads and some of the links in these weekly newsletters contain affiliate links to help support my work. If you’d like to show support, all you have to do is click on them and look around their site a little bit. 🙂 Disclaimer: Placement of the ads included in these weekly issues is not an endorsement. Please do your own research before using any AI tools.
AI research, explained in one morning email
Most of what shapes AI next year is sitting in a research paper today. Keeping up with them is a full-time job, and the abstracts don't help.
TLDR AI is the newsletter that does the translating. Each morning you get the handful of papers worth knowing about, curated by Anthropic and ex-Google engineers, with a plain-English summary of what they found and why it matters.
You come away smarter about AI research in the time it takes to finish your coffee.
Free, and delivered to 1.1M+ inboxes every morning.
HOT TAKE THIS WEEK
Two Things Can Be True About the DeepSeek Ban
Gif by HannahWitton on Giphy
Back in issue #36, I wrote: "I still don't necessarily understand this need to constantly compete with China in being 'first' or 'the best.'" This week, the race framing showed up in the headline itself: the BBC ran this story as the US banning DeepSeek "in the race for AI dominance."
So let me try to hold two things at once.
The security concerns are real. DeepSeek's own privacy policy says it collects, processes, and stores personal data in the People's Republic of China. In early 2025, the mobile security firm NowSecure reported that DeepSeek's iOS app sent some data without encryption and used hardcoded encryption keys. If you work at the State Department or Treasury, you shouldn't be typing work into that app. I'd say the same to a clinician typing patient details into any tool whose data lands somewhere they can't see or control.
The race rhetoric also serves the companies already in the lead. In issue #36, I pointed out that Dario Amodei's "pace the frontier" essay argued for keeping cheaper competitors, especially open-source Chinese models, from catching up. I called that slowdown pact a moat. Then in issue #39, we watched the message flip from the industry saying "slow down" to the government saying "speed up, so we beat China." This week's ban fits that story too. DeepSeek is cheap and openly available, and it competes directly with the American labs that, per the BBC, are both heading toward public listings and accusing Chinese firms of copying their models.
Both can be true. A tool can carry genuine data risks, and the people warning you about it can also benefit when you stop using it. (I'm an OT reading the AI and policy side here, not a national-security analyst, so I'm sticking to what's on the record.)
From a client's side, the question doesn't depend on which flag a company flies. What matters is where their information goes, who can access it, and whether anyone told them. That's the question I'd want us asking about every AI tool, American ones included.
