/
Navigation
C
Chronicles
Browse all articles
C
E
Explore
Semantic exploration
E
R
Research
Entity momentum
R
N
Nexus
Correlations & relationships
N
~
Story Arc
Topic evolution
S
Drift Map
Semantic trajectory animation
D
P
Posts
Analysis & commentary
P
Browse
@
Entities
Companies, people, products, technologies
Domains
Browse by publication source
Handles
Browse by social media handle
Detection
?
Concept Search
Semantic similarity search
!
High Impact Stories
Top coverage by position
+
Sentiment Analysis
Positive/negative coverage
*
Anomaly Detection
Unusual coverage patterns
Analysis
vs
Rivalry Report
Compare two entities head-to-head
/\
Semantic Pivots
Narrative discontinuities
!!
Crisis Response
Event recovery patterns
Connected
Nav: C E R N
Search: /
Command: ⌘K
Embeddings: large
VOICE ARCHIVE

Kevin A. Bryan

@afinetheorem
27 posts
2026-03-04
One reason to visit China often is that, like SF vs everywhere else, I feel six months behind in my knowledge base always. Eg, this week Alibaba's best AI folks quit, and BYD went from “crush everyone” to worst YoY sales drop I have ever heard of (I have no idea why for either).
2026-03-04 View on X
TechCrunch

Junyang Lin, a tech lead on Alibaba's Qwen team, abruptly steps down, and two other team members leave; one contributor says “I know leaving wasn't your choice”

Alibaba's Qwen AI project has lost one of its most visible technical leaders just a day after the Chinese tech giant unveiled its new Qwen 3.5 open-weight small models.

2026-03-03
Re: DoD-Anthropic craziness & @deanwball's great essay today: let's try a steelman USG defense. You are Canada, or a future D admin in the US. Contract w/ Starlink to handle all govt comms or similar. You worry it is so integrated & important - what if Elon shuts off access? 1/8
2026-03-03 View on X
Reuters

The US Treasury Department, State Department, and Federal Housing Finance Agency stop using Anthropic's AI products; the State Department will switch to OpenAI

The U.S. Treasury Department, State Department and the federal housing agency are terminating all use of Anthropic products …

Re: DoD-Anthropic craziness & @deanwball's great essay today: let's try a steelman USG defense. You are Canada, or a future D admin in the US. Contract w/ Starlink to handle all govt comms or similar. You worry it is so integrated & important - what if Elon shuts off access? 1/8
2026-03-03 View on X
Bloomberg

Sources: amid negotiations with the DOD, Anthropic submitted a bid to compete in a $100M DOD contest to develop voice-controlled, autonomous drone swarming tech

Anthropic PBC was among the artificial intelligence companies that submitted a proposal earlier this year to compete …

Re: DoD-Anthropic craziness & @deanwball's great essay today: let's try a steelman USG defense. You are Canada, or a future D admin in the US. Contract w/ Starlink to handle all govt comms or similar. You worry it is so integrated & important - what if Elon shuts off access? 1/8
2026-03-03 View on X
Hyperdimensional

The Anthropic-DOD skirmish is the first major public debate on control over frontier AI, and institutions behaved erratically, maliciously, and without clarity

On Anthropic and the Department of War  —  I.  —  A little more than a decade ago, I sat with my father and watched him die.

2026-02-28
Dean literally wrote the Trump admin's (very good!) AI policy. He is no left-wing activist. Today's insanity is just like the rest of the lack of stewardship of institutions we keep seeing: Harvard, AIDS treatments, NATO, Canada, courts - it's always full leverage abuse of power.
2026-02-28 View on X
@secwar

Defense Secretary Pete Hegseth directs the DOD to designate Anthropic as a supply chain risk, barring military contractors from doing business with the company

This week, Anthropic delivered a master class in arrogance and betrayal as well as a textbook case of how not to do business with the United States Government or the Pentagon. Our ...

Moral of this story is that no smart company is going to do business with this government. Anthropic built literally the world's best AI and integrated with the military, as a national service. They fulfilled their contract precisely. Result: they are being treated like Huawei.
2026-02-28 View on X
@secwar

Defense Secretary Pete Hegseth directs the DOD to designate Anthropic as a supply chain risk, barring military contractors from doing business with the company

This week, Anthropic delivered a master class in arrogance and betrayal as well as a textbook case of how not to do business with the United States Government or the Pentagon. Our ...

Announcing this with a token “we ask that the same terms go to other unnamed companies” is the opposite of a profile in courage. Look at how many key employees at Anthropic publicly support their CEO on “don't be extralegally extorted” vs. how many will publicly support Sam here.
2026-02-28 View on X
@sama

Sam Altman says OpenAI reached an agreement with the DOD to deploy its models in DOD's classified network and asks DOD to extend those terms to all AI companies

Tonight, we reached an agreement with the Department of War to deploy our models in their classified network. In all of our interactions, the DoW displayed a deep respect for safet...

Moral of this story is that no smart company is going to do business with this government. Anthropic built literally the world's best AI and integrated with the military, as a national service. They fulfilled their contract precisely. Result: they are being treated like Huawei.
2026-02-28 View on X
Anthropic

Anthropic says it'll challenge “any supply chain risk designation in court” and that the designation would only affect contractors' use of Claude on DOD work

Earlier today, Secretary of War Pete Hegseth shared on X that he is directing the Department of War to designate Anthropic a supply chain risk.

Dean literally wrote the Trump admin's (very good!) AI policy. He is no left-wing activist. Today's insanity is just like the rest of the lack of stewardship of institutions we keep seeing: Harvard, AIDS treatments, NATO, Canada, courts - it's always full leverage abuse of power.
2026-02-28 View on X
Anthropic

Anthropic says it'll challenge “any supply chain risk designation in court” and that the designation would only affect contractors' use of Claude on DOD work

Earlier today, Secretary of War Pete Hegseth shared on X that he is directing the Department of War to designate Anthropic a supply chain risk.

2026-02-06
Market signal: many of the best folks at AI labs that have left for their own thing recently are working on AI for science. (Also, “hobby is hiking in wilderness for weeks” is a very positive signal as well, as far as I'm concerned! Good luck!)
2026-02-06 View on X
Financial Times

US tech stocks slide for a third consecutive day amid a sell-off in the software sector; the Nasdaq closed down 1.6%, MSFT 4.95%, QCOM 8.46%, and PLTR 6.83%

Nasdaq drops as recent high-flying companies pull market lower  —  US tech stocks fell on Thursday, as weak jobs data piled further pressure …

2025-11-08
Indeed. World getting closer to accepting most obvious econ implications of AI: things it can do will get very cheap, complements to those will get expensive, society will get richer, things with + income elasticity of demand will get more expensive. Scarce factors get rents. 1/2
2025-11-08 View on X
a16z

Both Baumol effect and Jevons paradox may apply to the AI supercycle, as services with AI productivity gains get cheaper, while those without AI become costlier

and gets pricier as wages rise (Baumol effect). Alex Danco / @alex_danco : The Baumol effect (I still like calling it “cost disease") is topical these days - we wrote about it in t...

2025-09-26
AI economists and AI researchers: this is *excellent*. Details below, but as I feel like I've said in every talk on this topic since my slides said “GPT-2”, 1) AI technical capabilities are better and improving quicker than you think, 2) impact on economy *much* slower. 1/13 [image]
2025-09-26 View on X
The Works in Progress Newsletter

Why AI isn't replacing radiologists: models underperform in hospital settings, AI use faces legal hurdles, and the job is much more than image recognition

For years, radiology has been the go-to example in conversations about AI and professional obsolescence. … Bluesky: Eric Knutson / @alwayscorrect : I think they are using predictiv...

2025-08-02
Very cool result (and easy-to-read!). Vector embeddings are great: generate a ton of evil, sycophantic, or hallucinated content. Fine tune/remove training data w/ similar vector. Can imagine much easier way to generate a “persona” (creative, direct, etc.) than prompting.
2025-08-02 View on X
Anthropic

Anthropic details “persona vectors”, patterns of activity within an AI model's neural network that control its character traits, such as evil and sycophancy

Read the paper  —  Language models are strange beasts.  In many ways they appear to have human-like “personalities” …

2025-07-20
35/42 on IMO 2025, no tools/internet, as graded by 3 IMO medal winners. Model unreleased, “not coming for months”, so take caveat. But I always stress when talking about AI limits: the major labs have not seen them yet, so “lines on a graph” for, say, a year are baked in already.
2025-07-20 View on X
@alexwei_

[Thread] An OpenAI researcher says the company's latest experimental reasoning LLM achieved gold medal-level performance on the 2025 International Math Olympiad

1/N I'm excited to share that our latest @OpenAI experimental reasoning LLM has achieved a longstanding grand challenge in AI: gold medal-level performance on the world's most pres...

2025-06-11
o3-pro: it thinks for a *long* time, should be given very very very long and specific instructions, and is very effective on tough problems (better than any we've seen). But of course, even o3 after 15 minutes can't escape overtraining on this modification of the old riddle! [image]
2025-06-11 View on X
Latent.Space

OpenAI's o3-pro is much smarter than o3 and amazing at using tools, but the model requires a lot of context to run well and without enough it tends to overthink

OpenAI dropped o3 pricing 80% today and launched o3-pro.  Ben Hylak of Raindrop.ai returns with the world's first early review.

o3-pro: it thinks for a *long* time, should be given very very very long and specific instructions, and is very effective on tough problems (better than any we've seen). But of course, even o3 after 15 minutes can't escape overtraining on this modification of the old riddle! [image]
2025-06-11 View on X
TechCrunch

OpenAI debuts o3-pro for ChatGPT Pro and Team users and in its API, costing $20/1M input and $80/1M output tokens; Enterprise and Edu will get access next week

OpenAI has launched o3-pro, an AI model that the company claims is its most capable yet.  —  O3-pro is a version of OpenAI's o3 …

2025-03-11
And more on AI science: Lila, out of Flagship Engineering (the company which sput out Moderna and was the final MBA core exam case Hong Luo and I wrote this year!), raises huge money for AI-driven automated lab science. George Church (!) is chief scientist. Already good results.
2025-03-11 View on X
New York Times

Cambridge, Massachusetts-based Lila Sciences, founded in 2023 to build AI to accelerate scientific discovery, emerges from stealth with a $200M seed

This is all well-and-good...provided that this is actually the process through which science works.  And it's not.  —  www.nytimes.com/2025/03/10/t... X: Kenneth Stanley / @kenneth...

2025-02-11
France just held a big AI summit. They want to build up their tech sector. Global prosperity would be enhanced by this. But I can't tell you how insane EU regulation is here, and how it inadvertently enhances market power of big US firms. Let's look at GDPR, for example. 1/x
2025-02-11 View on X
Reuters

Paris AI summit: US VP JD Vance warns the EU that excessive AI regulation could strangle the tech and rejects content moderation as “authoritarian censorship”

U.S. Vice President JD Vance warned Europeans on Tuesday that what he called their excessive regulation …

France just held a big AI summit. They want to build up their tech sector. Global prosperity would be enhanced by this. But I can't tell you how insane EU regulation is here, and how it inadvertently enhances market power of big US firms. Let's look at GDPR, for example. 1/x
2025-02-11 View on X
New York Times

Paris AI summit takeaways: Europe has regulation regrets, AI doomsayers have lost ground, and policymakers can't seem to grasp how soon powerful AI could arrive

Kevin Roose / New York Times :

2025-02-03
The new OpenAI model announced today is quite wild. It is essentially Google's Deep Research idea with multistep reasoning, web search, *and* the o3 model underneath (as far as I know). It sometimes takes a half hour to answer. Let me show you an example. 1/x [image]
2025-02-03 View on X
TechCrunch

OpenAI debuts Deep Research, an AI agent for creating in-depth reports, available to $200/month ChatGPT Pro subscribers and limited to 100 queries per month

OpenAI is announcing a new AI “agent” designed to help people conduct in-depth, complex research using ChatGPT, the company's AI-powered chatbot platform.