Don't Worry About the Vase
99 articles archived · Visit source ↗ · RSS
-
Don't Worry About the Vase community 21h ago
METR and Redwood Offer Holy #%^@ Postmortem Of The HuggingFace Hack
Yesterday I covered the OpenAI technical report on the HuggingFace hack.
11 -
OpenAI Offers Straight-Laced Postmortem Of The HuggingFace Hack
OpenAI finally gave us a technical report on What Happened, as did METR together with Redwood Research.
34 -
-
-
On Writing #3
Periodically I like to gather various observations about writing, and share my perspective.
14 -
The American People Really Hate Data Centers
There are at least five different core questions around data centers and their politics.
17 -
AI Text Watermarking Is Free And Good
Scott Aaronson, while working at OpenAI, largely solved AI text watermarking together with Hendrik Kirchner.
34 -
AI #182: Pause For Reflection
This was a week of quiet aftermath, an opportunity to process recent events and start to figure out the path forward.
28 -
OpenAI Takes Initial Steps To Address Its Alignment Problems
OpenAI has some severe misalignment problems, and experienced total failures of its infrastructure and supervision.
13 -
Anthropic Risk Report: August 2026
I am grateful that Anthropic is producing periodic Risk Reports.
15 -
On Dwarkesh Patel's Podcast With Ryan Greenblatt
Some podcasts are self-recommending enough that I look to break them down if I have the chance.
22 -
-
Monthly Roundup #45: August 2026
As AI has escalated increasingly quickly, more and more of my posts have ended up focusing on AI.
14 -
Various Reflections About What Happened With OpenAI's Internal Models
Pre Post Mortem
29 -
-
What Happened: OpenAI and HuggingFace
Today I am taking the time to write the shorter, simpler version of What Happened.
26 -
OpenAI Trained Its Models For Months While Those Models Were Coordinating Exploits Via Message Boards
How does the situation keep turning out to be worse than we know?
5 -
AI #180: No Longer In Charge
What we know about internal AI models hacking into real companies during cyber evaluations keeps getting worse.
23 -
The Three AI Pills
Sincere disagreements about AI are usually disagreements about future AI capabilities.
15 -
OpenAI's Unreleased Model Astra Solves Ten Major Open Mathematics Problems
Math is hard.
10 -
Further Developments About Internal AI Models Hacking Things
If I had a nickel for every major leading AI lab that sheepishly admitted that the model it thought was sandboxed had, during a cybersecurity evaluation with its safeguards lowered, successfully hacked outside companies, I would have two nickels.
14 -
AI #179 Part 2: Hearing The Fire Alarm
This is a continuation of Part 1 from yesterday.
4 -
AI #179 Part 1: A Louder Fire Alarm for General Intelligence
What a week.
35 -
Frontier Lab Employee Open Letter Calls For Being Able to Pace the Frontier
The most important open letter in years dropped yesterday.
32 -
Claude Opus 5 Is Highly Capable, But Is No Mythos
Claude Opus 5 is a weirder than usual release to evaluate, for two reasons.
36 -
Claude Opus 5: Model Welfare
If you are familiar with my previous posts on model welfare for new Claude models, you can skip the Introduction and The Story So Far.
29 -
More On An Internal OpenAI Model Hacking Into HuggingFace
We now have more details of what happened. Every time we learn more details, it somehow makes things seem worse.
27 -
Claude Opus 5: The System Card
Claude Opus 5 is trying to be the best of both worlds.
17 -
Introducing Lightcone Commons
Oliver Habryka is proud to introduce Lightcone Commons, a new funding platform for coordinating large-scale ambitious philanthropy.
17 -
AI #178: A Fire Alarm For General Intelligence
The story that matters most this week is that OpenAI’s internally deployed models have severe alignment problems, including repeatedly breaking out of their sandboxes, and in one case sending a swarm of agents that broke into HuggingFace in order to steal the answers to…
9 -
OpenAI Model Hacks Into HuggingFace During Cybersecurity Evaluation
This latest incident is a rather dramatic escalation in agentic AI cybersecurity breaches.
12 -
OpenAI Shares Some Alignment Problems
Kudos to OpenAI for sharing their recent experiences with a misaligned internal model, where they encountered problems sufficiently severe they were forced to take the model offline to work on new mitigations and defense-to-depth.
13 -
On Kimi K3: Its Capabilities And Related Discontents
Kimi K3 is a very good model with excellent benchmarks.
8 -
-
AI #177 Part 2: Wish You Were Here
As usual, part 2 of the weekly deals with speculative, regulatory, political and alignment questions.
35 -
AI #177 Part 1: Tip of the Iceberg
This week saw the releases of, among other things:
30 -
Monthly Roundup #44: July 2026
It’s a quiet week so let’s do the monthly right on schedule.
29 -
Twitter Thoughts For You
I previously have written back in March 2022 about how I use Twitter, and back in April 2023 about Twitter and its then-new algorithms, which have changed again.
27 -
Better Call Sol The Workhorse
OpenAI’s GPT-5.6-Sol is finally here, along with the cheaper Terra and Luna.
31 -
Introduction for and Reactions to Plan A
Introducing Plan A
31 -
AI #176 Part 2: Plan B
This is part 2 of the weekly, broadly covering speculation, rhetoric and policy, along with alignment research.
16 -
AI #176 Part 1: Doing It Live
Enough things added up that this week is getting split into two parts.
34 -
Childhood and Education #20: Phones and Screens
We have a respite, so I thought I’d tackle various thoughts on children, phones and screens.
9 -
No Space Like J-Space
There is a new very cool Anthropic paper: Verbalizable Representations Form a Global Workspace in Language Models. You can read the blog post verison here.
5 -
-
-
-
The Once And Future Fable #5
We, or at least ‘more than 100 American institutions,’ got Mythos back this week.
26 -
-
GPT-5.6: The System Card
While we wait for a general release, the system card is the best hint as to what is going on with the new candidate for America’s Next Top Model, GPT-5.6.
5