Yo Shavit (@yonashav) / X

Yo Shavit

3,326 posts

Yo Shavit

@yonashav

ai resilience @foundationOAI. Past: @openai / @HarvardSEAS / @SchmidtFutures / @MIT_CSAIL. Tweets my own; on my head be it.

New York, NY

Joined June 2010

Pinned
Yo Shavit
@yonashav
Jul 5, 2023
The data used to train an AI model is vital to understanding its capabilities and risks. But how can we tell whether a model W actually resulted from a dataset D? In a new paper, we show how to verify models' training-data, incl the data of open-source LMs!
arxiv.org
Tools for Verifying Neural Models' Training Data
It is important that consumers and regulators can verify the provenance of large neural models to evaluate their capabilities and risks. We introduce the concept of a "Proof-of-Training-Data": any...
160K
Yo Shavit
@yonashav
Dec 22, 2024
Now that everyone knows about o3, and imminent AGI is considered plausible, I’d like to walk through some of the AI policy implications I see.
608K
Yo Shavit
@yonashav
Dec 5, 2023
One interesting realization from moving inside openai is that a lot of the time, we have no idea what roon is talking about either
160K
Yo Shavit
@yonashav
Sep 23, 2025
The world’s first real evidence of scaling was posted to a small pocket of researchers 5 years and 3 months ago. By the end of the decade, the world economy will have built a network of infrastructure mega-projects to put the railroads to shame. The 21st century goes very fast.
OpenAI
@OpenAI
Sep 23, 2025
More compute in the making. Announcing 5 new Stargate sites with Oracle and SoftBank, putting us ahead of schedule on the 10-gigawatt commitment we announced in January. openai.com/index/five-new…
150K
Yo Shavit
@yonashav
Mar 10, 2025
These results are a massive deal, and overhauled the way I think about alignment and misalignment. I think this suggests a new default alignment strategy. Results and takeaways 🧵
OpenAI
@OpenAI
Mar 10, 2025
Detecting misbehavior in frontier reasoning models Chain-of-thought (CoT) reasoning models “think” in natural language understandable by humans. Monitoring their “thinking” has allowed us to detect misbehavior such as subverting tests in coding tasks, deceiving users, or giving
168K
Yo Shavit
@yonashav
Dec 22, 2024
Replying to @yonashav
Observation 2: The corporate tax rate will soon be the most important tax rate. If the economy is dominated by AI agent labor, taxing those agents (via the companies they’re registered to) is the best way human states will have to fund themselves, and to build the surpluses for
80K
Yo Shavit
@yonashav
Dec 22, 2024
Replying to @yonashav
Observation 5: Technical alignment of AGI is the ballgame. With it, AI agents will pursue our goals and look out for our interests even as more and more of the economy begins to operate outside direct human oversight. Without it, it is plausible that we fail to notice as the
50K
Yo Shavit
@yonashav
Dec 22, 2024
Replying to @yonashav
Observation 3: AIs should not own assets. “Humans remaining in control” is a technical challenge, but it’s also a legal challenge. IANAL, but it seems to me that a lot will depend on courts’ decision on whether fully-autonomous corporations can be full legal persons (and thus
93K
Yo Shavit
@yonashav
Dec 22, 2024
Replying to @yonashav
Observation 1: Everyone will probably have ASI. The scale of resources required for everything we’ve seen just isn’t that high compared to projected compute production in the latter part of the 2020s. The idea that AGI will be permanently centralized to one company or country is
72K
Yo Shavit
@yonashav
Nov 24, 2023
If you are a public figure and tell your followers that “big new risks from advanced AI are fake”, you are wrong. Not only that, you’ll be seen to be wrong *publicly & soon*. This is not an “EA thing”, it is an oncoming train and it is going to hit you, either help out or shut up
190K
Yo Shavit
@yonashav
Oct 27, 2025
@sebkrier and I are pretty floored by the quality of MATS applicants
38K
Yo Shavit
@yonashav
Apr 20, 2025
Replying to @willdepue
may god have mercy on our souls
18K
Yo Shavit
@yonashav
Mar 24, 2023
If AI models get even better, their unchecked use will begin to pose serious dangers to society. Most people agree it’d be great if countries could agree on rules to prevent AI misuse/accidents, & avoid an arms race. But how could rules on AI actually be enforced? Paper thread:
206K
Yo Shavit
@yonashav
Sep 12, 2024
advancing the frontier of legitimate procrastination
11K