Yo Shavit
@yonashav
Followers
7K
Following
6K
Media
92
Statuses
3K
policy for v smart things @openai. Past: CS PhD @HarvardSEAS/@SchmidtFutures/@MIT_CSAIL. Tweets my own; on my head be it.
New York, NY
Joined June 2010
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! https://t.co/DKQf4eoC5y
arxiv.org
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...
7
66
374
Today, @ValthosTech launches with $30M from @Lux_Capital, @OpenAI, and @foundersfund. For the past few months I’ve been working with their world-class team (@MIT, @GoogleDeepMind, @PalantirTech) building America’s modern next-gen infrastructure. Hiring –– DM me.
Valthos builds next-generation biodefense. Of all AI applications, biotechnology has the highest upside and most catastrophic downside. Heroes at the frontlines of biodefense are working every day to protect the world against the worst case. But the pace of biotech is against
38
31
306
BREAKING: Mamdani's odds collapse in NYC Mayoral Election. If he continues falling at the rate he has the past 24h, Cuomo would be the projected winner.
2K
2K
17K
there are more things in heaven and earth, Horatio, than are dreamt of in your RL environment
1
10
143
(1/11) How do AI agents change markets? We speculate on this topic in our new article for the NBER “Economics of Transformative AI” volume with @gilirusak, @BenSManning, @AndreyFradkin, and @johnjhorton: “The Coasean Singularity? Demand, Supply, and Market Design with AI Agents”
12
49
168
עוד תיעודים של הפוגרום שביצעו מתנחלים היום בכפר הפלסטיני תורמוס עיא בגדה המערבית הכבושה
A horrific video shows Israeli settler terrorists attacking Palestinian olive pickers in the village of Turmus Ayya. In the footage, American activist @infinite_jaz, who documents settler crimes, is seen fleeing as settlers chase him and attempt to kill him.
213
702
3K
📢New piece on 🇨🇳 AI governance doc📢 China's key AI standards body put out "AI Safety Governance Framework 2.0" - taxonomy of risks (focus shifting) - technical mitigations (more sophisticated) - roadmap for future technical standards @Scott_R_Singer & I break it down 🔗👇
1
4
13
Spritz Vibe. Limited Edition. Frosted over & fresh for the season, Spritz Vibe Sparkling Snowball Frost Limited Edition is here! CELSIUS. LIVE. FIT. GO.
469
733
11K
Yeah, my main issue with the Mechanize thesis is that while the incentives make it inevitable, different social configurations are not. Putinist full automation would result in a different social and power structure than Swiss full automation. We can affect which we’ll land in.
As always, I think Mechanize is analytically sharp, but doesn't stick the normative landing. Delaying full automation is good because it gives us time to restructure economic and political institutions that hinge on labor being valuable. For that, we need every year we can get.
9
5
121
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.
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. https://t.co/AWJQt02XZ4
13
79
1K
Excited to share details on two of our longest running and most effective safeguard collaborations, one with Anthropic and one with OpenAI. We've identified—and they've patched—a large number of vulnerabilities and together strengthened their safeguards. 🧵 1/6
8
62
300
The Cathedral has been destroyed, the youth is consumed by nihilism, mental illness and relativism. We can argue whether that's the fault of The Cathedral itself, but I'm not seeing anything promising rising from the ashes, at least not coming from those who attacked it.
It was sparked in its day! You can see its ancestry in early 19C Exeter Hall abolitionism. It is possible to read the 19C and stay within the liberal canon. The old regime that Dostoyevsky’s demons opposed was not a revolutionary regime built on a chiliastic ideology. Ours is
18
21
234
Making cooperation great again with @yonashav 🖇️
MATS 9.0 applications are open! Launch your career in AI alignment, governance, and security with our 12-week research program. MATS provides field-leading research mentorship, funding, Berkeley & London offices, housing, and talks/workshops with AI experts.
4
8
168
me: what could i have done in my past to deserve this karma me during the mesozoic era:
240
31K
196K
The speculated B30A would be a really good chip. “50% off” is false reassurance. -½ B300 performance, ½ price = same value (just buy 2x) -Well above (12x!) export control thresholds -Outperforms all Chinese chips -Delivers 12.6x the training perf of the H20 -Better than H100 1/
11
52
236
congrats to @__nmca__ for becoming more cracked than every Meta employee and every Chinese national
Two weeks ago, we launched The Metis List. Since then, we've spoken with many of you and have updated the ranking accordingly. 128 top AI researchers, ranked by their peers.
0
0
10