Professional reference class tennis player. I like non-fillet frozen fish, packaged medicaments, and other oily seeds.
Joined June 2013
- Tweets3.2K
- Following134
- Followers2K
- Likes5.3K
Eric Neyman retweeted
The story keeps getting crazier. To recap everything that has been reported so far:
- The US intelligence community discovered that Jensen Huang, the CEO of the largest company in the world, was sleeping with Alice Huang, a Chinese spy and an executive at Megaspeed. (FT)
- Multiple independent investigations have concluded that Megaspeed is likely a Chinese cutout, which helps Chinese companies access AI chips or circumvent export controls. (NYT and Bloomberg)
- The US intelligence community also concluded that Alice Huang’s goal was to convince Jensen to lobby the U.S. government to loosen export controls on advanced AI chips. (NYT)
- The intelligence community relayed this information to “a small group of top US government officials” earlier this year. (FT)
- Over the last year, Jensen Huang has repeatedly lobbied the U.S. government to loosen export controls on advanced chips. @BISgov has done so, not only by permitting the export of H200 chips to China, but also not enforcing export controls that would have required a license for exports of the most advanced Blackwell chips to Megaspeed and Southeast Asia.
- Although @CommerceGov leadership was presumably briefed on this information by the U.S. intelligence community, @BISgov has taken no actions to tighten controls on Megaspeed or Southeast Asia.
- Megaspeed remains a preferred cloud partner of Nvidia, and continues to receive Nvidia Blackwell chips despite employing a Chinese spy who reportedly is romantically engaged with the CEO of Nvidia.
- Tens of billions of dollars worth of AI chips have been exported to Southeast Asia this year alone, and Malaysia and Singapore are now two of the four largest importers of AI chips in the world.
What is going on here??? We need answers. The question of who gets the most advanced AI chips in the world is literally the defining strategic question of our time. And not only does it seem like the MSS may be directly influencing the debate on export controls in the United States, but also @CommerceGov is refusing to listen to the warnings of the IC on this issue. What is a POSSIBLE good explanation for why Megaspeed is still receiving chips? Why is there not a license requirement for AI chip shipments to all of Southeast Asia? This is beyond strange at this point.
The fallout from this story will be massive. Obviously, the motivations underpinning Nvidia’s policy views opposing export controls must now all be questioned. But also, how did @BISgov sit on this information and has still taken no actions? We need answers.
US intelligence agencies believe @Nvidia CEO Jensen Huang had romantic relationship with alleged Chinese spy
as.ft.com/r/97c2f9f7-ad2e-44…
Eric Neyman retweeted
New: Spoke w/ @RepDonBeyer on his AI safety plans post-midterms
— He's hoping to roll out a draft of his bill with Rep. Jacobs this year to tee it up for potential Q1 action
— He made his case for an AI select committee
— Despite big Dem wins, he talked up AI bipartisanship
Eric Neyman retweeted
Here’s a pretty important for thing for a new Dem congress to get to work on right away!
This report and reactions from people I trust (like Jake) are making me physically ill. As far as I can tell, Anthropic's central plan is to continue this acceleration of AI capabilities until we're getting decades of AI capabilities growth in a year. Given our very limited ability to understand or control AI systems, this would be extremely reckless and completely unacceptable, and it should be the single top priority of civil society and the US government to stop them and other companies from doing this.
Eric Neyman retweeted
This report on AI R&D automation (RSI) is a big update towards ASI within 1-2 years. It says AI progress is sped up by RSI right now by around 2x. In AI 2027, this happens around March 2027.
I was seriously wrong about RSI speed up; I previously expected much less in 2026 (around 1.15x-1.3x). It seems as though there was specific privately available evidence that would have changed my mind.
I haven't seen an unredacted version of this report; this is purely about the public report. Because so much of the underlying evidence was redacted from the public report, I think conclusions about things other than the observed 2x faster rate of progress on benchmarks depend significantly on trust for the authors of the report. (E.g., conclusions about whether this speed up is from RSI depend on trust.) I do trust the authors of the report a lot to be careful and rigorous and generally do a good job. (Note that I also work at METR.) Less extensive redactions would greatly reduce this need for trust.
Since in my current role I may have access to other confidential information that could be relevant to timelines one way or another, I unfortunately cannot state my all-things-considered views on ASI timelines or the probability of a software intelligence explosion right now. Here, I’ve done some analysis of the implications using my previous public statements.
A 2x acceleration in overall AI progress from RSI corresponds to a much larger than 2x RSI speed up for AI software R&D (algorithms and data); I did an estimate using my public preexisting modeling of drivers of AI progress and the public report which yielded a very rough estimate of 4.5x for how much RSI is currently accelerating AI software progress (relative to no AI automation). In the AI 2027 scenario, this much acceleration happens around March 2027.
Here are the worked details of this estimate based on my public modeling. There are two main reasons RSI speed up of AI software R&D must be larger:
- Not all of AI progress is from better software (some is from scaling up compute in training runs). I've previously speculated that around 2/3 of AI progress is from software improvements which implies that 2x overall AI progress acceleration requires 3x faster software progress.
- Diminishing returns: A fixed amount of progress requires more work over time, so to ongoingly maintain 3x faster software progress via AI automation, you need increasingly large speed boosts from AI automation (as in, speed ups relative to how fast software R&D would go without AI automation). Based on this report, it seems that there is a substantial amount of additional AI progress from automation that has occurred.
We can estimate the effects of diminishing returns by first determining how many additional years of software progress we've seen (relative to the preexisting trend) and then estimating how much software R&D would slow down after a year of software progress if you held inputs fixed due to diminishing returns. I've previously estimated software R&D would slow down by around 2x for each year of software progress if you held inputs fixed (lesswrong.com/posts/FG54euEA…) and based on the report, it seems as though we've perhaps seen roughly 0.35-1.0 years of additional software progress incorporated into trained models so far (geometric mean of ~0.6). This combines to 3 * 2^(~0.6) = ~4.5.
(Relevant public modeling and numbers I'm using: lesswrong.com/posts/FG54euEA…, lesswrong.com/posts/Jga7PHMz…)
Eric Neyman retweeted
I've always had substantially longer timelines than my friends at Anthropic — I've been concerned that they're in an echo chamber and believing the hype about their own models too much, and that they're underestimating bottlenecks and unknown unknowns. And there have definitely been specific occasions when they have overstated how much uplift they were getting from their models. Nonetheless, if METR is correct here (and I think they are among the most trustworthy sources on this topic), then my Anthropic friends have been much more right about timelines than me.
This report is also some evidence that a software-only singularity is plausible. I now think that we are more than 50% likely to have full blown superintelligence in the next 18 months, absent a slowdown. I need to call my family.
Lost in all the noise around this tweet is that artificial superintelligence may literally kill us all.
Congresswoman, thank you for taking this issue with the seriousness that it deserves.
Polls underestimated Democrats almost everywhere, but there was also a regional effect: polls missed by a particularly large amount in the South. Here are the competitive Senate races with significant polling, ordered by size of polling miss.
Here's how much polls underestimated Democrats in each state (based on latest results / my extrapolation):
NC: 9.2
TX: 9.1
GA: 8.5
NH: 7.1
SC: 5.5
MN: 5.4
MS: 5.0
ME: 4.5
IA: 3.5
MI: 2.7
OH: 2.3
KS: 1.9
AK: ~0
NE: -4.9
Polls underestimated Democrats the most in the South.
Eric Neyman retweeted
There's a correct answer to the question of which Democratic faction is better at winning elections, and I was gratified to see my fave moderate Dems like Bobby Pulido and Jamie Ager overperform last night.
But the core story of the midterms is Republicans deciding to lose. The last two years have been a complete and utter dismissal of public opinion. The Trump administration has taken zero interest in addressing voters' top concerns, and Republican congressional candidates took no interest in breaking from a deeply unpopular president. It was a generational defeat, and it was deserved.
I'm grateful to live in a country where the kind of abject disregard for the electorate's preferences that Republicans have shown results in such a large shift in the balance of power. You shouldn't be able to govern like that without consequences—and, it turns out, you can't.
It's a good day in America. Lots of work to do, but it's a good day.
Eric Neyman retweeted
No legislator in the country has done more for AI safety than Senator @Scott_Wiener.
He campaigned strongly on his AI safety record and earned the support of hundreds of AI safety advocates from the very beginning of his campaign.
I can’t believe we can now call him Congressman-elect Wiener.
San Francisco Supervisor Connie Chan has conceded to state Sen. Scott Wiener in the race to succeed Nancy Pelosi in Congress, saying “we do not have a clear path to victory,” though the race has not officially been called for Wiener. sfchronicle.com/election/art…
Eric Neyman retweeted
I know the "smart" conventional wisdom is that midterms don't predict future election results. But then I look at this. And 2006 is the only modern lame-duck term with a wave like this.
Eric Neyman retweeted
Bad policy, bad character make bad politics: erratic tariffs, impulsive Iran war, blanket pardons, apparent corruption, weaponized justice, ICE brutality, mountainous deficits, hiding Epstein, abandoning allies and coddling enemies. Hoping for a course correction, for the country and for my party.
In my opinion, it would be a big mistake for Democrats to pursue this. I may write more about this in the coming days/weeks.
Yesterday, voters were profoundly at odds with Republicans over AI. Furthermore, 90% of voters (!!) said that government corruption was a crucial or major factor in their vote. In the end, Big Tech billionaires enabling government corruption really hurt the Republicans.
Last night was an *awful* night for polling -- we saw the largest polling bias in decades. But NYT/Siena's final polls were really good!
In the end, our final polls did not skew toward either side. Our old weighting approach would have overestimated Dems by 3.
The lesson: nonresponse was probably a bigger problem than in prior midterms, but weighting to a 2024 electorate (like many other pollsters) obviously wasn't a solution, either
Eric Neyman retweeted
To illustrate how the value of early voting analysis:
Two months ago, I prob had a worse forecast than any of these.
On my final forecast, finalized a few hours before polls close, I called 22 races wrong, and had a Brier score of just 0.03379 (using the same races as Lakshya)
This is a wonks-only postmortem, so please forgive me for one more: I think @NateSilver538 used to say that he'd prefer you just use Classic in a competition.
But if you allow it (and we should!), then the best Brier Score model of the cycle is clearly Silver Bulletin's Deluxe.
Eric Neyman retweeted
Statement from Humans First, the conservative pro-AI safety and anti- data center expansion group helmed by longtime tea party leader @AmyKremer & GOP operative @ChrisRBarron, on last night:
“Republican leadership in Washington gambled our majorities on big AI money and lost. They ignored their own voters, dismissed legitimate concerns, and put the interests of powerful donors ahead of the families and communities they were elected to represent. The leaders responsible for those decisions should resign from their leadership positions. Accountability cannot stop with the voters who paid the price.”
“This spring, I met with Speaker Johnson’s staff to warn them that a revolt was brewing at the grassroots level over this issue. I begged the Speaker’s office to do something, anything, to show that they understood these concerns. Those warnings went unheeded. Speaker Johnson should step aside and allow House Republicans to choose new leadership."
"If Republican leaders will not stand up to big AI and put their own voters first, they should step aside for leaders who will.”