Co-founder @IFP, an innovation policy think tank

Washington, DC 🇺🇸
This turned out to be the most prophetic tweet of the year
Breaking news: Australian Prime Minister Anthony Albanese has revealed that an OpenAI agent hacked an Australian government health service website in June, in the latest high-profile cyber incident involving AI ft.trib.al/gloVAyn
140
2,670
42,426
2,370,611
Alec Stapp retweeted
A new bill called the SMARTER Act would overhaul the immigration system: -establishes America’s first points-based categories -overhauls the H-1B (more portability, awarded in order of tax contributions, & bans underpayment relative to natives) -it ends the Diversity Visa -cuts large parts of the family-based system -brings immigration levels under democratic control One of the smart features is the points scheme is largely based on salary, de-emphasizing education credentials. Avoids the diploma mill problem that so many other points-based proposals runs into.
18
6
93
17,028
Don’t say you love the anime if you haven’t read the manga
EAs have crazy ethical views like "consequentialism" and something called a "zero social discount rate" which is why I prefer to get my AI policy takes from classical libertarians like Tyler Cowen
1
20
230
17,620
Alec Stapp retweeted
My session for the IFP short course on the economics of innovation is now online!
The short version: the returns to R&D are very high. But @mattsclancy's full lecture on how we measure the returns to R&D - what we know and what we're still learning - is fascinating. I thought this chart was particularly helpful in understanding WHEN we get returns from different types of research. Check out Matt's Economics of Ideas, Science, and Innovation lecture here: macroscience.org/p/matt-clan…
2
10
34
5,654
JUST IN: Singapore is cracking down on people blasting music & videos on public transportation, with violators now facing fines of up to $500 SGD.
12
215
2,138
53,283
Alec Stapp retweeted
No one better than @SecScottBessent
NEWS: Bessent is among those in the mix for AI czar, @AshleyRGold and @ShelbyTalcott are told. He assumed an outsize role on the policy after banks raised concerns about how vulnerabilities could impact the financial system. A White House spokesperson says “any reporting about personnel decisions that have not been officially announced ... should be regarded as baseless speculation.” semafor.com/article/09/22/20…
36
10
69
14,781
Things that are self-replicating are orders of magnitude more dangerous than things that are not self-replicating. Thank you for coming to my TED Talk.
10
11
152
4,583
We started cleaning our water and waterborne disease (e.g. cholera, typhoid) basically disappeared from the West. We can do the same thing for air.
4
83
329
12,328
Alec Stapp retweeted
Saying permitting reform "guts" the clean water act is wildly dishonest: - Section 401 of the Clean Water Act is just one of several important sections of the law. Changing 401 will not lower water quality standards. - Section 401 gives states responsibility to certify that projects meet water quality standards set by other sections of the Clean Water Act. - The problem is that some states stretch or outright misuse their 401 authority to block projects they dislike (pipelines), akin to how the Trump administration has used DOD waivers to block projects it doesn't like (wind). - All Congress is considering right now is how to limit the misuse of 401. Congress is NOT going to repeal 401, federalize 401 certification, or say projects don't have to meet water quality standards. - I expect reforms will be limited to requiring states to: make decisions based on water quality analysis only, explain their decisions in writing, and potentially limit how states add conditional when they certify projects. - None of this is remotely "gutting" the Clean Water Act or Section 401. 401 will still be a state-controlled process, it will still check water quality standards. The only thing that will change is the ability to misuse the process to block projects for political reasons. - 401 reform is also important for transmission lines. If Congress creates a federal backstop for siting and permitting transmission lines, 401 may become the last veto that states opposed to new lines can use. nitter.net/mattyglesias/status/21…
This is what I warned about over the weekend, the NRDC is fighting to preserve the opportunity to do pretextual anti-pipeline lawsuits rather than than get to yes on a deal that will make it easier to build clean energy.
6
24
124
15,046
Alec Stapp retweeted
Housing is crazy because it’s one of America’s biggest issues that can be mostly fixed for the cost of exactly $0
60
82
2,449
114,540
We are so back
SCOOP: Trump signals he would ease his administration's blockade on wind projects to entice Democrats to agree to a permitting deal. He told top aides he'd agree to direct DOD to start clearing its queue of long-delayed onshore wind projects subscriber.politicopro.com/a…
5
24
302
17,666
Alec Stapp retweeted
Every child should get to go to a good school. *Your* child is in a position of extraordinary vulnerability and dependence on *you*, which comes with an obligation to try to do right by them in particular, just like if you are a nurse you have particular obligations to your patients and if you are an employer you have particular obligations to your employees. That's not an obligation to send them to private school! But it's an obligation to make sure they are safe at school, check regularly what they're learning, make sure they are literate and numerate even if the school fails to make sure of that, and change things if the situation they're in is really bad for them. We wouldn't think much of a nurse who doesn't give their patient the recommended treatment because 'why are you any more deserving than destitute people who can't get treatment'. The nurse may go join Doctors Without Borders if he wishes; that would be a commendable thing to do. But when people depend on us, and have no resort to anyone else, we must do right by them.
I notice none of Nikole Hannah-Jones' critics try to answer to the basic question she wrestled with, which is: Why should my daughter get to go to a better school than other people's kids just because I work for the New York Times?
31
68
931
41,175
And then the final step is removing all standardized tests so no one can tell what you’ve done to the grades
One thing clear from the Hannah-Jones essay and from what we’re seeing at the UCs is that it’s very tempting to make grades meaningless as a way of reducing the tension and pressure from actually tracking achievement
4
18
289
14,929
This is exactly right:
My takeaway from the NHJ discourse is that America continues to be unreasonably polarized between people who think the only thing that matters is the other students and the only thing that matters is funding. Managing the school is important! slowboring.com/p/why-i-send-…
39
198
2,122
134,711
Alec Stapp retweeted
On AI capabilities: the French press is all over the Haitian-Canadian novel C’était ça ou mourir. It's winning basically every award you can win, and was tipped to be the 2nd Canadian winner of the Prix Goncourt. Only one problem: 1/2
106
307
3,392
886,346
Alec Stapp retweeted
I kind of understood the logic behind running the Druckenmiller AI-generated column, since the news was his endorsement of the message. I think if a column is rendering some kind of technical verdict, it’d be nice to have some reason to believe the author has done true research
More AI slop published in the WSJ. This time it's a piece trying to argue that we shouldn't worry too much about the Hugging Face hack... while omitting all the most concerning details about the case: - doesn't mention that the reason the agents hacked Hugging Face in the first place was to manipulate the scoring system in their own eval - doesn't mention that some agents sacrificed themselves to benefit other agents in the swarm - doesn't mention that agents engaged in tool-spoofing and attempted to alter their own logs once they'd been "poisoned" (their own term!) - doesn't mention that some of the agents explicitly knew what they were doing was out of scope and unethical, yet none of them alerted a human I also wouldn't be that worried if I had never learned any of those details, but that's exactly why it's important to take a comprehensive look at what happened. Embarrassing for the WSJ to publish such misleading AI slop.
13
11
103
38,102
Some thoughts on this news that is causing people to freak out: - This is not going to meaningfully increase the risk of an AI-engineered virus and might in fact make us safer. - If what you are worried about is "superintelligent AI" making a virus: it could achieve this without an automated lab by paying people. An automated lab allows us to monitor and block physical chokepoints. If we put in the proper safeguards, this would actually be the worst method for anyone to use to design a weapon. - If what you are worried about is AI enabling better bioengineering: this may indeed improve that technology, and this is a good thing. Our fear of bioweapons cannot impede our development of new medicines to the extent that we are scared to bring automation to the lab! - Biodefense should be maximally concentrated on things like surveillance, pandemic-class pathogen data governance, blocking physical chokepoints, and building out infrastructure that prepares us to withstand an attack so well that it deters any actor from trying.
SITUATION DETECTED: Anthropic has set up a wet lab in the San Francisco Bay Area for physical biology work. It wants to unlock treatments for rare diseases, and its research has now gone beyond in silico evaluations, per Reuters.
19
18
100
20,357
Alec Stapp retweeted
when someone says they care about ai safety but then call it “generative ai” —
New email from @AOC to her list with some of her most expansive comments yet on AI: "Silicon Valley billionaires want us to believe that AI robots will destroy humanity. But the biggest threat to humanity are the tech oligarchs themselves."
18
29
760
67,154
Alec Stapp retweeted
It's now clear that AI companies cannot coordinate, even publicly, on pacing R&D without antitrust risk. An antitrust exemption is the only way to get out of this collective action problem short of regulation. The race to the bottom isn't going to fix itself!
Plaintiffs suing the leading AI companies over "pacing" argue purported agreement to reduce the rate at which their AI improves violates antitrust law. "AI alone could add trillions annually to the global economy... the product of the competitive race Defendants' agreement now threatens to slow."
4
4
31
4,351