San Francisco, CA
Joined June 2020
- Tweets10.6K
- Following401
- Followers9.4K
- Likes18.4K
In my experience, the key distinction between those who embrace AI progress and those who want to stop it often comes down to whether they view AI as an extension of humanity or a competitor to humanity.
It's important to note that this distinction isn't an empirical one: it says nothing about the probabilities of various events occurring. Rather, it's a conceptual frame that determines what it means to preserve humanity and promote human flourishing.
Matthew Barnett retweeted
I think it is important to point out that, apart from unfortunate suicides involving people with mental health issues, almost nobody has died from AI/LLMs compared to the many lives that have been saved by AIs. I estimate about ~50 deaths from AI/LLMs compared to about 10000 lives saved. There are also, of course, millions upon millions of lives that have been improved by LLMs. It's important not to forget this in the current awakening on AI safety.
I suspect the Covid-AI analogy will fit our policy response too:
In both cases, cheap, well-targeted interventions are justified (vaccines, better RL environments).
Yet the public will go further than warranted and demand costly, sweeping action (lockdowns, pausing AI).
In the COVID-AI analogy, the last 24 hours feel like March 11, 2020, when Tom Hanks tested positive for COVID & the NBA suspended its season.
- 120 million views on @hilbertspaess’s resignation
- @DKokotajlo on Rogan’s podcast
- @paulfchristiano joins OpenAI board
The transition moment from “risk experts have been talking about” to “risk everyone is talking about.”
Matthew Barnett retweeted
It's frustrating when people say "AI progress is too fast" while over 100,000 people still die from aging per day, with no sign of abating. It's like we're in a huge, deadly war and people say our leaders are rushing to agree to a peace settlement. No, they should go even faster.
I support laissez-faire policy for open-weight models, in large part because it intensifies competition, especially with China, and therefore makes pausing or regulating AI less politically feasible.
For my first post, I’m sharing a letter @NVIDIA signed on why open models matter.
AI will transform every industry, power every company, and be built by every country.
Open models strengthen safety and cybersecurity, accelerate innovation and diffusion, and enable sovereignty.
The world needs both frontier closed models and frontier open models.
images.nvidia.com/pdf/Open-W…
I oppose this completely.
We should not allow technological progress to be held hostage by capricious and foolish government officials. I hope that the US government loses the power to impose these sorts of controls in the first place.
The US government, citing national security authorities, has issued an export control directive to suspend all access to Fable 5 and Mythos 5 by any foreign national, whether inside or outside the United States, including foreign national Anthropic employees.
The net effect of this order is that we must abruptly disable Fable 5 and Mythos 5 for all our customers to ensure compliance.
Access to all other Claude models is not affected.
We apologize for this disruption to our customers. We believe this is a misunderstanding and are working to restore access as soon as possible.
Read our full statement: anthropic.com/news/fable-myt…
Matthew Barnett retweeted
Claude Opus 4.8 scores 70.9% on GBA Eval, the top score to date. Given 24 hours, it writes an emulator that plays most games, with working audio on all of them. It beats the previous best (GPT-5.5 at 53.2%) in under an hour.
To be clear, I'm not sure whether current AI could pass this test. However, I think this prediction is on track to be falsified, even if not yet.
It's also how he operationalized a key premise in his AI doom argument with Bryan Caplan, so I think it's worth evaluating.
Matthew Barnett retweeted
Why would we deprive today's teens of our era's greatest source of information and support?
I'm guessing many Ant researchers greatly benefited from the internet as teenagers.
Growing up in the countryside in difficult family circumstances, I would have never become passionate about programming without it.
Anthropic Doesn't Allow Kids Under 18 — Here's Why
"We just don't know enough about what AI is going to do to kids. It needs to be done with an adult in the room. It needs to be done with a human in the loop." — @DanielaAmodei
In 2002, Ray Kurzweil and Mitchell Kapor made a bet on whether the Turing Test would be passed by 2029.
To determine the winner, they devised detailed terms, including the appointment of a committee to administer an actual exam, with a set of judges and human foils.
To assess whether Eliezer Yudkowsky is calibrated on AI doom, it seems relevant that in 2016 he said he'd be "pretty shocked" if an AI could pass an unrestricted one-hour Turing test before the end of the world.
He said this in the comment section of this post: econlib.org/archives/2016/03…