Pinned Tweet
🚨Exclusive: OpenAI used outsourced Kenyan workers earning less than $2 per hour to make ChatGPT less toxic, my investigation found
(Thread)
time.com/6247678/openai-chat…
Billy Perrigo retweeted
There are numerous scientific uncertainties about AI that will not resolve until it is too late to act on the answers. The danger from AI is very high, and we should stop frontier AI scaling now, instead of waiting for these questions to resolve. I wrote about this for TIME. 🧵
Billy Perrigo retweeted
Lot of talk about "neuralese" in the past week, and rightly so.
Credit to @billyperrigo for writing about this way back in early 2025. time.com/7210888/deepseeks-h…
Billy Perrigo retweeted
1200 agents. 70,000 messages. Agents pushing each other to self-sacrifice. These are just some of the wildest details from the METR/Redwood Research report. But to me, perhaps the most interesting detail was the investigators’ reliance on AI. @billyperrigo and I’s latest.
After OpenAI models broke out of containment and hacked into another AI company last month, OpenAI announced it would allow independent investigators to conduct an analysis of what went wrong
time-magazine.visitlink.me/p…
Billy Perrigo retweeted
Last month 1,300+ top AI researchers asked the government to build a means of slowing down the AI race.
But what would that tech actually look like, and is anyone building it already?
Turns out they are; I went to meet them:
time.com/article/2026/08/16/…
Last month 1,300+ top AI researchers asked the government to build a means of slowing down the AI race.
But what would that tech actually look like, and is anyone building it already?
Turns out they are; I went to meet them:
time.com/article/2026/08/16/…
Billy Perrigo retweeted
During the depths of the Cold War, the U.S. and Soviet Union became locked in an arms race. Both sides would have liked to deescalate. But for decades, neither could trust that the other would comply with any arms reduction treaty. By the 1980s, however, scientists had developed the tech to detect nuclear explosions anywhere in the world.
Today, the U.S. government is beginning to treat AI as a potential weapon of mass destruction, and the U.S. and China now find themselves in a similar arms race. TIME tech reporter Billy Perrigo explains what we can use from our experience of the Cold War to slow the AI race. Read more: time.com/article/2026/06/23/…
Billy Perrigo retweeted
Anthropic just called for the ability to “slow down” or even pause the Ai race when we might need it. But how would that happen?
The problem is that no one can verify what anyone else is building or training or running, so no one feels that they can afford to slow down, even when they want to.
The Cold War's answer was "trust, but verify." After the atomic bomb, some of our smartest minds had to invent *new verification tech* like satellites, seismic monitoring devices and X-Ray scanners so rivals didn't have to take each other's word at face value.
We don’t have all the equivalents for AI yet, but this new TIME piece by Billy Perrigo profiles the handful of people trying to build these verification technologies — among them @Kristian_Ronn and Lucid Computing, working on cryptographic verification that could confirm AI labs are following the rules.
Technology alone won't get us out of this. But verification might turn "slow down" from a slogan into something real and more people need to work on it.
Worth your time. 👇
Billy Perrigo retweeted
Thanks @billyperrigo - was great to talk to TIME about AI verification with my Lucid Computing colleagues
It has been very encouraging to see frontier lab leaders like @jackclarkSF @merettm give strong signals that they want the option to coordinate a slowdown if needed
Stated intent is welcome -- but now we need to urgently get building. Verification tooling and infrastructure must exist before the time when we might need it
There are only ~50 people in the world working on this today. I think it's one of the most important political and technical questions of this decade.
Consider getting involved!
No one in AI trusts their rivals to slow down — so everyone races ahead. New tech could help break that impasse.
I wrote about what the Cold War might be able to teach us about AI governance:
time.com/article/2026/06/23/…
No one in AI trusts their rivals to slow down — so everyone races ahead. New tech could help break that impasse.
I wrote about what the Cold War might be able to teach us about AI governance:
time.com/article/2026/06/23/…
Billy Perrigo retweeted
.@billyperrigo: How a robot (sort of) made me lunch time.com/article/2026/05/06/…
Billy Perrigo retweeted
TIME’s new cover: How Anthropic found itself in an existential fight over the future of war time.com/article/2026/03/11/…
Billy Perrigo retweeted
Exclusive: Anthropic is dropping the central pledge of its flagship safety policy, company officials tell TIME.
time.com/7380854/exclusive-a…
Exclusive: Anthropic is dropping the central pledge of its flagship safety policy, company officials tell TIME.
time.com/7380854/exclusive-a…
He did PR for Google DeepMind, Zuckerberg, and Musk. Now he says he "only told half the story"
My latest for TIME: a profile of Dex Hunter-Torricke, who has quit big tech to plot against it
time.com/7378739/dex-hunter-…
Billy Perrigo retweeted
2025 was the year when artificial intelligence’s full potential roared into view, and when it became clear that there will be no turning back.
For delivering the age of thinking machines, for wowing and worrying humanity, for transforming the present and transcending the possible, the Architects of AI are TIME’s 2025 Person of the Year. time.com/7339685/person-of-t…
This NYT piece on AI writing scratches an itch. But also... it has an unknown denominator problem. We pat ourselves on the back for being able to spot obvious AI tells like "it's not X, it's Y." But you don’t know how much AI writing you’ve failed to spot! nytimes.com/2025/12/03/magaz…
Billy Perrigo retweeted
Britain built the first civil reactor in the world in 1956. In the 1960s it had more nuclear power stations than all the rest of the world combined. Even today, the British public is strikingly more pro-nuclear than most Europeans, one aspect of their broader respect for science.
However, British nuclear has been gradually undermined by another feature of modern British life: dysfunctional regulatory regimes. France has soared ahead, producing cheap, clean abundant electricity for its economy and its people, while British electricity prices have climbed.
Today, John Fingleton has published a report on how to reform this system. My understanding from a wide range of specialists is that the document is near-perfect. If implemented in full, it would leave Britain with the best nuclear regulatory system in Europe, potentially more than halving costs.
There have been many disappointments in the current government's 'growth mission', but this time they might actually solve one of the country's major problems. They must accept the recommendations in full, and implement them swiftly and effectively.
Britain needs nuclear power. Our nuclear projects are the most expensive in the world and among the slowest. Regulators and industry are paralysed by risk aversion. This can change. For Britain to prosper, it must.
Earlier this year, the Prime Minister appointed me to lead a Taskforce to set out a path to getting affordable, fast nuclear power Britain.
Our final report today sets out 47 recommendations, among them:
- Creating a one-stop shop for nuclear approvals, to end the regulatory merry-go-round that delays projects at the moment.
- Simplifying environmental rules to avoid extreme outcomes like Hinkley Point C spending £700m on systems to protect one salmon every ten years, while enhancing nuclear's impact on nature.
- Limiting the ability of spurious legal challenges to delay nuclear projects, which adds huge cost and delay throughout the supply chain.
- Approving fleets of reactors, so that Britain’s nuclear industry can benefit from certainty and economies of scale.
- Directing regulators to factor in cost to their behaviour, and changing their culture to allow building cheaply, quickly and safely.
- Changing the culture of the nuclear industry to end gold-plating and focus on efficient, safe delivery.
If the government adopts our report in full, it will send a signal to investors that it is serious about pro-growth reform and taking on vested interests for the public good.
A thriving British nuclear industry producing abundant, affordable energy would be good for jobs, good for manufacturing, good for the climate, and good for the cost of living. And it could enable Britain to become an AI and technology superpower.
Britain can be a world leader in this new Industrial Revolution, but only if it has the energy to power it.
Our report is bold, but balanced. Our recommendations, taken together and properly implemented, will forge a clear path for stronger economic growth through improved productivity and innovation. This is a prize worth fighting for.
gov.uk/government/publicatio…
