It might seem quite surprising that Eliezer Yudkowsky and Bernie Sanders have wound up allies — but there's a case to be made that AI safety would always need to take on a populist, anti-elite form. The AI safety movement needs normies: www.transformernews.ai/p/the-ai-saf...

AI Safety advocates have known for a while now that the near-future's going to get very, very weird. What they didn't know is that they might need people who aren't. The AI safety movement needs normies: www.transformernews.ai/p/the-ai-saf...

Our newest newsletter's up now, if you want to catch up with the quite frightening advances Anthropic has made with Mythos (which is perhaps most discernible from looking elsewhere in the newsletter, in how it's being responded to by the political world) www.transformernews.ai/p/the-contra...

New week, new newletter, and the big story in this one is: did anyone notice that what Jensen Huang said about China on Dwarkesh Patel's podcast this week didn't really make any sense? Here's why his contradiction has enormous ramifications. www.transformernews.ai/p/the-contra...

If you want to spend less time on social media but, like us, you simply can't miss a second of all the beefing and cattiness among the upper-echelons of the AI industry, we can not recommend strongly enough that you subscribe to our wonderful newsletter. www.transformernews.ai/p/pentagon-a...

"Chatbots are not a cure-all for mental health... [but] chatbots are already filling critical gaps." In a bracing new opinion piece, Jess Miers and Ray Yeh argue on behalf of AI as a vital mental health resource. www.transformernews.ai/p/less-liabi...

In 2023, California congressman Ted Lieu introduced what he called the first piece of federal legislation written by AI, using ChatGPT to generate a resolution expressing support for Congress’s focus on AI. Now AI is becoming prevalent in writing laws — so what's the risk in using it?

“Who the fuck knows how to align superhuman AI?” We might be too presumptuous when we think human intelligence can handle AI superintelligence. So here's a radically counter-intuitive idea for you: the case for automating alignment research. www.transformernews.ai/p/ai-alignme...

"Despite individual CEOs expressing concern about the risks posed by their technology, none seem willing to slow down unless everyone else does, too." AI alignment researchers want to automate themselves. Here's why that could be a pretty big problem. www.transformernews.ai/p/ai-alignme...

“The future is already here – it's just not evenly distributed" — William Gibson. Psst...have a click around on our homepage to keep around with the bizarrely futuristic stories that make up this present moment: the weird, weird era of AI. www.transformernews.ai

Whether you want to know all the Mean Girls goss on Altman and Amodei's beef, or how ChatGPT has taken Oppenheimer's old job, or about the surprising popularity of insipid-sounding AI-generated dating shows, we've got it *all* for you in our newsletter: www.transformernews.ai/p/ai-populis...

AI labs are getting more and more into the habit of letting AI train itself, and even contribute to AI safety work. And since the science shows that LLMs can't really be trusted yet, that *might* just lead us into some very dark places... www.transformernews.ai/p/ai-alignme...

After our terrifying break of a single week, we're delighted to inform you that the Transformer weekly newsletter is back! And yet again, it has all the AI news you could possibly desire all in one very readable place! Incredible stuff. Here's the link: www.transformernews.ai/p/two-fronts...

The latest Transformer weekly newsletter is here to keep you on top of all the big AI moves — and there've been a lot of big, big moves lately. You can quickly catch up with the rest of what's going on right here: www.transformernews.ai/p/two-fronts...

It's the cutting edge of the cutting edge, the final frontier of the jagged frontier: with our weekly newsletter, you can stay on top of ongoing research about advancing AI capabilities. So why not have a curious little browse around, or even subscribe? www.transformernews.ai/p/two-fronts...

Illinois, which has classically been considered the political bellweather state of America, has aptly become the bellweather state for AI regulation — which means a lot of money and a lot of friction are starting to appear in the area. www.transformernews.ai/p/ais-next-b...

In our latest opinion piece, Konrad Körding and Ioana Marinescu argue that AI will likely have a more limited impact on jobs than people think, due to the physical reality of physical work — and point out what policy makers should do in light of this. www.transformernews.ai/p/the-key-de...

So it seems not everyone's happy about Jensen Huang's relationship with Trump... www.transformernews.ai/p/not-everyo...

If you want to understand NVIDIA, you should understand its unusually close relationship with the White House. Not everyone’s happy about Jensen Huang’s direct line to Trump www.transformernews.ai/p/not-everyo...

The White House is trying to make AI a partisan issue again. And their priorities are quite revealing — the threat of wide-scale loss of jobs is absent, and the framework pushes for “commercially reasonable" child safety requirements. Read more here: www.transformernews.ai/p/the-white-...

A wave of legislation targeting chatbots such as ChatGPT and Claude has emerged in six states since the start of the year — but there are vital gaps in these bills which show they might not be up to task when it comes to protecting children.

Our chances of solving the AI alignment problem are heavily dependent on there being a political will to solve it — so if there's no political drive for AI regulation, we could be in a lot of trouble indeed. No, the AI alignment problem isn't solved. www.transformernews.ai/p/no-ai-alig...

How an AI industry-backed lobbying group has spent half a million dollars on buying a grassroots movement. www.transformernews.ai/p/how-to-buy...

Bring up "recursive self-improvement" in Washington, and people are likely to laugh. Bring it up in San Fransico, and the mood quickly turns somber. The big story our latest newsletter: AI's already more responsible for developing AI than you think. www.transformernews.ai/p/the-fuse-i...

“The instability of that single week in February was a preview of what it feels like when the increasing ability of AI starts to interact with markets, jobs, and governments all at once.” In our newsletter, the future's starting to freak people out. www.transformernews.ai/p/the-fuse-i...

In the 14th Century, Mansa Musa would give away so much gold on his travels that it'd devalue the metal and cause economic crisis in the region. Nowadays, Anthropic employees have pledged to give away billions (and billions) to EA-aligned nonprofits. So what's going to happen?

Along with all the other pretty dramatic AI news lately, the big story you might be missing is just how much money is being spent lobbying both for and against AI regulation. Both sides claim the lead in AI’s high-stakes midterm race: www.transformernews.ai/p/alex-bores...

"We come from different organizations, different backgrounds. We have different concerns: safety, ethics, labor, democracy, art, human dignity. We’ve all arrived at the same conclusion: slow down." So what gives with this new wave of anti-AI protests? www.transformernews.ai/p/scream-if-...

The Transformer weekly newsletter is here! The big AI story has been around Anthropic, the Pentagon and the White House, but our Big Story this week is focusing on the big, big money moving around AI and campaign financing. Stay on top of all of it here: www.transformernews.ai/p/what-the-f...

Introducing the Transformer AI Campaign Finance Tracker — the ultimate tool for monitoring AI's influence on elections.

You've probably got.a lot of questions about autonomous weapons, and you likely have a lot of anxieties about them too. We've got a lot of answers, though we probably can't help with those anxieties. What you need to know about autonomous weapons: www.transformernews.ai/p/what-you-n...

He's fought against regulating cryptocurrency, and now he's helping OpenAI avoid regulation too. Here's what Chris Lehane has brought to OpenAI, and how it symbolises just how much they've changed in the last few years.

He helped crush crypto's enemies, and he might be the clearest symbol there is of what OpenAI is transforming into, or even how it lost its way. On the "guerilla warrior" who taught OpenAI how to fight: www.transformernews.ai/p/the-gueril...

Pentagon CTO Emil Michael called Dario Amodei “a liar [with] a God-complex”, and OpenAI have raised an ungodly $110b, which could help with building the machine God. Catch up with all the week's AI news via the God-like omniscience of our weekly update: www.transformernews.ai/p/anthropic-...

"Humans have come up with lots of creative ways to kill each other over the last few millennia. Our deadliest mass murders, however, have mostly been incidental." Concerning the nightmare scenario: just how worried should we be about AI biorisk? www.transformernews.ai/p/ai-biorisk...

"It is not hard to picture how such an administration would weaponize AI-enhanced surveillance." This is exactly how the leading AI labs could already be leading us into a technological dystopia. www.transformernews.ai/p/the-dod-fi...

"We cannot be at all confident that OpenAI can or will actually prevent the Pentagon from using its models to conduct mass surveillance on Americans. We just have to take the company, and the government’s, word for it. History suggests we shouldn’t." www.transformernews.ai/p/openai-pen...

Mass surveillance has never been very popular with the US public. But the untested ground of AI-enhanced state surveillance could be an existential threat to the future of American democracy itself. www.transformernews.ai/p/the-dod-fi...

"As the Snowden files revealed, mass surveillance is nothing new. But AI could take it to an entirely new level, creating a true digital panopticon." In enabling AI-enhanced surveillance, could these major AI labs foster a dystopian nightmare? www.transformernews.ai/p/the-dod-fi...

Is the increased productivity from AI causing more work for AI users? Is the left taking the right approach on incoming superintelligence? And do British AI researchers be like "anfropic?" Catch up with all the AI discourse with our weekly newsletter. www.transformernews.ai/p/white-hous...

If AI isn't replacing you (and it seems to be replacing a lot of people), it might still be forcing you into an evermore burdensome grind. For all its benefits, instead of alleviating our burnout, is AI becoming the very thing that's burning us out? www.transformernews.ai/p/all-watche...

"We would all hope that AI companies would proceed carefully. However, it appears that caution has been thrown to the wind." In our newest piece, Lord Hunt argues that the UK has to spearhead a pause on the development of the most advanced AI models. www.transformernews.ai/p/why-we-nee...

Glib, clever, semi-ironic dismissals of AI as just a "next-word predictor" or as "not really intelligence" might be stunting the left from actually seeing what AI is, what AI might become, and what the left might actually be able to do about AI. www.transformernews.ai/p/the-left-i...

Alex Bores wants to fix the Democrats' AI problem, with a “willingness to stand up to the political influence that is pushing for nothing to happen”. But what exactly is it going to take to make something happen instead? www.transformernews.ai/p/alex-bores...

On our weekly newsletter, as ever, we've got everything you need to know about what's been happening in the world of AI. For instance: a bunch of people have been leaving AI Labs lately due to quite severe concerns. So should you be severely concerned too? www.transformernews.ai/p/alex-bores...

The left is used to rather glibly dismissing AI as something unlikeable, insignificant, and faddish — but this risks them missing their chance to see it for what it is, and to actually co-ordinate on doing something about it. Our latest piece: www.transformernews.ai/p/the-left-i...

“This ‘let it rip’ approach that is coming from the federal government right now is one that won’t serve us well." Read our interview with House candidate Alex Bores, plus everything you need to know in AI policy, in the latest Transformer Weekly:

Last week the AI frontier jumped forward — twice. But the more advanced AI becomes, the more difficult it is to tell whether novel AI models are dangerous: evaluation techniques simply cannot keep up with the jagged frontier. So we're flying blind. www.transformernews.ai/p/going-in-b...

The political world is unsettlingly chaotic, the AI world is dizzyingly rapid: but if you want to stay on top of how they affect each other, you can always rely on our very calming and well-organised weekly newsletter. www.transformernews.ai/p/going-in-b...

Nvidia is (maybe?) open to investing as much as $100b in OpenAI! SpaceX has acquired xAI, creating a $1.25t company! Google plans to double AI spending this year up to $185b! For these and more inconceivably large numbers, check out our latest newsletter. www.transformernews.ai/p/going-in-b...