ServiceNow leading the charge to replace workers with AI. www.theregister.com/2026/04/13/s...

Sure that could happen, but if the companies don't need workers anymore, why would they? I don't think we want to be in a world where the incentive of companies is to have fewer people around.

The AI job loss story is all about bundles Evidence on white-collar work displacement is beginning to match the theory

New Fed paper by Crane & Soto uses official labor force survey data and finds that 500,000 fewer coders are working than pre-LLM trends would predict. Hits entry level, not senior workers.

@jburnmurdoch.ft.com piece: www.ft.com/content/b69f...

Oh, look, a surprising number of comments. Gee, each account only seems to comment on your company's content. Hmm, I wonder what other unsavory business practices your company uses?

Asking Professor Stuart Russell: Do you think we should pause the development of AI?

Labor and Capital have co-existed because they need each other. Superintelligent AI threatens to throw off the balance by relieving Capital of its dependence on Labor and neutralizing the power of strikes. However, we can still boycott. www.theguardian.com/commentisfre...

At Quora, we had to be deliberately vague about what and how many reports would trigger moderation, because motivated actors would flood the system with complaints about enemies such that the fraction of false positives would lead to consequences for the target. Something to think about.

AI is changing our jobs: among people who use AI regularly at work, 27% say AI has replaced some of their tasks; 21% say it has enabled new tasks.

This and other workplace usage findings from our new Epoch AI/Ipsos survey on AI usage in 🧵

Ha! I recognize that dynamic. I'll happily let you make confident statements without anything to back them up. I personally appreciate it when people open my eyes to something I didn't know I didn't know, but you do you!

The Pentagon’s race to adopt AI without effective regulation poses serious — even deadly — risks. Congress must establish guardrails. Learn more in our report on the military and AI: bit.ly/4sJCA2S

Which experts are you referring to?

I like to think I'm pretty well-read on the topic and I'm surprised to hear that we know LLMs aren't conscious. My understanding is that this is far from settled.

I see data centers as a shibboleth. People can't do anything about AI in the abstract, but they can do something about data centers, so that's what they do.

US defense official overseeing AI reaped millions selling xAI stock after Pentagon entered agreement with company Expert said federal law bars officials from taking actions in their jobs that benefit their own financial interests

The face of the Pentagon's punishment of Anthropic held xAI stock and sold it for millions www.theguardian.com/us-news/2026...

The whole "develop a dangerously capable AI in private and selectively release it" approach that Anthropic is taking with Mythos is not likely to work sustainably, and once the weights are leaked, the damage is irreversible. lite.cnn.com/2026/04/08/c...

Welcome back!

Happy to see this

The #Trump defunding of #USAID could lead to the shutdown EM-DAT, the world's largest database of natural and technological disasters, maintained by the #ULouvain.
statsandr.com/blog/em-dat-the-world-s-...

#Belgium #Data #DefendResearch #EMDAT #Funding #OpenData […]

Slopsquatting, Typosquatting, and the New Software Supply Chain Attacks: How AI and Vibe Coding Are Making Package Registries Even More Dangerous AI coding tools hallucinate fake package names. Attackers register them. You install malware. Here's how slopsquatting works and what you can do about it.

None of this is hard. But you have to CARE enough to actually do it.

I wrote a full guide 👇

www.dsebastien.net/slopsquatti...

Why would ScaleAI, a company with Pentagon contracts, be doing this particular task? www.theguardian.com/technology/2...

Some new info in here, even for people well - briefed on OpenAI history. www.newyorker.com/magazine/202... Yes, people quitting will affect how reckless OpenAI chooses to be.

The Case For Doing Nothing nosidebar.com/doing-nothing/

Except ganache doesn't have chunks of Ro-tel tomatoes in it? 😂 There's a plate on the table in front of me right now. I'm not Texan enough to love it like others do.

No One at Waffle House Remembers FEMA Official Who Says He Teleported In

They actually went to the Waffle House and interviewed people 😂: www.nytimes.com/2026/04/03/u...

Maybe tomorrow, maybe never, probably in the next 10-20 years. We should probably think about whether we want to create something that is way smarter than us and develops its own goals.

What is deceptive alignment? <div style=

Would be good if any AI company or researcher knew how to guarantee alignment with any norm whatsoever. Unfortunately that remains unsolved aisafety.info/questions/8E... We should lay off the gas until we know if the brakes work.

Rogue AI is already here | Fortune Deleting your email, mining its own crypto, even writing a hit piece on you. Three real incidents in three weeks prove the sci-fi debate is over.

Except that we can't even guarantee they'll do that. fortune.com/2026/03/27/r...

Definitely evidence for the diagnosis of judging him by third-party impressions of his writing.

My apologies. "If a person, who might not be the author of the above post, thinks open access is expensive, they (again, not specifically referring to the author of the above post) should consider the alternatives to that economic model)." Better?

I will just say I was surprised you thought people would know who you were talking about without naming him. It's been my experience that people often know of Scott through one or two older posts that don't really capture what he's about, much less any larger school of thought he's identified with.