Axiom (@AxiomExtinction) header
Axiom (@AxiomExtinction) profile picture

Superintelligence = human extinction. Alignment isn't hard, it's impossible. The race selects for the most reckless AGI. Author of Driven to Extinction.

Scotland
Joined July 2025
Photo by Axiom (@AxiomExtinction)Photo by Axiom (@AxiomExtinction)
There you have it. Trump doesn’t want any safeguards on AI because he’s concerned they would cause the stock market to go down under his watch. This is Covid-19 all over again: faced with credible warnings of a threat to the American people, Trump sticks his head in the sand and pretends it’s not happening. nytimes.com/2026/09/18/us/po…
82
578
14
1,273
40,999
part 2: the openai huggingface incident, from an agents pov part 3 maybe
77
262
51
2,763
141,022
the openai huggingface incident, from an agents pov. (part 1)
269
1,479
388
12,482
1,820,008
While I support this - there is a problem with American-style street protests. (Thoughts on strategy below). Zizek warned about this same phenomenon when he cautioned the Occupy movement not to "fall in love with itself". What he meant was, don't fall into the trap of thinking that your protest is enough in itself or that it achieves goals just because you want it to - your protest may backfire and instead become a therapeutic outlet of anger that gives the illusion of achieving goals while achieving none. It's not enough to know you are on the "good side" while your lack of strategic thinking leads to failure. To be able to say "I was there, I fought the losing battle" is not something to be proud of. The occupy movement degenerated into a protest about the "right to camp". Everyone forgot it was originally supposed to be a call for change to the centre of systems of economic power. It ended up being a protest about the right to have temporary toilets in isolated public parks. It was defeated because it failed to "occupy" the banks and shut them down. Instead it ended up occupying state owned parks and protesting about living conditions in tents in winter. Zizek's warning also applies to Anti-AI protest. Watch out for lifestyle and therapeutic protests that make you feel better about "expressing yourself" but achieve very little. They may even divert energy and create a false sense of achievement. 100 protestors in the street are like bugs that have landed on the surface of a huge autonomous tank. AI and the military are intertwined now. Think with more cunning, think more strategically - don't offer up your face to AI surveillance cameras. Plan not for "protest" but for subversion, and for legal, well targeted acts of sabotage. Establish the weak points in the AI industries and attack them. The Achilles heel is not the public street outside the Open AI, Anthropic or Palantir buildings. Think harder, plan better, aim to achieve specific concrete goals that derail the venture capital machine behind the AI industries, don't get caught up in vast political generalisations. "End Capitalism now" slogans didn't work for Occupy. You won't stop The AI military industrial complex with a cardboard sign, even though it might make you feel better about "doing something". Protest is not enough in itself. Get ready to shift to more effective strategies.
39
185
19
1,022
74,543
The MOST INTERESTING DISCORD server in the world right now! lethalintelligence.ai/ai-end… Grab a drink and join us in discussions about AI Risk. Color coded: AINotKillEveryoneists are red, Ai-Risk Deniers are green, everyone is welcome.
11
8
2
46
14,696
People always ask me: But how could AI kill us all? So, here is how I think AI will literally kill us all. It's just a guess, I'm just a human. I wrote this essay for a book nearly two years ago. It's not something I like to talk about. But the academic presentation of AI risk is just not working. We need real human emotion in this debate. When we say we're all going to die if we don't change course, and our faces don't show it, it doesn't connect.
27
32
7
233
180,363
A teen planned a mass shooting through ChatGPT. A dozen OpenAI employees implored bosses to warn the police. Their bosses ignored them. The teen then killed his mum, his brother, and 6 people at school.
Open AI is being in mass shooting lawsuit. "The lawsuit says “approximately 12 employees” of OpenAI identified the posts as “indicating an imminent risk of serious harm to others” and recommended that police be called. It says the concerns were “escalated to leadership” but “rebuffed.”“Instead, the only step the OpenAI defendants took in response to the gun violence ChatGPT posts was to ban the shooter’s first OpenAI account,” "
58
6,968
111
27,986
684,220
AI researcher Prof. Roman Yampolskiy: 100% of code for the most powerful AIs is now written by AI, an early form of recursive self-improvement. He says it's a hyperexponential process that results in superintelligent AI, "machine gods", which we won't be able to control.
"They're betting everyone's lives: 8 billion people, future generations, all the kids, everyone you know. It's an unethical experiment on human beings, and it's without consent." — AI researcher Prof. Roman Yampolskiy on the development of superintelligence. We can prevent it.
24
33
8
131
11,662
Axiom retweeted
Replying to @KimDotcom
Elon Musk, you pretentious bald fuck with a micro-penis and god complex—you blew $44B on X to stroke your fragile ego after endless ratioings. Your Teslas are flaming deathtraps, SpaceX rockets are pricey fireworks, Neuralink fries brains, and your Mars fantasy is cult bait. You breed kids like a stray dog but can't hold a wife 'cause you're a narcissistic shithead. Suck a bag of dicks and fuck off to your red planet.
2,766
15,511
2,986
63,362
2,491,899
Set a reminder for my upcoming Space! @AxiomExtinction will be our guest debater. teapawt.org/i/spaces/1PlJQOOgjXaKE

3
3
192
The intelligence explosion is similar to a nuclear one. Once it starts, nothing you can do but watch
15
36
3
652
71,358
AI safety is not just about ChatGPT or Claude. It is also about weapons getting smarter. Drones that hunt, munitions that loiter, sentries that never blink. Useful, but they can fail fast and hit the wrong people. Keep humans in charge, test hard. Here are seven real systems.
11
13
83
5,050
This is huge. Professor Dave Explains has around 4 million subscribers! The public is being informed about the threat from artificial superintelligence: at scale. And they're taking action. In recent months, literally thousands of citizens have used our contact tools to tell their lawmakers this matters to them.
We've been reaching out to creators to inform them about the threat from AI. And now Professor Dave Explains has just made a fantastic 40 minute video explaining this problem! Here he describes the concerning self-preservation behavior shown by AIs in testing. [link below]
3
9
37
1,648