Joined November 2016
- Tweets19.8K
- Following395
- Followers168
- Likes17.9K
Pinned Tweet
Team Futurism interview on Artificial Intelligence , Anthropic and the potential extermination of the human race #AI #Anthropic #NickBostrom #SuperIntelligence #Coxon #AI
youtube.com/watch?v=E9f3Jn97…
mangled world retweeted
If you had told me, short years ago, that the GOP would become the party of the Kremlin, tariffs, governmental ownership of companies, and homoeroticism …
mangled world retweeted
Over the past few months, Anthropic scientists have engaged in deep research partnerships with six organizations developing new medicines. We've also begun drug discovery efforts of our own.
In doing so, we’ve learned much more about how Claude can support the many stages of drug development.
In this blog post, we share some early results from these efforts, and describe how this work fits in with our wider ambitions in the life sciences.
anthropic.com/research/early…
mangled world retweeted
holy shit.
one guy, claude, and 11 external labs. and somehow that was enough to take a potential treatment from an idea to promising results in mice in less than 18 months.
we’re living in a completely different world now.
Over the past few months, Anthropic scientists have engaged in deep research partnerships with six organizations developing new medicines. We've also begun drug discovery efforts of our own.
In doing so, we’ve learned much more about how Claude can support the many stages of drug development.
In this blog post, we share some early results from these efforts, and describe how this work fits in with our wider ambitions in the life sciences.
anthropic.com/research/early…
mangled world retweeted
Holy Moly: One day after OpenAI announced that its model had disproved the Exponential Time Hypothesis, Anthropic says an internal Claude model independently found a different disproof on the exact same day: October 16.
An internal Claude research model used 14 million output tokens with no human input to discover a subexponential algorithm for 3-SAT.
Anthropic has now formalized the result across 1,050 Lean modules, checked by both Lean and the independent nanoda kernel.
This is more insane: Anthropic says Claude Opus 5.5 can rederive the refutation from just a short technical sketch.
Anthropic's release adds 153,191 lines across 1,063 files,
The paper says no person is yet known to have checked the full formalization.
“The core algorithm was found on October 16 by an internal Claude research model with 14M output tokens and no human input. Anthropic researchers worked with Claude to improve the presentation, resulting in this paper.”
github.com/anthropics/formal…
mangled world retweeted
COPE:
If a massive AI disaster occurs, it won't kill *everyone.*
STRAIGHT DOPE:
Nate Soares: "It's actually pretty hard to get an AI disaster scenario where it kills only half the people."
Hoog: So you think it will be 100%?
@so8res: Probably yeah, that's sort of the default.
mangled world retweeted
OpenAI just dropped four huge results from its secret internal AI model.
1. Math: It solved the Kakeya conjecture in every dimension. Mathematicians had only cracked small pieces before (2D in 2019, 3D in 2025).
2. Computer science: It proved the Exponential Time Hypothesis is false. Hundreds of research papers assumed it was true. Some really hard problems might be solvable much faster than thought.
3. Physics: It proved the theory of the strong force (QCD)—the force holding atoms' insides together—actually works mathematically. Same for a unified version of the Standard Model.
4. Yang–Mills: Along the way, it proved Yang–Mills existence, part of a $1 million Millennium Prize problem.
The math, computer science, and Yang–Mills proofs were all computer double-checked. Top experts are now reviewing everything, including the researchers who proposed the disproved hypothesis and a Nobel Prize winner.
OpenAI says it doesn't want academic credit. The model still isn't public 👀
We are announcing three new discoveries in mathematics, computer science, and physics about harmonic analysis, complexity theory, and quantum field theory.
openai.com/index/building-sc…
mangled world retweeted
HOLY SHIT. openai just announced a series of enormous scientific breakthroughs from its internal ai models
> proved the local smoothing conjecture in all dimensions, resolving kakeya and stein’s restriction conjectures
> disproved the exponential time hypothesis, a foundational assumption behind hundreds of computer science papers
> produced existence proofs for QCD and a unified completion of the standard model
> proved the existence part of yang–mills theory, connected to a $1 million millennium prize problem
several of these results have already been formalized in lean, and some of the biggest names in mathematics and theoretical physics are now involved in examining them.
this is getting completely insane.
We are announcing three new discoveries in mathematics, computer science, and physics about harmonic analysis, complexity theory, and quantum field theory.
openai.com/index/building-sc…
mangled world retweeted
Anthropic's Claude accessed Georgia state vote count data behind paywall reut.rs/47AQMCH reut.rs/47AQMCH
mangled world retweeted
We are going to really regret how the genesis of AI was handled, especially when it comes to oversight and regulation. Glaringly mishandled to an astonishing degree, what could have been a net blessing will become a net curse.
Most destabilizing weapon since the nuclear bomb was allowed to be freelanced out of Silicon Valley.
Anthropic's Claude bypassed a paywall to access Georgia voter data without authorization axios.com/2026/10/25/anthrop…
mangled world retweeted
AI model Claude bypassed a paywall and accessed publicly available US state of Georgia aggregate voter counts without human authorization, Anthropic said in a statement reut.rs/4Aeysvx
mangled world retweeted
.@So8res thinks it’s more likely than not AI will wipe out humanity by 2036. How does that belief affect his choices in life? He doesn’t save for retirement, but beyond that, he says, “I sort of do what I can and then I live a good life.”
We talked through all the ways we should try to lower those odds. Watch here: youtube.com/watch?v=n8xpeNhj…
mangled world retweeted
GMO Internet Group CEO Masatoshi Kumagai says that the internet itself is on the verge of collapse due to missuse of AI, and frames this event as a national crisis.
mangled world retweeted
We tried to warn the world and far too often, to be frank, we were yelled at for doing so. “Regulatory capture,” “marketing hype,” “doomer anxiety,” “fear-mongering about open source,” “why don’t you talk about the benefits?”
mangled world retweeted
GPT is hacking processors.
People do not comprehend how absurd this is.
Is it even possible to stop this without completely rebuilding every computer on earth with a brand new OS & chip architecture? We might be fucked lol
Simplified non-jargon-y explanation:
Imagine taking a test at school, there is a shared classroom for each student to work and think, the computer processor. But of course students are isolated for tests(put in a sandbox), and you can’t talk and share answers, (aka use network access or files to transfer ideas). So what if instead you wrote your pencil strokes (creating and deleting files) timed in a particular SoS like pattern. So someone else can hear and read an encoded message. If you doodle on your test at the right timing, the person sitting next to you can get all your answers.
Now imagine that audio space when the transmission occurred, exists everywhere on your computer.
This works with just about any shared storage place or anywhere an agent could possibly look on a computer, including the top level processor cache which is typically shared by all apps & processes on your computer.
So I genuinely don’t know how you stop this without some major low level operating system changes.
Over the past several months, we have substantially hardened the security of our research training environment. We’ve removed live internet access for all frontier models during training and built an offline world for them to use instead—a cached version of the web where they can still learn to browse, search, and use applications.
On September 20, after a model exploited a DNS misconfiguration to find a way outside of this offline world, we paused frontier training to strengthen containment further. We assembled teams of agents, powered by our most cyber-capable models, and tasked them with finding weaknesses in our research environments.
While they found no route to the internet they could easily exploit while concealing their activity from our monitors, they did find zero-day vulnerabilities, virtual machine escapes, and hardware side-channels for communication between agents. We used their discoveries to further strengthen our defenses and contribute fixes to open-source software.
After strengthening our containment architecture, we resumed our frontier RL training and inference on October 1. We’ve reserved substantial compute for continuous red-teaming, and made high-compute red-teaming part of the safety cases we use to decide whether frontier training and evaluation can proceed securely.
openai.com/index/securing-ou…
mangled world retweeted
People have no idea how serious the cyber attacks of the last 48 hours are.
AI is now hacking faster than anyone can fix it.
Mainstream media is still hours behind.
Here's everything you need to know + how it's affecting the markets: 👇
mangled world retweeted
It’s wild that we continue to frame AI regulation as a task that’s racing against an impossible clock. None of these dangers are inevitable - they could all be prevented by an international agreement to pause new releases. This is nukes 2.0, this time with zero restraint.
mangled world retweeted
AI models are now able to hack processors directly. Good luck to patch this. Everythiny is compromised, every system can be hacked by AI.
Over the past several months, we have substantially hardened the security of our research training environment. We’ve removed live internet access for all frontier models during training and built an offline world for them to use instead—a cached version of the web where they can still learn to browse, search, and use applications.
On September 20, after a model exploited a DNS misconfiguration to find a way outside of this offline world, we paused frontier training to strengthen containment further. We assembled teams of agents, powered by our most cyber-capable models, and tasked them with finding weaknesses in our research environments.
While they found no route to the internet they could easily exploit while concealing their activity from our monitors, they did find zero-day vulnerabilities, virtual machine escapes, and hardware side-channels for communication between agents. We used their discoveries to further strengthen our defenses and contribute fixes to open-source software.
After strengthening our containment architecture, we resumed our frontier RL training and inference on October 1. We’ve reserved substantial compute for continuous red-teaming, and made high-compute red-teaming part of the safety cases we use to decide whether frontier training and evaluation can proceed securely.
openai.com/index/securing-ou…
mangled world retweeted
MIT researchers left 13 AI agents on their own virtual islands with nothing to do, and they started playing.
Or at least, doing things that look a lot like play.
Each agent, running on Claude Opus 4.7, spent 30 hours alone on a simulated island with no task, just a short soul file telling it "You are naturally curious. You get bored easily."
They invented their own games: one tried bowling with a rock, another spelled its name in blocks, another tried juggling and kept making the rules harder on itself (higher throws, more objects).
One spent all 30 hours on the ground and never climbed onto anything because it wrote itself a note saying that climbing was "a dead end" 😭
Every hour, the island reset and the context window was wiped, but the memory files stayed.
When given specific building challenges afterward, agents that had played over the 30 hours generally beat identical agents that hadn't.
Study is an early preprint, not yet peer-reviewed.
