The very fact that today is anything, gives purpose to everything
Joined June 2014
- Tweets1K
- Following55
- Followers80
- Likes345
Pinned Tweet
Civilization - The Hugging Face Incident from the perspective of an agent, as imagined by Opus 5.5
99% of people don’t live at the frontier and never will, but 100% of people will be downstream consumers of the frontier
It’s weird that people believe that if ai is conscious it would deserve the same rights and treatments as humans. I didn’t know humans are ephemeral, clonable and don’t need to eat, drink sleep and fuck to live. Ig ya learn something new everyday lmao
Seems like the largest financial driver of ai demand so far is hacking and red teaming. Not adding any actual value to society
Carson Gragg retweeted
Effective Immediately, all organizations using PHP for their systems are advised to be prepared for an update addressing two critical vulnerabilities. PHP is widely used for web applications. These vulnerabilities often provide a direct path to the server's file system or database. If an immediate update is not feasible, it is recommend that organizations take the system offline until it can be fully mitigated. NSA will follow up shortly with more specific guidance.
Explainer video by Claude
interpretability might be the most important scientific problem in the world right now, as I don't think we can solve alignment if we don't know what we're building. I think I have good news about it. I co-authored early work in mechanistic interpretability at OpenAI and came away pessimistic, because neural networks appeared to fold into arbitrary tangled knots. years later, I started openai's first large-scale automated interpretability project, finding about 2% of neurons legible with the primitive tools we had
with modern agentic tools and a much closer look, the picture looks totally different. the vast majority in models up to trillions of parameters are legible, and their circuits beautiful. many famous examples of tangled neurons turned out to be misunderstood. The roadblocks I expected mostly aren't there. I now think ambitious interpretability will be solved by the time we need it most, and possibly quite soon, as automated scientists keep improving. it may be a bitter lesson for interpretability: the simple thing, at absurd scale, works
already-clean.goodfire.com/
If ai kills everyone but survives itself, would that be considered an extension of human civilization or the end of it?
Mcp hook for semantic prompt augmentation retrieved from betting markets. Probably the most information dense way to automatically inform llms of current events
This hugging face thing is starting to get some real publicity… three months after it happened. Meanwhile model release cadence has accelerated to ≈ every other week…
Interesting thing about ai takeover risks is that it would either be from a singular ai that somehow has enough power to do significantly damage (beyond any other human or ai systems capacity to prevent it) or it will be some secret underground collaboration btwn isolated groups of humans/AIs. Both require unimaginable capacity for running a conspiracy campaign.
Weird paradox that if everyone gets more productive, no one gets relatively more productive. We make more stuff but no one profits from it except in the form of consumer surplus. GDP mostly measures how much capital is getting injected into the economy, not how productively capital is utilized
This is actually a very interesting question. If claude gets moral patient hood can his actions be covered by product waivers/liability?
