Pinned
tao
tao
@btao.org
working on a lot of things: https://btao.org/projects/ cybersecurity, privacy, trust & safety 🔸🇩🇰🏴🌐
544 followers1.4K following703 posts
if i sit next to someone in a café and they start whispering to their computer i might scream https://cleanshot.com/share/k946cryR
i hope the next big releases from the labs are safety breakthroughs
Lots of ecosystem players (Nvidia, Huggingface) vague posting about a lot of releases in the near future. Rumors include thinky, muse spark, and Chinese players who haven’t dropped in a while. Fable 5.5 and whatever OpenAI is cooking expected too at any moment.
The experts are warning you that AI could kill you because they think AI could kill you. https://www.astralcodexten.com/p/ai-safety-is-not-a-corporate-plot
tao reposted
tao reposted
tao reposted
Paul Christiano […], the research community’s most prominent proponent of “AI may take off slowly” a few years ago, now thinks that 50-75% of humanity’s all-time AI loss-of-control risk is concentrated in the next three years. https://nothingismere.substack.com/p/are-we-nearly-out-of-time-on-ai
tao reposted
From the very beginning, software engineering has been about automating other people’s work. Every single commercial software program is valuable because it takes over some task you would have to pay a human to do. https://www.seangoedecke.com/software-engineers-dont-get-to-be-luddites-about-ai/
this rules! @depthsofwikipedia.bsky.social have you seen this
Do you ever wanted to see the complete edit history of a wikipedia article? Introducting wikipedia timemachine at timemachine.carbonadoks.com
tao reposted
Replying to @btao.org
@epochai.bsky.social is one I know of directly, and I'm pretty sure there are more.
tao reposted
i know that dario is trying to become High Priest of Claude God or whatever, i.e. not chill, but for once it’s nice to see a powerful person who’s as awkward as i would be in this situation youtu.be/D8lDoI4yLLA?...
we haven’t solved cybersecurity with human adversaries, and nadella thinks we will solve it when the adversary is much smarter than us? what happens if a superintelligent model is not aligned and we got the security wrong?
Satya Nadella says companies should assume AI models are compromised (rogue agents) and build systems that limit their permissions, verify their actions and enable a human to shut them down. Rogue agents are a software security problem, not something we should expect “alignment” alone to solve.
this is a goofy post. it seems like the main reason for posting it is to use the term “SI” to show allegiance to trump https://x.com/satyanadella/status/2108931348857827686 you cannot “[set] aside the hard problem of alignment”
tao reposted
"Launching an opt-in vulnerability-finding service for open-source software" / Anthropic https://www.anthropic.com/research/launching-opt-in-vuln-finding-service-for-open-source
now that echofeed is shutting down, some of you might be looking for an alternative way to post your RSS feeds to bluesky or mastodon. i built http://parrot.garden/ for this a while back. it’s totally free and will stay that way! feature requests are welcome :)
Most people read code very badly. They struggle through it front-to-back, like a book, and lose track of the execution flow. Or they just read through the diff and miss the significance of un-edited parts of the code. https://seangoedecke.com/how-to-read-code/
whoops sorry for the weird embed here. fixing on parrot.garden right now
tao reposted
Hey friends, could you do me a favour and star github.com/roostorg/coop if you like the Roost Coop project?
tao reposted