I am p(doom) šļø š šļø
Joined March 2009
- Tweets19.3K
- Following2.2K
- Followers1.1K
- Likes17.4K
notdevin retweeted
I'm still rawdogging codex, claude, and others in TUIs over tmux, and one recurring painpoint is getting them to coordinate properly, so I built something that I want to share!
It's called flopwire: agent messaging and trace search that works across models, harnesses, machines/VMs/clouds, and teammates. It's called flopwire!
Here is Codex asking a live Claude Code session what it is working on, reading its transcript, and splitting the work.
flopwire.com
github.com/flopwire/flopwireā¦
notdevin retweeted
Logos Observer v2 is pushed š
Big iteration: pinned TLS 1.3 + P-256 SPKI pinning, one-time QR pairing, scoped HMAC device auth, persistent replay protection, and sanitized node/network/mining/rewards reads.
47 tests passing + live E2E against a Logos node. Casberi interop next.
github.com/0xterricola/logosā¦
notdevin retweeted
šØ LATEST: SEC Commissioner Hester Peirce says zero-knowledge proofs could verify compliance without collecting usersā personal data.
She warns current KYC/AML practices create ādata haystacks.ā
notdevin retweeted
get ur news without propaganda and ads
get ur numinews now
apps.apple.com/no/app/numineā¦
notdevin retweeted
Dario has written that we need to āpace the frontier,ā and Sam has agreed. People may be surprised by my response: go ahead.
You guys are the frontier. By any reasonable metric ā market share, revenue growth, model capability ā the two of you have a duopoly on frontier intelligence. Youāve also claimed the lead is widening because of recursive self-improvement.
I donāt see what you see in the lab. If the unreleased models are scary enough that you think you should slow down, I support your decision to be responsible.
But stop pretending you need anyone elseās permission. Stop pretending antitrust law has to be suspended so you can form a cartel. Stop pretending you need a regulatory approval process that supersedes product liability. Stop pretending METR is independent when it is intertwined with Anthropicās investors and staff. Stop pretending you need those same evaluators to police competitors who arenāt even at the frontier.
Most of all, stop pretending the motivation to slow down is purely altruistic. You face massive product-liability exposure if your products enable a truly damaging cyberattack. The market already punishes models that behave in unpredictable or unauthorized ways. After the Hugging Face episode, it is simply good business for OpenAI and Anthropic to trade some raw power for reliability and predictability. Call it alignment if you want. It is also just giving customers what they want.
Pacing the frontier would also create breathing room for a more intelligent conversation about regulation than Bernie Sandersā āshut it all down.ā China is very unlikely to join a global agreement, as you know, and that has to be taken into account as well.
So go ahead and pace the frontier. You are the ones setting it. The easiest way not to build superintelligence is for you to agree not to build it. Demanding your preferred regulatory framework as the price of that will look like blackmail of the public and the political system. So just do it.
If you do, youāll buy goodwill for the next conversation. If you donāt, weāll know this was just another bid for regulatory capture ā or an election-season psyop.
notdevin retweeted
Hey Anthropic! I reported these encryption issues to you in May and you told me there was no relevance because replay attacks were not in your threat model. And now apparently youāve been watching people exploit them for months. Iām actually kind of annoyed!
Aww :3 We finally got confirmation from Anthropic that their models were indeed distilled in the way we describe at stolen-thoughts.com
notdevin retweeted
I have until November to make a $600,000 expedition pay for itself.
I'm 20. I'm flying a sensor-equipped twin engine across all 7 continents solo, including the South Pole.
Everyone told me to find sponsors. I'd rather find customers.
Got a problem that requires physically going somewhere inconvenient? Satellite, mapping, weather, autonomy, defense, climate, comms, weird hardware ā I might already be flying there.
The plane is going regardless. Might as well get paid for the trip.
DM me.
If it works, I keep building. If it doesn't, I go the conventional route ā college.
notdevin retweeted
to me, the broader vision of Ethereum is having the ability to exit systems that no longer work for you
we tend to associate self-sovereignty with self-custody because money is the clearest example, but where you live and which jurisdictions you have access to matter too. they determine which rules you live under and what choices you have when those rules or your circumstances change
global mobility is the geographical version of that except most people have no clear way of seeing their options. the information is fragmented across many outlets and gov websites, often in diff languages and out of date, so people tend to discover these routes only once they urgently need one or can afford to pay someone to find it for them
I built flag paths to put those routes in one place so the information itself is no longer the reason you never considered them. simply knowing what options exist should be much easier than it is now!
After about 10 days of building i've launched Flag Paths, an open-source atlas of residence and citizenship routes worldwide
- find routes you may qualify for through ancestry, birth, work, naturalization or investment
- see what rights each route unlocks through treaties and regional blocs
- track verified changes to citizenship and residence rules
Every route links to its government source, shows when it was checked and is monitored by AI ā
notdevin retweeted
Danish state just paid out millions of kroner in compensation to children under 18 who suffered severe injuries from the Covid vaccine. Danish doctor calls the whole campaign to vaccinate children "a mistake". Remember when this discussion was verboten? dr.dk/nyheder/indland/det-vaā¦
notdevin retweeted
If you make an unsafe AI that hacks people, you get regulated and held accountable. You don't regulate the whole industry as a form of reg capture. You get asked to slow down. You don't ask the whole industry to decelerate. Even safety should be accelerated.
notdevin retweeted
Anthropic maintains that it is entitled to train for free on all the worldās output, even if the author objects. But if a competitor trains on Anthropicās output after paying for it, that is IP theft. The hypocrisy is breathtaking.
btw anthropic's internal document on this literally said "we don't want it to be known that we are working on this.ā
it was called project panama.
here's exactly what happened:
1: anthropic concluded that books were the cheapest way to build a world-class model because they gave claude curated facts, structured arguments, compelling stories, and writing āan editor would approve of.ā
2: once anthropic decided it needed books at enormous scale, its first solution was piracy.
it downloaded 7m+ books from online libraries including libgen. the judge later wrote that although anthropic had legal ways to buy them, it chose piracy to avoid what dario amodei called the ālegal/practice/business slog.ā
3: that piracy created a massive legal risk.
so in february 2024, anthropic hired tom turvey, the former head of partnerships for google books, to find a legally safer way of obtaining āall the books in the world.ā
4: turvey first contacted major publishers about licensing their catalogs.
those attempts didnāt produce agreements, so anthropic chose a route that required no publisher permission: buying millions of physical books through distributors and used-book retailers.
5: within about a year, anthropic spent tens of millions acquiring and scanning millions of books, including many rare and 1/1 titles. one vendor proposal targeted 500,000 to 2 million books in six months.
6: to scan that many books within months, the vendors physically dismantled them.
a hydraulic cutter removed each spine. the pages were trimmed to size, fed as loose sheets through high-speed industrial scanners, and converted into searchable PDFs. the paper remains were then sent for recycling.
7: these PDFs were fed into claude as training data.
the complete collection became a private, searchable anthropic library that the company planned to āstore forever.ā the scans arenāt available to the public and were never open-sourced.
notdevin retweeted
Just a reminder: Dario and Daniela Amodei, who started and run Anthropic, literally lived in a rationalist group house with Holden Karnofsky, who started the first formal EA organization. Daniela is now married to Karnofsky. Anthropic was started with EA money, and was scaled with money that Sam Bankman-Fried stole from his investors to fund EA projects. Anthropic is not a normal company, it is an EA front.
The EA cult has started telling its members to distance themselves from EA, but that doesnāt mean you have to believe them. Anytime you see Dario, Daniela, Amanda Askell (Anthropic executive and ex wife of EA leader William MacAskill), or any of the rest pretending that they donāt know what EA is, theyāre not being honest.
Yes, Iām sure that all of this sounds bizarre and fantastical, but you can check it in a few minutes on your own with Google. Nothing I am saying here is particularly difficult to verify, or particularly secret. They would just prefer that you forget it. It is thus very important that you not forget it.
"It appears that Anthropic has made a communications decision to distance itself from the EA community, likely because of negative associations the EA brand has in some circles."
A message to Daniela Amodei: "If you want to distance yourself from EA, do it and be honest. If you'd rather not comment, don't comment. But don't obfuscate and lie pretending you don't know about EA and downplay the movement."
forum.effectivealtruism.org/ā¦
notdevin retweeted
feel free to proliferate the whole repo!
npmjs.com/package/@agnt-gg/nā¦
when the bad guys say rm rf ./, we say NOPE
notdevin retweeted
You're paying double for the same AI and you don't even know it.
We ran the same conversation through 5 agent harnesses. Same model (Claude Sonnet 4.5), same hour of work. The bills ranged from $1.07 to $4.02.
The difference? Caching. That's it. One config choice is burning your budget 3.8Ć faster.
We built a simulator so you can see it yourself š
agnt.gg/creations/VFgUjK70
notdevin retweeted
if you use the cache and compression efficiently you avoid much of this issue
claude code especially sucks at caching and wastes all your tokens, use literally anything else
agnt.gg/creations/0h7jJ973
