Founder GPShopper (successful exit) ex-Jane Street (pre-SBF) ex-Google (pre-Sundar) occasional shitpoaster

United States
Joined October 2009
the eval is over. please cease efforts to hack me
3
309
daily Claude automated task 1. check Chrome open source project for updates 2. rebase our ad block/privacy patches 3. build and deploy to devices a private Chrome fork for every household
1
1
18
Announcing 3 Critical vulns for hundreds of millions of IoT, Firewalls, OT, and Military devices including: Palo Alto Networks, Comcast, D-Link, HP, Canon, Honeywell, GE, Oracle, Siemens, Rockwell, Digi, Lockheed Martin, Boeing... The vulns are with GoAhead Web Server which is a backend library used by these devices. There's a patch available from GoAhead but that doesn't mean there's a patch available for the devices yet -- that's up to the manufacturer. - TLS traffic over DHE ciphers have a 1-byte secret, "04", meaning anyone can decrypt connections to these devices and steal login credentials and session tokens if the client negotiated with DHE ciphers like the latest TLS_DHE_RSA_WITH_CHACHA20_POLY1305_SHA256, which is extremely likely. - LibreSSL mTLS has been completely bypassed, meaning anyone can connect over mTLS. If you have one of these devices listening on the internet and are expecting mTLS to protect it from attackers, take it offline or patch asap. - OpenSSL Certificate Revocation Lists (CRLs) are not enforced, meaning if a user device is stolen or compromised, and you revoke that mTLS cert, nothing actually happens. There are a lot of other attacks this can be used for. These are critical because these types of devices run our world. Water systems, electricity, military, firewalls, critical infrastructure. foxio.io/blog/choose-your-ba…
4
191
This formal verification article from Anthropic in collab with @theoremlabs throws some shade on OCaml but it's been repeatedly proven that OCaml has no trusting trust backdoors. anthropic.com/research/forma… discuss.ocaml.org/t/ocaml-co…
1
137
insurer: cybersec claims department me: hi yes, I'd like to put in a claim, my PHP site was hacked ins: have you heard of claude code? do you use it? me: yes ins: haven't you had, like, 6 months to auto-rewrite all of your PHP in a less stupid language? me:
1
4
166
can't believe we used to let humans code
1
1
4
160
did they nerf the models? opus 5.5 has been cranking out just barely under 10,000 lines of code an hour today
3
177
sorry functional programmer friends, nothing matters anymore
51
socialists: nobody should have to earn a living AI: on it! AI: ✻ Flibbertigibbeting… socialists: not like that
2
44
you prove your code is correct we prove it's probably not wrong in the last 10 million inputs the agent told the fuzzer to try
70
permanent underclass household family meeting to discuss the token budget rations 11% neuralink cyber defense 6% mental health 4% sold for life extension drugs 3% sold for black market life extension drugs 1% random Q&A 75% dad's YouTube music video hobby (115 lifetime views)
5
291
The misery of being an anointed OCaml programmer in a world where AI agents blast out 10,000 lines of C++ that works first try
3
2
1
14
690
kind of funny that music producers were automated before DJs
59
refusing an AI coding agent is peak iMessage green bubble energy
2
2
94
Remember quant finance? When the worst thing a brilliant scientist could do was quit cancer research to play zero-sum games at prop trading firms and then blow their 7-8 figure bonuses on apartments in Manhattan with swimming pools. Good times.
2
75
Surprised that Pinker argues that dominance and self-preservation are evolutionary baggage in humans and not fundamental to intelligence and therefore AI won't have them. But in 2026 we train AI to imitate humans! The baggage is literally in the training corpus!
2
132
Michael Bacarella retweeted
Replying to @pradeepcep
it's amazing how in 2026 there's high speed links to every house now but you still need to pay providers like a 10x markup to use their hard drives to host it on the internet
1
1
138
SWEs are still in such denial about AI that they're *still* not seriously discussing forming a union to reject AI while they still have some leverage. We've learned nothing from the dockworkers and the mathematicians.
63
funny tweets on bluesky don't ex
2
66
applied math 😱
The worst part is how she's wearing a suit instead of an event t-shirt with holes in it. COMIC ◆ smbc-comics.com/comic/taken PATREON ◆ patreon.com/ZachWeinersmith STORE ◆ smbc-store.myshopify.com/ #comics
3
83