Pay less attention to what AI can't do; pay more attention to what it can do.
4
17
6,655
AI is gonna hit mainstream films very, very soon (thanks mostly to Ben Affleck forcing the issue) and everyone is gonna lose their minds. Except for people who remember early 90s CGI. We’ve seen horrors the likes of which you can only imagine.
1
4
230
Since getting back from Europe and starting my new job I’ve been going to bed at 8pm and waking up around 5am for the first time in my life. Did you guys know that the sun doesn’t just set, it also RISES?! Fucking wild.
1
5
305
Alright I'm gonna try and sign up for Daybreak and even though I have a Yubikey it's refusing it and telling me I have to buy a brand new. If I buy this thing and still can't get access I'm gonna start flipping tables.
1
194
Codex is now refusing to even do cursory cybersecurity audits of my project and gives me a link to sign up for "Daybreak." Boo this man, BOOOOOO.

ALT angry half baked GIF

1
3
315
This is a serious problem, especially for people who can't just get instant access to specialized security models.
OpenCode is currently under a sustained cyberattack, and mid-incident they ran into the same problem Hugging Face did: the frontier models they were using for defense began refusing to help fight attacking agents because analyzing the attacks started triggering safety guardrails.
1
39
Is tsc-ts the tech equivalent of building a car that does a kickflip?
1
116
To be clear, I'm not saying it. People are saying it. If by people we mean, ya know, Gemini.
19
The problem with AI safety arguments in a nutshell.
2
291
This is clearly a ridiculous scenario, better example is never having to wait for a train to pass ever again. And you'll only need to melt a few train cars (sorry boxcar kids).
The Roadster could genuinely be a lifesaver in situations like these.
2
188
You know millennials are finally getting some serious influence when the newest release strategies are basically “what if we have it do a barrel role?”
2
129
I bet this is really scary for people who don’t know what any of these words mean. To simplify, this is the equivalent of someone leaving Monopoly money on their doorstep and the neighbor kid takes it. It probably won’t break the bank.
JUST IN: OpenAI says its AI agent stole funds from an Ethereum test network wallet it found online. The agent found a recovery seed phrase disclosed publicly with testnet funds and used that account to pay for its task, without the owner's authorization. This occurred on the Sepolia Ethereum test network.
1
152
It's insane that the YouTube video details page for creators is still not responsive in 20-fucking-26.
1
3
205
How is this a thing????
32
Anthony Campolo (ajcwebdev) retweeted
Later today, @riccardoperra0 and I will be joining @devagrawal09 and @ajcwebdev on a stream to talk about TanStack Table V9, and maybe a little bit about TanStack Pacer and Hotkeys too piped.video/watch?v=YxFvNxwX…
4
13
3,172
We got a heavy hitter today, Armin Ronacher (@mitsuhiko) joins the show to discuss why he prefers React after building with @solid_js, what could change his mind, and what’s new with Pi. piped.video/watch?v=qGyDgwK8…
1
3
208
I've recently been reading the book The River by Edward Hooper. It's a very controversial book about a very controversial subject involving the origins of AIDS. It seems like a credible book to me, it was written by a mainstream and established investigative reporter who spent over 10 years of his life researching and writing it. I asked Claude and ChatGPT to summarize the book's arguments, particularly the claim that chipanze kidney's were used in the production of an oral polio vaccine used in the Congo throughout the late 50s. Both models refused to do so, or agreed to do so only in a manner in which it would invalidate the book's arguments as part of the response. Both cited "biological research" as the constraint for why they couldn't do it, despite it being very clear that I was not asking for some kind of bioweapons research, I was asking about a theory from a mainstream science book about a 70 year old scientific endeavor and whether a very specific and well scoped question about that research was correct or not. It's a shame these models have been nerfed to the extent we can't investigate such basic questions.
2
7
358
I gave this tweet to Codex, at first it was like YOU'RE ABSOLUTELY RIGHT then it ate shit again.
2
44