crash test smarty
contxt.bsky.social
crash test smarty
@contxt.bsky.social
Hillary Clinton?
December 9, 2025 at 4:21 AM
Does it matter whether they are the "worst of the worst"? We know the American criminal justice system has unequal outcomes based on race. Every one of these people at least deserves due process, but we don't need to accept the judgement of a racist system either.
November 25, 2025 at 8:36 PM
wagon wrist, even
November 7, 2025 at 3:53 AM
SNAP doesn't cover sanitary products like tampons, pads, etc but those are often in high demand/short supply at food banks.
November 4, 2025 at 8:29 PM
this but the dial says "killing trans people"
November 2, 2025 at 5:02 PM
Sadly Mr Dees passed in 2020 after a long fight with ligma
September 29, 2025 at 4:52 PM
It's 2002 and people in my small town are campaigning *against* a new Walmart because it'll hurt small businesses! The nostalgia isn't just ahistorical, it's incompatible with their blood-soil-bootstraps white supremacist vision of rural America
September 29, 2025 at 4:50 PM
This has the aesthetic of a David Dees piece lmao
September 29, 2025 at 4:42 PM
Going to the consulate in the overseas concentration camp and requesting a new passport
April 17, 2025 at 3:14 AM
Somebody shot all my neighbors with a Ballardization ray, turning them into flesh-eating renters
April 17, 2025 at 2:44 AM
These people are so rich they can afford reading tutors for their dogs, so that tracks
April 4, 2025 at 1:56 AM
It's funny to see a guy use a chess metaphor (pieces, positional gain) who has one of the all-time dumbest takes on chess
April 2, 2025 at 3:34 PM
The rub is that if you're building explicit guardrail systems for a sensitive system, then why not just use those guardrails to build non-AI software? Making decisions about complex, ambiguous systems is and always has been the hardest part of automation.
March 31, 2025 at 1:50 AM
The answer, I suppose, is that these people want to hook up their AI systems to power stations, hospital systems, etc. But the real danger isn't that the model is too intelligent, it's that it isn't smart enough.
March 31, 2025 at 1:46 AM
Same for the assertion of AI's cataclysmic danger. Current models "don’t go beyond what can be found in textbooks or on search engines". What would a more dangerous model do, exactly?
March 31, 2025 at 1:45 AM
As usual the questions is never asked or answered here: is there any technical evidence for the approach of "powerful AI"? These CEOs get so much free marketing
March 31, 2025 at 12:46 AM
I think most of us rely on experts to communicate the details of, say, inflation, mRNA vaccines, effects of migrant workers on labor markets. But there's a total partisan divide on who is an expert and when we can trust them. Idk how long that's been the case in the US
March 17, 2025 at 3:08 PM
Might be a reflection of critical governmental, economic, and social systems being too complicated for almost anyone but experts to understand
March 17, 2025 at 3:02 PM