Reposted by TimAaron Rupar @atrupar.com · 07/10/2026Troy Jackson after Collins takes credit for rural healthcare cuts in big beautiful bill not being as bad as they could've been even though she voted to advance the bill: "Sen. Collins is like an arsonist who helped set the fire, and then she wants credit for bringing a cup of water to put it out" 242104492419
Tim @teejaykay.bsky.social · 02/10/2026This is pretty extraordinary. Definitely the first time I’ve seen an AI video conversation where the AI looked and sounded real. Usually there’s too much delay in responses and their tone is too flat. (It’s a bit flat here too, but it comes across more as “deadpan” than “fake”) 000
Reposted by TimChristopher Mathias @letsgomathias.bsky.social · 02/10/2026New: Chris Rufo & his co-host devoted a podcast to "The Indian Question," framing Desi people as a problem to be solved. Their "solution?" Mass expulsion. For those allowed to remain: an “assimilation hazing process” w/ instances of “vulgar, defamatory racism" talkingpointsmemo.com/news/chris-r...talkingpointsmemo.comChris Rufo — Known For ‘DEI’ and ‘CRT’ Panics — Tries to Make ‘the Indian Question’ a ThingTo the degree that a “respectable” or “moderate” right exists today in... 1131339483
Tim @teejaykay.bsky.social · 01/10/2026In fact, ALL peas are exactly pea-sized! Glad I could help 000
Reposted by TimZach Weinersmith @zachweinersmith.bsky.social · 28/09/2026WOW, RCP polling average has Trump less popular than he was right after Jan 6 (which, to be clear, was *extremely* unpopular) 1125626
Tim @teejaykay.bsky.social · 25/09/2026Pretty cool! The free space in Claude bingo should’ve been the em dash 😆 020
Tim @teejaykay.bsky.social · 24/09/2026Remember in the 90s when computers got great at chess and the goalposts for “real AI” instantly moved? 4 years ago AI could hardly render an image. Now it does photorealistic video. Bluesky and AI detractors have buried their heads in the sand, and will continue to do so to their own detriment. 001
Tim @teejaykay.bsky.social · 24/09/2026The replies here are variations on “OpenAI hacked them on purpose” and “AI isn’t smart enough to do this on its own”. No and no. OpenAI is too dumb to properly constrain their very smart AI. And yes, “smart”, broadly defined, is accurate. Conscious, no. But incredibly adaptable and capable. 110
Tim @teejaykay.bsky.social · 24/09/2026Sesame Street and Barney. I have several songs from both of them memorized to this day, and one will get stuck in my head at random 010
Reposted by Timweeder @weeder.bsky.social · 16/09/2026Printing something is one of the most depressing things you can do at 6AM 2342545
Tim @teejaykay.bsky.social · 17/09/2026At least Nancy’s teacher will be impressed with her in-depth report on the Q3 numbers! 1120
Tim @teejaykay.bsky.social · 15/09/2026I don’t know who these two people are but I will assume Hee and Haw 000
Tim @teejaykay.bsky.social · 15/09/2026Gus adjusted his tie before he collapsed, he was way more put together 080
Reposted by TimMark Allen Jr. @markallenjr.com · 15/09/2026Imagine the job security of being gone for 3 months without communication to your boss or any indication that you'll be returning any time soon & still having a job. I say imagine because literally no job would let you get away with this. The people of Kentucky should fire him. 59335
Tim @teejaykay.bsky.social · 13/09/2026static.klipy.comHank Scorpio: My Goodness, What An Idea!Alt: Hank Scorpio: My Goodness, What An Idea!why didn’t I think of that 010
Tim @teejaykay.bsky.social · 11/09/2026Man, if I had a nickel for every time Egg Fu the oriental mastermind had a robot twin behind the bamboo curtain who bakes the unique metal men in a concoction of catastrophe 000
Tim @teejaykay.bsky.social · 08/09/2026I was gonna say “world peace” but I’m not sure we want any of the current players trying to tackle that problem 121
Reposted by TimMicroplastics Sommelier @leastactionhero.bsky.social · 03/09/2026i don't agree with their contemptible geography policy, but i do agree with their whatever-is-going-on-here policy 2291
Tim @teejaykay.bsky.social · 31/08/2026I’m just caught up on the fact that they had to specify “one” pair of sunglasses I’d be much more intrigued at “Tom Cruise Defies All Reason in Four Pairs of Aviator Sunglasses” 050
Tim @teejaykay.bsky.social · 19/08/2026I mostly agree! My point is just that it would be awfully strange if a big bundle o’ neurons were the only way to get to consciousness. It seems possible, in principle, to make those same connections in other ways. 000
Tim @teejaykay.bsky.social · 18/08/2026I agree. I do think there are a lot of people on this site who, due to negative effects of LLMs, have been polarized into believing that no AI/machine could ever be conscious. 150
Reposted by TimZach Weinersmith @zachweinersmith.bsky.social · 08/08/2026More actual lifesaving stuff from deepmind, which will probably not get too much coverage outside of nerd circles: deepmind.google/blog/weather...deepmind.googleAI model achieves breakthrough in forecasting cyclonesWeatherNext enables accurate cyclone forecasts that can give an extra day of warning. Now we are open sourcing the model. 99911
Tim @teejaykay.bsky.social · 06/08/2026I don’t think current models will trigger true doomsday scenarios, but this is a good illustration of how advanced AIs will take a poorly specified goal and run wild with it. It’s terrifying, and OpenAI is basically just bragging about it, probably with zero repercussions or real lessons learned. 000
Tim @teejaykay.bsky.social · 02/08/2026And then there’s another letter that’s supposed to be DOUBLE that?? Pssh yeah sure guys, I’m sure it’s right next to the triple-Q 1270
Reposted by TimJon Bois @jonbois.bsky.social · 01/08/2026reply to this post with a time you tried to use a button and it didn't work 1817121
Tim @teejaykay.bsky.social · 01/08/202610 new math discoveries from OpenAI. LLMs are often dismissed as “stochastic parrots.” And yeah, that’s sorta how they work? But with results like this, you have to conclude that a sufficiently advanced stochastic parrot somehow gains some emergent, completely unexpected abilities. 010
Tim @teejaykay.bsky.social · 31/07/2026An English scientist talking about an English playwright is stretching the bounds of “pop culture USA” a bit 😆 000
Tim @teejaykay.bsky.social · 31/07/2026Omg I had this when signing up for physical therapy. I enjoyed Esperanto, Middle Dutch and Ancient Egyptian 040
Tim @teejaykay.bsky.social · 24/07/2026Eh. It’s silly enough to think these AI detector tools are remotely accurate. Now they’re also claiming they can distinguish “AI assisted” writing? 010
Reposted by TimAmanda Litman @amandalitman.bsky.social · 18/07/2026Really really enjoyed this interview with Mamdani - as always, a masterclass in communication. His answer on creating a big tent party is both excellent and low key savage. www.nytimes.com/2026/07/18/m... 181666302
Tim @teejaykay.bsky.social · 16/07/2026Yeah, I’ve had it before in other contexts… not so good as an Oreo it turns out 😆 100
Tim @teejaykay.bsky.social · 16/07/2026I got some Oreos from an Asian market and one kind was like grape peach and the other was red bean paste? Both deeply weird. The creme on the grape peach is like a medicinal grape liqueur, but with just normal Oreo on the outside The bean paste one is sort of… slightly floral and musty 230
Tim @teejaykay.bsky.social · 16/07/2026In “Going Postal” by Terry Pratchett, at a certain point there’s so much smog that it’s a “fug”: “you could have cut cubes out of the air and sold it for cheap building material” 0130
Tim @teejaykay.bsky.social · 13/07/2026Gotta love a mythical half animal half plant creature just showing up to say hi on your trailcam 000
Tim @teejaykay.bsky.social · 13/07/2026Wow and also wow your alt text is just really good writing 250
Tim @teejaykay.bsky.social · 12/07/2026Ah yes, an old southern white man with shitty views who’s been a Senator for decades finally went and died, just as we’d all expe— *puts finger to earpiece* 040
Tim @teejaykay.bsky.social · 12/07/2026Ah yes, an old southern white man with shitty views who’s been a Senator for decades finally went and died, just as we’d all expecte— *puts finger to earpiece* 4532
Reposted by TimZach Weinersmith @zachweinersmith.bsky.social · 11/07/2026Another wild AI math result. Here's the prompt, which interestingly is mostly about method and probably would work pretty well for a human organization: cdn.openai.com/pdf/04d1d1e4... 822624
Tim @teejaykay.bsky.social · 09/07/2026Theoretically, you could maybe codify the other requirements in such a way that it guarantees the thing is good. But a lot of those are pretty abstract and would need orders of magnitude more description. 000
Tim @teejaykay.bsky.social · 09/07/2026Because the rules are (generally) necessary but not sufficient conditions for the thing to be good. 100