Imma be real with you guys the "TESCREAL" stuff is basically the libleft version of rightist paranoia over "Cultural Marxism", where you just smash together a half-dozen things that annoy you and go "these things are all connected and being pushed by a nefarious secret cabal of my political enemies"
30.09.2025 00:59 โ ๐ 372 ๐ 47 ๐ฌ 30 ๐ 15
Alternative theses: bsky.app/profile/dyst...
30.01.2026 07:22 โ ๐ 1 ๐ 0 ๐ฌ 0 ๐ 0
(not meaning to, idk, police you or anything, just working through this out loud)
30.01.2026 06:17 โ ๐ 2 ๐ 0 ๐ฌ 1 ๐ 0
Lol yes that does seem to be what happens. I guess I was interested to see what looked like a number of concessions on technicalities in the quoted thread, and I don't think even those might be warranted!
30.01.2026 06:16 โ ๐ 4 ๐ 0 ๐ฌ 2 ๐ 0
4o would beg to differ on the backswing there
30.01.2026 06:04 โ ๐ 2 ๐ 0 ๐ฌ 0 ๐ 0
Also curious as to in which sense a given person means they're stochastic. Talking about randomly initialising the weights seems daft; temperature/softmax seems like a daft target; what's left? There was a paper recently IIRC showing (intuitively?) same input same output (barring temp)
30.01.2026 03:57 โ ๐ 3 ๐ 0 ๐ฌ 1 ๐ 0
The most compelling explanation I've heard is that it's overfitting on the social media war: common knowledge is, and I guess status is tied up in declaring that, Big Tech cannot live up to its own promises and only delivers enshittification.
30.01.2026 03:44 โ ๐ 6 ๐ 0 ๐ฌ 1 ๐ 0
this is exhaustive and also has a hilariously put-upon tone that I appreciate. Focus on the power use and the bad effects of the outputs and the negative social externalities!
30.01.2026 03:18 โ ๐ 58 ๐ 9 ๐ฌ 1 ๐ 1
I don't know what team you consider yourself on, and don't care if you block me, but please stop shooting your own team in the foot by telling them factual inaccuracies that will cause them not to take action when it matters. You are abdicating the most important conversation of history.
30.01.2026 03:03 โ ๐ 3 ๐ 0 ๐ฌ 0 ๐ 0
I like that you only followed me after this reply lol didn't realise there was going to be a test ๐
29.01.2026 07:24 โ ๐ 1 ๐ 0 ๐ฌ 1 ๐ 0
Yeah it was tbh even through a bunch of sauce
29.01.2026 07:01 โ ๐ 1 ๐ 0 ๐ฌ 1 ๐ 0
Not least of all because you then need a RNG for your RNG lol
23.01.2025 03:05 โ ๐ 1 ๐ 0 ๐ฌ 0 ๐ 0
Bad (illegal?) solution(?): 99 flips - randomise the order of the interval and flip until you get heads, then pick the one at that position. If tails at 99, pick the 100th.
23.01.2025 03:04 โ ๐ 1 ๐ 0 ๐ฌ 1 ๐ 0
26.12.2024 18:14 โ ๐ 113 ๐ 10 ๐ฌ 5 ๐ 0
I think it must be driven by the Lump of Labour fallacy... While other foreign born Americans do get & create jobs, they're not the plum prizes (that people think *they* could get or something??). Eli's reply seems right to me.
27.12.2024 00:08 โ ๐ 0 ๐ 0 ๐ฌ 0 ๐ 0
falaffle
17.12.2024 07:08 โ ๐ 4 ๐ 0 ๐ฌ 1 ๐ 0
Incredible, no notes, more please
15.12.2024 12:02 โ ๐ 1 ๐ 0 ๐ฌ 0 ๐ 0
CEO killing, real bad stuff seemingly coming out of Syria, o1 eval... tough news week for a new dad (this one)
10.12.2024 08:22 โ ๐ 1 ๐ 0 ๐ฌ 0 ๐ 0
Lot of great bits in the fish episode recently - was a great challenge to assumptions I didn't even realise I had!
10.12.2024 06:01 โ ๐ 0 ๐ 0 ๐ฌ 0 ๐ 0
2 conspiracy theory ideas:
- Musk enshittifying competitors
- Altman drumming up demand for that worldcoin UUID thing
06.12.2024 09:03 โ ๐ 1 ๐ 0 ๐ฌ 0 ๐ 0
I prefer {renv} but same idea
06.12.2024 06:53 โ ๐ 0 ๐ 0 ๐ฌ 0 ๐ 0
Would it be right to infer you define alignment as 'does what user intends' rather than 'priorities notkilleveryoneism'? I'd say this is different evidence for different definitions...
06.12.2024 06:49 โ ๐ 1 ๐ 0 ๐ฌ 1 ๐ 0
Google co-founder Larry Page is reportedly not worried about machine superintelligence, not because he doesn't think it could kill everyone, but because he doesn't seem to care!
04.12.2024 19:18 โ ๐ 2 ๐ 1 ๐ฌ 1 ๐ 0
Some of the stuff Cruz points to just seems like reasonable object-level beliefs rather than pandering? I'd be surprised if Cruz (or much of the rest of the right) would have proved more useful/stalwart coalition members - e.g. for SB1047 it seems like a16z et al subverted people of all stripes.
03.12.2024 07:42 โ ๐ 1 ๐ 0 ๐ฌ 0 ๐ 0
I'm disinclined to take this motte & bailey as too much evidence for anything; presumably Cruz would have found another bailey.
03.12.2024 07:37 โ ๐ 1 ๐ 0 ๐ฌ 0 ๐ 0
suitcase pile looming directly above the head, no less
01.12.2024 21:17 โ ๐ 2 ๐ 0 ๐ฌ 1 ๐ 0
this Deep Utopia could have been a flowchart
30.11.2024 22:57 โ ๐ 1 ๐ 0 ๐ฌ 0 ๐ 0
Just a little post for all you EA forum users as a treat
30.11.2024 21:08 โ ๐ 2 ๐ 0 ๐ฌ 0 ๐ 0
Anyone know when Draft-Amnesty-Week-Draft Amnesty Week is?
30.11.2024 21:07 โ ๐ 0 ๐ 0 ๐ฌ 1 ๐ 0
Should have been Ghost in the Smell
30.11.2024 20:59 โ ๐ 1 ๐ 0 ๐ฌ 0 ๐ 0
Interested in #Infosec and #ProductiveDisagreement | #StayAtHomeDad, worked in #Biotech, co-built tiny companies in renewables and structural engineering sectors, ex #HumanRights observer | likes #ElixirLang, #Boardgames |๐ธ 10% @givingwhatwecan.bsky.social
A LLN - large language Nathan - (RL, RLHF, society, robotics), athlete, yogi, chef
Writes http://interconnects.ai
At Ai2 via HuggingFace, Berkeley, and normal places
Professor at Wharton, studying AI and its implications for education, entrepreneurship, and work. Author of Co-Intelligence.
Book: https://a.co/d/bC2kSj1
Substack: https://www.oneusefulthing.org/
Web: https://mgmt.wharton.upenn.edu/profile/emollick
temporarily embarrassed orbital weapons platform
Political progressive, economic liberal, bisexual centrist โข Spanberger-Mamdani Democrat โข Washington, DC apologist โข GOP delenda est
ultralight power and energy, physics, ai, building in secret in public
๐ london
๐ฆ emojistats.bsky.sh | ๐ท๏ธ https://labeler-starter-kit.bsky.sh | ๐ฉ https://bluesky-heatmap.fly.dev
๐ช @sortinghat.bsky.sh ๐ @zodiacsigns.bsky.sh ๐งฉ @mbti.bsky.sh ๐ค @eras.bsky.sh โ
@a.rnbw.ca
๐ค @songofsongsbot.bsky.social
recovering cryptographer building ML models, doing systems work, security, etc.
Como todos los hombres de Babilonia, he sido procรณnsul; como todos, esclavo; tambiรฉn he conocido la omnipotencia, el oprobio, las cรกrceles.
very sane ai newsletter: verysane.ai
random bloggy bits: segyges.leaflet.pub
croissanthology.substack.com
croissanthology.com
Ex-philosopher, ex-Tweeter.
Email: info@contrapoints.com
He/him. Fox gamedev irl. Machine pervert. Furry stuff here and opinions about technology, games, movies, AI, VR. Pfp + Banner ๐จ by @Perpleon.bsky.social
https://vrchat.com/home/user/usr_d9ee62da-234d-445a-8acc-720e0d33f8f0
๊ฎ surfed on by the information superhighway
๊ฎ ๐ @linneaisaac.bsky.social
๊ฎ she/they ๐ณ๏ธโโง๏ธ
๊ฎ fiction/art/blog/games @ https://vgel.me
๊ฎ llms at acsresearch.org
charts and graphs
follow me on twitter https://twitter.com/norvid_studies
Independent AI researcher, creator of datasette.io and llm.datasette.io, building open source tools for data journalism, writing about a lot of stuff at https://simonwillison.net/
Locked in and posting regularly on here now
i'm trying to get back in touch with my feelings and talk about them more
๐ฆฃ tortle@schelling.pt
๐ค @ohwizenedtortle