@crystlizedd

wow its that guy

18+
Joined June 2025
artificial analysis if it was good
1
22
374
crystl :3 retweeted
Que putas está pasando que varios países están sacando comunicados literal diciendo que se prepare todo el mundo
1,323
3,256
215
69,736
8,444,208
crystl :3 retweeted
Things you see right before the market implodes
This is actually what the New York Stock Exchange (NYSE) looked like last night
101
1,169
25
20,867
779,323
crystl :3 retweeted
JUST IN: Open-source AI models surge to a record 78.4% of token volume on Vercel’s AI Gateway.
61
43
12
669
59,468
crystl :3 retweeted
they literally have no idea what kind of creature he is that is NOT a normal cat 😭
40
795
73
16,587
261,060
crystl :3 retweeted
RSA-896. In the last 16 days the public factoring record jumped from 829 bits to 862, now to 896. Through AI assistance.
636606729769440499166579950236036751749912014371509557713570027508971809534551913252252094954941974952859310861988904737359709200557919 is a factor of RSA-896 saweis.net/posts/rsa-896.htm…
22
14
3
623
49,483
holy shit its verity
I will die your "friend". I will. I will die your "helper". I will. I will die your "A.I. assistant". I will.
4
18
327
crystl :3 retweeted
Don’t anthropomorphise LLMs, they hate that.
35
70
7
662
21,309
oh my god i hate claude for coding. what. i decided to try fable 5.1 for coding (previously only used for chat) and this is mind-numbingly bad. how do people use this? it's so bad! we need a moratorium NOW
6
2
19
621
crystl :3 retweeted
2026 is the year of linux desktop
1,006
425
229
8,720
663,561
crystl :3 retweeted
“We want evaluators more independent than METR!” *monkey’s paw curls*
We’re partnering with Accenture on independent evaluation of frontier AI—part of our recent commitment to embed evaluators at Anthropic. Both we and Accenture expect to invest at least $1 billion to build capacity in this area over the next five years. anthropic.com/news/accenture…
Readers added context they thought people might want to know
Anthropic presents this as an "independent evaluation" but will directly fund Accenture's work and has a prior commercial partnership with the firm for deploying its models, including training ~30,000 Accenture professionals on Claude. anthropic.com/news/accenture… anthropic.com/news/anthropic…
18
29
3
964
46,523
Fable 5 was the best model until it came back, Astra is like how prime fable felt ngl
remember what life was like back then?
8
so what really is the difference between chatgpt not training on my prompts or using my data to improve their model, and them just wholesale feeding the contents of my chats to advertisers for reidentification?
1
1
3
75
crystl :3 retweeted
hi i'm an anthropic employee and the rumors are true we quantized fable to 1.58 bits and serve that at peak times and there is nothing you can do about it
1
1
65
1,350
This you big dawg?
They added a jet into CS2
1
493
4
15,481
167,229
Moonshot, can we get a ~300b Kimi K3.1 Flash please 😭
30
crystl :3 retweeted
Kimi K3 is the closest model to Fable in style. If they improve efficiency in 3.1, I don't think it'll make sense to pay for Fable and Astra anymore.
29
10
3
264
15,471
crystl :3 retweeted
Bonsai-2 27B レビュー 結論: GSQ-RCOを使え
13
9
2
165
17,752
crystl :3 retweeted
Adding this to AGENTS.md
124
25
28
1,237
147,320
crystl :3 retweeted
automatic translation
170
6,287
103
77,329
714,277
crystl :3 retweeted
Startup idea: open air market for exotic meats across the street from Anthropic’s wet lab
68
355
17
6,217
125,870