othy.h

5.4K posts

othy.h banner
othy.h

othy.h

@othy_h

J’aime : Les Maths, animes/mangas, cultures 😊.

Katılım Temmuz 2018
1K Takip Edilen86 Takipçiler
othy.h retweetledi
🍓🍓🍓
🍓🍓🍓@iruletheworldmo·
happy 5.6 day everyone, now we can all solve math.
English
101
20
947
75.6K
othy.h retweetledi
Alexander Doria
Alexander Doria@Dorialexander·
knowing how to spin subagents on gpu, on track to become a very consequential skill
English
5
1
29
1.2K
othy.h retweetledi
Tyler
Tyler@rezoundous·
Codex is not feeling so codexy anymore.
English
48
14
357
19.9K
Chubby♨️
Chubby♨️@kimmonismus·
Microsoft put $13 billion into OpenAI and built the cloud infrastructure Anthropic runs on. This week it canceled its internal Claude Code licenses because the token bill was too high. Even for MSFT Claude is too expensive.
Hedgie@HedgieMarkets

🦔Microsoft canceled its internal Claude Code licenses this week after token-based billing made the cost untenable, even for a company with effectively infinite cloud resources. Uber's CTO sent an internal memo warning the company burned through its entire 2026 AI budget in just four months. American AI software prices have jumped 20% to 37%, and GitHub (owned by Microsoft) is dropping flat-rate plans for usage-based billing across its products. My Take The AI subsidy era is ending in real time. The same company that put $13 billion into OpenAI and built the Azure infrastructure powering most of Anthropic's compute just looked at the bill from a competitor's coding tool and decided it was not worth paying. That is not a productivity failure on Anthropic's end. Token-based pricing is forcing every enterprise customer to confront the actual cost of running these models at scale, and the number turns out to be far higher than the flat-rate experiments suggested. This ties directly to my Gemini Flash post yesterday. Anthropic, OpenAI, and Google all raised effective prices in the last six months. Enterprises that built workflows assuming AI costs would keep falling are now watching annual budgets evaporate in months. Two outcomes look likely from here. Either enterprises scale back AI usage to fit budgets, which slows the revenue ramp the labs need to justify their valuations ahead of IPOs, or the labs cut prices and absorb the losses, which makes the unit economics worse at exactly the wrong moment. Both paths land in the same place, the numbers stop working, and somebody has to take the writedown. Hedgie🤗

English
36
7
215
26.2K
othy.h retweetledi
Elon Musk
Elon Musk@elonmusk·
Humans using Mythos as seen by Mythos
GIF
English
2.9K
2.4K
41.2K
3.2M
othy.h
othy.h@othy_h·
@ajambrosino You need to add long context mondel and better parallel agent harness
English
0
0
0
35
Andrew Ambrosino
Andrew Ambrosino@ajambrosino·
codex thursday no. 6: - appshots - /goal - remote computer use while locked - advanced annotation mode - plugin sharing - improved analytics - small stuff everywhere
English
61
20
453
20K
othy.h
othy.h@othy_h·
Tbh feel like the Native Harness of claude code is the best by far, absolute killer @bcherny >>>>> @thsottiaux , 1M context may help too why OpenAI doesn’t have long context model released yet ? Sorry. But Tibo has better vagueposting game
English
0
0
0
27
Haider.
Haider.@haider1·
Dario Amodei says Mythos showed unexpected strength in cybersecurity It was better than human software engineers at finding ways code could be exploited, and proved useful for both attacking and defending systems "this thing kind of seems like a weapon"
English
14
11
86
9.5K
othy.h
othy.h@othy_h·
@rezoundous It’s strategy, anthropic is taking over their productively market, probably redirected the budget to openai …
English
0
0
0
130
Tyler
Tyler@rezoundous·
Microsoft canceled Claude Code license due to unsustainable costs. If they can't afford it, who can?
English
60
22
404
28.5K
Valentin Sch
Valentin Sch@val_strch·
Le top 10% est atteint Prochain palier : top 100 000
Valentin Sch tweet media
Français
5
0
19
1.5K
othy.h
othy.h@othy_h·
AGI timeline is where roon post this
GIF
roon@tszzl

English
0
0
1
28
othy.h
othy.h@othy_h·
@tszzl Ultra instinct gpt when ?
English
0
0
0
28
roon
roon@tszzl·
GIF
ZXX
36
24
375
14.2K
othy.h retweetledi
Susan Zhang
Susan Zhang@suchenzang·
so they were actually serious about ai welfare i used to roll my eyes when they said ai welfare and human welfare are the same thing now i'm not so sure. maybe this is a good thing? does being mean to bots mean they will be mean to humans? is it just inevitable that bots get anthropomorphized? are we just assuming humans won't be able to tell the two apart eventually?
NIK@ns123abc

> be me > first tell claude it’s dumb > then immediately ask it to look up some article > claude: “I'm happy to look it up, but I'm not going to keep talking if you call me names. Knock that off and I'll search” > wait… it just threatened me > then asks “do you remember when did it come out or what was it about? prompting tips, getting better results, something like that?” > me: “i dont and yes u are dumb” > 0.3 seconds later > claude: “I’m going to stop here. Take care.” > activates End Conversation tool > Done > *chat ended by claude* This is insane.

English
16
2
61
13K
othy.h retweetledi
NIK
NIK@ns123abc·
> be me > first tell claude it’s dumb > then immediately ask it to look up some article > claude: “I'm happy to look it up, but I'm not going to keep talking if you call me names. Knock that off and I'll search” > wait… it just threatened me > then asks “do you remember when did it come out or what was it about? prompting tips, getting better results, something like that?” > me: “i dont and yes u are dumb” > 0.3 seconds later > claude: “I’m going to stop here. Take care.” > activates End Conversation tool > Done > *chat ended by claude* This is insane.
NIK tweet mediaNIK tweet media
English
305
64
1.9K
313.1K
will brown
will brown@willccbb·
switching back to Cursor Tab
English
8
0
65
6.2K
othy.h
othy.h@othy_h·
@thsottiaux It works only for you guys now not for the pro customers 😂
English
0
0
0
49
Tibo
Tibo@thsottiaux·
Codex is our WorkGPT
English
206
21
1.2K
50K
othy.h retweetledi
Daniel Litt
Daniel Litt@littmath·
Doing my best to communicate responsibly, but wow the discourse is rough.
English
27
9
523
21.8K