Nick Kulikaev
1.6K posts
Nick Kulikaev
@NickKulikaev
Quit corporate career to become a billionaire. https://t.co/dDp9Qd0KdB | https://t.co/PvWgBiKjO7 | https://t.co/arHkeK8u18 | https://t.co/JrDPZzJQvV
Katılım Temmuz 2019
353 Takip Edilen124 Takipçiler
probably the harness, not the model. one-shotting complex stuff flails on everything, 4.8 and 5.5 alike. spec it tight, make it write a RED-GREEN test first, run a second agent to tear the diff apart, and 4.8 closes most of that gap on its own. i wired that exact loop into a free claude code plugin if you wanna test it: github.com/inferencegod/a…
English
Nick Kulikaev retweetledi
@TBGamer9000 @AlexFinn Sure but it’s not going to be opus 4.7 level. I agree that having your own compute with at least 128gb is something to aim for. But running top models locally doesn’t seem feasible near future. So Alex oversells it.
English
@NickKulikaev @AlexFinn To be fair there should be more optimized/quantized ton 5.2 versions coming.
English
I was wrong
I've been saying for months that open source AI models are 6 months behind frontier
They caught up. GLM 5.2 is as good as Opus 4.8
This changes everything. If you run GLM 5.2 locally no government can take it away. You become sovereign
And even if you run through APIs, its a fraction of the cost
The battlefield is different now. If open source is as good as frontier, and people have cheaper alternatives, governments can't be as quick to regulate. It will destroy the frontier AI labs
All of this is such a massive win for the people
If you are not paying attention to local models yet, you are making a tremendous mistake
Z.ai@Zai_org
Introducing GLM-5.2: Frontier Intelligence, Open Weights - Significant improvements in coding and agentic tasks - Strong long-horizon capabilities with a 1M context window - Two levels of reasoning effort: GLM-5.2 (max) pushes the limits, while GLM-5.2 (high) strikes a strong balance between performance and token efficiency - MIT-licensed open weights - Same API pricing as GLM-5.1 Tech Blog: z.ai/blog/glm-5.2 Weights: huggingface.co/zai-org/GLM-5.2 API: docs.z.ai/guides/llm/glm… Coding Plan: z.ai/subscribe Chat: chat.z.ai
English
@Atilla_Guerilla @AlexFinn Curious what was your token cost per hour when fable was available?
English
@NickKulikaev @AlexFinn On API, I believe Anthropic might be subsidizing/rebating my employer
English
@NickKulikaev @AlexFinn My ClaudeCode usage is around 20-30k a month
English
@garrytan Feels like 15% productivity gain is on the low end. Not average. I can’t even make myself work with any other model right now.
English
Rough estimate on $ productivity lost by Fable 5 ban: $12M per hour
Frontier AI-coding daily actives, mid-2026: 5M devs
Fully-loaded cost: $90/hr
Work routed to Fable in 48 hours: 17.8%
Fable is on average ~15% more productive
Effective throughput loss per dev = 17.8% × 15% ≈ 2.7% of output
2.7% × $90/hr = ~$2.40/dev/hr × 5M devs = $12M per working hour
English
@0xSero I can hardly find anyone who’s still drinking in SF.
English
@thekitze unlike when? Italy is always spectacular - unlike Google UX.
English
@ChrissGPT What is Salesforce doing there? Selling crappy saas?
English
English
@levelsio Pieter, did you change something about commenting restrictions? For about a month I couldn’t comment on your posts. Whatever that filter was it was annoying as hell. Twitter is all about engaging in conversations.
English
You have about 50x to 100x higher odds to die in a private jet than flying commercial
The reason isn't that the planes are bad btw, they're great, but private jets flights are much less strictly regulated than commercial flights
They can fly more custom routes, fly into tiny airports and when the passenger (usually a rich person) really needs to be at a meeting in time they might decide to fly through bad weather when a commercial flight would cancel or delay the flight
Simply speaking private jets have a lot more freedom which is good and bad
Commercial flights are extremely strictly regulated, fly extremely boring routes, the same every day or every few hours back and forth, highly predictable and if one pilot encounters some issue like bad weather they're able to tell the next flying the same route
Even if I could afford to fly private, I'd prefer to just board with fast track and go first or business class, it's much safer!
Breaking Aviation News & Videos@aviationbrk
A NetJets Cessna 680 Citation Latitude crashes on a highway while making an emergency landing to Laredo International Airport, Texas. The crash was reported around 10 p.m. It was not immediately clear how many people were on board or their condition. Nobody on the highway was injured during the crash, police said. The flight departed from Los Cabos International Airport around p.m. MT and was on its way to Austin-Bergstrom International Airport when it crashed, according to tracking data.
English
@thsottiaux @Dimillian But can I use it with Fable when it comes back?
English
@thekitze Google design team has done more damage to humanity than barbarians. Spreading absolute primitive cavemen level aesthetics.
English
google cannot make software to save their life
incredibly ngmi company
English
@Youssofal_ Yes, 5.5 feels twice slower even in fast mode. And 3 times dumber. Feels like a waste of time now. Taking a break.
English
I used to be excited for Codex rate limits.
Now after Fable I really don’t care anymore. I have 3 codex subscriptions all sitting at 100%.
I don’t even log in anymore.
Tibo@thsottiaux
This was fixed. You know what's coming 👀 Give us 24 hours to reset the Codex rate limits across all plans.
English
@marclou @trust_mrr You can just do any thing now but it reminds me of posting in X. If you have a large following any thought fart post gets reach and attention. If you don’t, you get 3 views. Same with whatever thing you build.
English
My vibe-coded startup marketplace @trust_mrr just reached 30K buyers 🌟
It has helped 100+ startups get acquired in my sleep, run 100% on autopilot, no employees, no investors, and costs <$500/mo.
You can just do the thing!
English