Andrei Lahunou
138 posts
Andrei Lahunou
@Systerr
IT, blockchain, agents, development https://t.co/uxeJgAvMzG UK Global Talent Founding [email protected]
London 가입일 Mart 2009
164 팔로잉114 팔로워
a 3D game i made completely by GLM 5.2 and literally nothing else.
plotva.geta.moe
English
This is a huge. I start using dev versions recently and they are fast. Marketing that says 10x speed up was right
TypeScript@typescript
📣 TypeScript 7's Release Candidate is now out! 📣 The new native port is almost here. Try it out on your codebases, and make sure your team is ready for the upcoming 7.0 release! devblogs.microsoft.com/typescript/ann…
English
@Taniyatweets_ I do. But should I? If it match specs and another Ai confirms it why should I care what's inside ?
English
@JunaidAckroyd What if no one from them? What if eventually @subquadratic will be a thing?
English
Who is winning the AI race?
- Anthropic
- OpenAI
- Deepmind
- xAI
English
@klmivn Так і ёсць. 20 даляраў план на паглядзець. Працаваць с ім нельзя аніяк
Українська
Бляха, я хз як карыстацца claude з іх лімітамі. Відаць, толькі на Max можна нешта рабіць актыўна, 20$ - так, пагуляцца. Трэці раз за дзень выбіраю за кароткія сессіі па 20-30 хвілін
Українська
@shadcn That's a cool insides and open a whole new usage surface. Install just is adding file - no any imports that means for a lot of user and agent less work to install
English
@Agentplace_io 6/6 Agentplace isn’t just an agent builder.
It’s a 24/7 PhD-level teammate that truly understands your business — for cents, not salaries.
Building agents is just one feature.
English
@Agentplace_io 5/6 It shows all the math, money flow, and updated charts in one clean tab.
Then tell it: "Review this every 4 hours and alert me if the winner changes."
Done.
English
1/6
Everyone thinks @Agentplace_io is just for building AI agents.
It is. But the same no-code builder can also become a ruthless business analyst that reasons over your real data in plain English. 🧵
English
45% of queries don't come from humans. Who do you think produces them?
English
We have tests where Chinese models beats available USA models. And it costs less. What's next?
ModelScope@ModelScope2022
GLM-5.2 is now live on ModelScope! Built for long-horizon tasks, with Solid 1M context and stronger coding capabilities. 🚀 🤖 modelscope.ai/models/ZhipuAI… #2 on Code Arena among all available models globally. Highest-ranked open-source model on FrontierSWE, SWE-Marathon, and PostTrainBench, performance between Claude Opus 4.7 and 4.8. 🧠 Solid 1M context: KV8, LayerSplit, IndexShare 💻 Coding: Terminal-Bench 2.1 open-source SOTA (+17.5% over GLM-5.1), MCP-Atlas within 0.8% of Opus 4.8. Engineering feel between Claude Opus 4.7 and 4.8 ⚡ Long-horizon execution: from a single requirement description to a fully deployable multi-platform app in one continuous run vLLM, SGLang, Transformers supported.
English
Maybe that's the reason why @aisdk has only the simple agent out of the box?
Vercel@vercel
Introducing eve, an agent framework. 𝚊𝚐𝚎𝚗𝚝/ 𝚊𝚐𝚎𝚗𝚝.𝚝𝚜 𝚒𝚗𝚜𝚝𝚛𝚞𝚌𝚝𝚒𝚘𝚗𝚜.𝚖𝚍 𝚝𝚘𝚘𝚕𝚜/ 𝚜𝚔𝚒𝚕𝚕𝚜/ 𝚜𝚊𝚗𝚍𝚋𝚘𝚡/ 𝚜𝚌𝚑𝚎𝚍𝚞𝚕𝚎𝚜/ Like Next.js, for agents. vercel.com/blog/introduci…
English
How to be the best on marker
- build best product
- OR find money and acquire the best product
Both approaches works well
SpaceX@SpaceX
SpaceX has exercised the option to acquire @cursor_ai in an all-stock transaction with the goal of building the world’s most useful AI models. For the past few months, SpaceXAI has been jointly training a model with Cursor, which will be released in Cursor and Grok Build soon. We look forward to working closely with the Cursor team to advance our frontier AI capabilities
English
Andrei Lahunou 리트윗함
Linux kernel 7.1 is released (removing 140,000 lines of code that supported some legacy software).
Rising AI-assisted code debate aside, the new release brings a rewritten NTFS driver and some improvement for AMD and Intel CPUs.
The new release also dropped supports for several legacy hardware. The idea is to have a leaner kernel which is easier to maintain as AI keeps on finding security flaws in obsolete drivers and thus making the job of maintainers difficult.
English
We got early access to le chaton fat from @MistralAI and honestly? It beat all our expectations. Faster and more powerful than Mythos. Never underestimate a fat kitten.
English