پن کیا گیا ٹویٹ
AI Things
1.7K posts
🇫🇷 Mistral AI in Talks for €3B Funding at €20B Valuation
Mistral AI is in talks to raise €3B at a €20B valuation, providing the Paris-based company with resources to compete in the global AI race against US and Chinese rivals
Source: Bloomberg
English
The correct way to handle AI cyber security: No guardrails.
Grummz@Grummz
The only way to make software more secure is to allow AI to scan for vulnerabilities and fix it. But this is what got Fable 5 pulled. The stuff that fixes vulnerabilities also reveals them, obviously. The only solution is no guardrails. Use AI to fix it all and find it all. Otherwise you just leave hacks to state level actors and leave everyone else vulnerable.
English
What Europe should do right now:
1. Call all the European researchers working on AI and return them back with same salary (or they can stay but switch career).
2. Fill EU places having GPUs with money, and put those people there.
3. AI partnerships with China + India.
English
@initjean I wonder what holds EU that much back, I feel like its not "only" or "primarily" because of regulations.
Europe has money, human capital, tech etc. Even only ASML is a big edge.
They may need to take some steps back to a point of a newer and different AI chips approach.
English
i hope Amazon loses another data center
Stephanie Palazzolo@steph_palazzolo
Breaking: Amazon CEO Andy Jassy was among the tech leaders who raised concerns to senior Trump officials this week re: security risks in Anthropic's newest models. Those convos set in motion the government's new export controls on foreign national access to Mythos and Fable.
English
@michuk @MistralAI Europeans may need to take a step back and start from ai chips point. May they start with a new, different approach to ai chips, since it looks like biggest bottleneck.
English
This is dangerous for Europe.
We should act super fast & together.
Our best chance is all hands on board to accelerate @MistralAI, the only European company that has models on nearly the same level than the US and Chinese leaders.
All European national sovereign funds should immediately put 1B EUR each into Mistral and make sure the company always stays European (not just French).
And as soon as reasonably possible, we should start using those models day to day for work, first in governments and state companies, then elsewhere.
The only way to win is to stay independent and own the tech ourselves.
Anthropic@AnthropicAI
The US government, citing national security authorities, has issued an export control directive to suspend all access to Fable 5 and Mythos 5 by any foreign national, whether inside or outside the United States, including foreign national Anthropic employees. The net effect of this order is that we must abruptly disable Fable 5 and Mythos 5 for all our customers to ensure compliance. Access to all other Claude models is not affected. We apologize for this disruption to our customers. We believe this is a misunderstanding and are working to restore access as soon as possible. Read our full statement: anthropic.com/news/fable-myt…
English
OpenRouter just introduced Fusion API, a compound model that claims Fable 5-level intelligence at half the price.
Sounds impressive, but I’d like to see how it holds up in real-world use.
Has anyone tried it yet?
OpenRouter@OpenRouter
Introducing the Fusion API, the smartest compound model in the market. Fusion achieves Fable-level intelligence at half the price. How it works 👇
English
Notably, the budget panel was comparable with Claude Fable 5 in performance.
A panel of Gemini 3 Flash, Kimi K2.6, and DeepSeek V4 Pro, fused together, beat solo GPT-5.5 and solo Opus 4.8 outright.
And it landed within 1% of Fable 5 while costing roughly half the price.
English
Introducing the Fusion API, the smartest compound model in the market.
Fusion achieves Fable-level intelligence at half the price.
How it works 👇
English
@Im_IrushiK Normal people would lose i think
Government controlled ai vs doomer controlled ai
English
@Evinst3in I think they are not that much close with Anthropic to set something like this.
English
What if Anthropic and the government planned this in advance and we all fell for it?
English
all previous "nerfs" were nothing compared to this
English
🚨 New GLM-5.2 is out
▸ Third place on KingBench🥉
▸ One million tokens context window
▸ Preliminarily ahead of GPT-5.5, Gemini 3.5 Flash, and Opus 4.7.
The @Zai_org team's new model is off to a good start. Let's see if those results get confirmed by more benchmarks!
English
@techikansh whether we get or not will clarify if this fable ban is only special to Anthropic or not
English
Kimi Code is as good in FrontierMath Tier 4 as GPT-5 mini and GPT-5.4 nano 🙈
English
@pseudokid btw mimo is a truly making me sad model unfortunately. i love kimi,qwen. I think deepseek v4 pro could be smarter but its kinda not the leading Chinese model. Glm, Qwen, Kimi are good. Minimax isnt bad but they may need to have bigger models.
English
Okay we have Kimi K2.7 and GLM 5.2 today
I have to get used not seeing GLM 5.2 as "GPT 5.2"
These names bro smh
AICodeKing@aicodeking
GLM-5.2 on KingBench (3). Thoughts: The model has superb taste. It is greater at UX than UI. The code is always very clean. It is great at One-shot wonders. I asked it to fine-tune a whole local model and it did it in 30mins! This is just a great model to use all-round. 1/n
English
@wawilow108 @varrock I think thats not the point. If thats the case, we should see close performance on Arc-Agi, maths etc.
English
@varrock The Chineese LLM's are trained by and for Mandarin speakers, there are no incentives to make the best english speaking model
English
OpenAI has removed 5.2-Pro from ChatGPT. The best model for math/ML is not available anymore. Yes, it was mostly through an unbelievably high reasoning budget, but still. A dark day - and I'm not being facetious :-(
English
@bayeslord I think America is still the best place and apparently with all human and financial capital they will keep their edge
But the problem in my mind is different
English
