My main take away of the Deepseek paper is not scientific but organizational: we need an European industrial plan in AI right now. No safety summit, no peppered compute grants, no funding processes that take two years.

The DeepSeek R1 training procedure confused me at first. My brain refused to accept this powerful model could be incredibly straightforward. Let me break down this elegant beast for you 🧵

Just like the voices in my head

"We are living in a timeline where a [Chinese] company is keeping the original mission of OpenAI alive - truly open, frontier research that empowers all."

Europe: But we had AI so well regulated! How could we lose the race?

The most exciting thing about DeepSeek R1 isn’t that it’s open-source or matches OpenAI’s o1 in reasoning tasks—or even that it’s 90–95% cheaper. The real breakthrough? It proves that LLMs can improve reasoning through pure reinforcement learning, without massive CoT datasets.

Now that Trump has announced $500B for AI, Europe will double down by announcing 1000X more AI regulations.

👀

IQ is the only real unfair advantage

𝐋𝐞𝐭’𝐬 𝐛𝐞 𝐫𝐞𝐚𝐥: 𝐌𝐨𝐬𝐭 𝐀𝐈 𝐭𝐨𝐨𝐥𝐬 𝐭𝐨𝐝𝐚𝐲 𝐚𝐫𝐞 𝐟𝐥𝐚𝐬𝐡𝐲 𝐝𝐞𝐦𝐨𝐬 𝐭𝐡𝐚𝐭 𝐟𝐚𝐥𝐥 𝐚𝐩𝐚𝐫𝐭 𝐢𝐧 𝐫𝐞𝐚𝐥-𝐰𝐨𝐫𝐥𝐝 𝐞𝐧𝐭𝐞𝐫𝐩𝐫𝐢𝐬𝐞 𝐩𝐫𝐨𝐝𝐮𝐜𝐭𝐢𝐨𝐧. There’s a persistent myth that fully unsupervised agents—or swarms of them—can

100% we have been running Deepseek in production large scale use cases and it behaves 100x better than Geminin 2.0 Flash Exp on rate limiting.

Why AI Won't Cause Unemployment Marc Andreessen Reposted Jan 24, 2025 "In retrospect, I wish I had known more about the hazards and difficulties of [running] a business." -- George McGovern Fears about new technology replacing human labor and causing overall unemployment have

i essentially dont use claude anymore

Here are the top AI Papers of the Week (Jan 20-26): - DeepSeek-R1 - Can LLMs Plan? - Chain-of-Agents - Scaling RL with LLMs - Humanity’s Last Exam - Agentic RAG Overview Read on for more:

We just launched our latest Feely work - Noxus- shaping the future of enterprise AI! We crafted their brand identity around cellular transformation, designed the website, and brought it to life with custom animations. Thanks, João Pedro Almeida for your trust ⚡️

don’t train LLMs build products on top of them finally blatantly clear thanks to R1 (-> the billions raised by labs was wasted)

Great piece from Mario Draghi: ft.com/content/13a830….

github.com/Noxus-AI/no-llm Noxus has released it's first open-source library! Ever wanted to build a service with a lot of LLM, but don't want the hassle of integrating every single provider? Look no further, we got the solution right here! Feedback is always appreciated ❤️

Sem qualquer promo, a Noxus encheu o Alhub no 1° “Enterprise Al Talks". Próxima semana Londres. Falámos de use cases de AI nas empresas e no Estado, regulação e como atrair capital/inovação. Aos interessados, a Noxus está a contratar um Founder Associate - GTM. Link em baixo.

Great news from Lisbon. We’ve just tripled our office space! Still in the middle of decorating. Waiting for @levelsio to visit us at the AI Hub! Noxus is growing fast. Teams expanding in London, Lisbon, and beyond. Let’s keep building Europe! #Noxus