🔥The 9th Round of Easy Loan, Earn $40 Reward is in progress❗️ ⏰ Promotion Period: January 15th - Feburary 15th, 2025 👉 Register now and check more details at gate.io/campaigns/358

pip install genai-processors

Forget about that "ding", they just iced the model release: x.com/sama/status/19…

youtube.com/live/G7isIBuAd…

theconversation.com/ai-is-driving-…

reading 📑 arxiv.org/pdf/2507.06261

kimi.com Kimi.ai is🧠

I have a model that can spot a money link on any page. It's near 100% accurate. Think about what I just said.

Wrote this one for myself to refer back to it when I need it, but you're welcome to take a looksie :) dejan.ai/blog/analysis-…

GPT-4o and possibly the rest lost about 30% IQ in the last few days. Thank god for Gemini and Claude.

Unable to allocate 22.5 TiB for an array with shape (2486928, 2486928) 😬OK FINE! I'll just use FAISS.

GSC data. Why 16 months and not 12,18 or 24? Did we ever find out?

by Lily Ray 😏 searchengineland.com/ai-search-boom…

note to self: <ctrl94> is Gemini's internal system token for thinking

My team is going to love this: youtu.be/2zly04mUjxY

What do humans and AI have in common? We don’t read. Instead we rely on attention mechanisms to process text information. When optimising content for AI and humans you must get to the point early and optimise content to reduce cognitive load. dejan.ai/blog/human-fri…

For my text chunk-loving friends: chunk.dejan.ai

Really... 42 seconds? I hope all reasoning models die a swift death in favour of model architectures which internalise the process instead. And it's not just the speed. It's the fact that the model talks to itself instead of me and makes objective-misaligned decisions for me.