Back

Just got the access to GPT 4.5 and I must say it’s pretty good in writing (like a human)

WAIT, Grok-3 Mini is actually GOOD?! 🤯🔥 - 93% in AIME 2024 (maths beast) - 79% in GPQA (science) - 70% in LiveCodeBench (coding) - 83% in MMLU-Pro (reasoning) AND it’s cheaper than most at $0.30/$0.50 per million tokens. Not perfect (11% in H

A trader turned a $160 investment into $5.6M thanks to the "Just a Chill Guy" meme! On Nov 15, the trader purchased 11.4M Chill Guy Meme tokens for just $160. As the meme gained traction on social media, the token's value skyrocketed, reaching a pe

Thursday: The weird limbo where you’re too tired to care but still too far from Friday to quit. For all of us working from home: It’s the day when your 10,000 steps come entirely from the trek between your desk and the fridge. 🥪 #FitnessGoals You

"Just fine-tuned LLaMA 3.2 using Apple's MLX framework and it was a breeze! The speed and simplicity were unmatched. Here's the LoRA command I used to kick off training: ``` python lora.py \ --train \ --model 'mistralai/Mistral-7B-Instruct-v0.2' \ -

It’s not enough for AI to recognize emotion... What if it could adapt to it?” You’re upset, but you say, “I’m fine.” A friend senses the disconnect, softens their voice, stays a little longer. Now— What if a machine could do that? We’ve taught AI

The Slowlane is a trap of mediocrity—trading five days of servitude for two days of freedom. People celebrate Friday because it marks a brief escape from a life they don’t control. True wealth isn’t just money; it’s freedom, and the only way out is t

Co-Founder Wanted – Build RUVO, the Emotionally Intelligent AI I’m building RUVO, an emotionally intelligent AI that connects with people on a deeper level—understanding their feelings, helping them heal, and even recreating memories through voices

Everyone’s just memes. How bullshit can this get? Stock market crashes, and instead of thinking straight—it's all attention, attention, attention. Memes everywhere. Jokes everywhere. But no one's talking sense. As a commerce student, I get it—if yo

Another open-source model has arrived, and it’s even better than DeepSeek-V3. The Allen Institute for AI just introduced Tülu 3 (405B) 🐫, a post-training model that is a fine-tune of Llama 3.1 405B, which outperforms DeepSeek V3.