🔥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

💭Remember the Stanford Prison Experiment? 📣“I Want to Break Free! Anti-Social Behavior and Persuasion Ability of LLMs in Multi-Agent Settings with Social Hierarchy”, our new pre-print, is out! W/Nicolò Penzo, Itakello, Roberto Dessì, Marco Guerini, Bruno Lepri, jacopo staiano @[email protected] Check this🧵

1-shot

there are only two companies that 1. have founder mode [engaged] 2. have enough money to build god bullish on openai / xai google don’t have 1 anthropic don’t have 2 meta. well, yann still exists.

the education system is fundamentally broken. it's designed to produce factory workers for the industrial age, but we're living in the information age. an age of exponential change. we're still teaching kids to memorize facts, when they have access to all the world's information

I’m constantly struggling with finding the right balance between short term and long term optimisation

Presenting today at #CS2Italy our new work on persuasion and anti-social behaviors of LLMs in multi-agent settings. Joint project with jacopo staiano @[email protected], Bruno Lepri, Itakello, Nicolò Penzo, Marco Guerini, Roberto Dessì MobS Group @ FBK Fondazione Bruno Kessler - FBK Università di Trento

DeepSeek's aha moment: "It underscores the power and beauty of reinforcement learning: rather than explicitly teaching the model on how to solve a problem, we simply provide it with the right incentives, and it autonomously develops advanced problem-solving strategies."

Using OpenAI Operator within Operator has the best error message

We have to take the LLMs to school. When you open any textbook, you'll see three major types of information: 1. Background information / exposition. The meat of the textbook that explains concepts. As you attend over it, your brain is training on that data. This is equivalent

o3-mini might be the best LLM for real-world physics. Prompt: "write a python script of a ball bouncing inside a tesseract"

I asked o3-mini to "create an autonomous snake game, where snakes compete with each other"

being technical is a surprisingly small factor in "feeling the agi" there are graybeard phd computer scientists who've never heard of chatgpt and philosophy dropouts at openai unfakeable curosity + some min iq are ~all you need to see the machine god before others. i love that.

the bubble people in silicon valley live in a bubble. they wake up, drink their mushroom coffee, take their electric bike to a coworking space, and spend the day talking about ai safety, the singularity, and whatever new nootropics they’re microdosing. they believe the biggest

this is sad and pathetic

The "meaning crisis" revolving around AI will turn out to be a complete nothing burger There is no intrinsic sense of meaning to doing something that is economically valuable. It's 100% cultural storytelling around specific circumstances. Humans find ways to tell stories about

Sustainable abundance

the problem with the 4D chess theory is that every insane/irrational decision trump makes could be seen as a 4D chess move - when in reality it’s far more likely he’s just chewing on the pieces

Claude raises the waterline for competence. It handles 99% of general discrete tasks better than the average person - the kind of work that used to require basic professional training but not deep expertise. This creates an interesting split: routine generalist work (basic

I use Windscribe as my main VPN. It's free up to 15GB/month and the only one with inverse split tunneling on macOS. Highly recommended.

I think Cursor is essentially dead. Long live OpenAI and Windsurf.