Alejandro Piad Morffis

If you want to learn more, read the full article: blog.apiad.net/p/why-artifici…

But this is not all. NNs are also, in practice, extremely well-suited to take advantage of the immense increase in data and compute availability of the last couple of decades. But that's a story for another thread.

Why are (artificial) neural networks so powerful? No, it's not because they are "biologically-inspired." That has little to do with it. It boils down to: - Their theoretical properties. - Their practical capacity to scale. Let me explain 👇

@svpino Sup? How's it been over here? I see a lot of zealots in your replies, so, business as usual I guess...

@alepiad Hi there you!

AI coding replacing programmers is the modern version of calculators replacing mathematicians.

@svpino Also, you can't compare LLMs to calculators! Calculators don't hallucinate.

@svpino I don't know, Rick. Been a long time since I last used my pocket mathematician. ;)

If you're the kind of nerd (like me) who prefers to chat with an LLM in the terminal instead of a website or app, check this project I'm working on. It's still pretty rough around the edges, but it knows Bash and it can mess with your system in fun and unexpected ways.

Here's a fun little LLM-based project I'm working on: github.com/apiad/lovelaice It's a bot that lives in your terminal and knows a bit of Bash, Python, and Linux. It can run commands and create files and maybe wipe out your hard drive if you're not looking. Take it for a spin?

@plflaw Sorry, been off Twitter for a while. Would you mind sending me an email, would love to discuss it with you. apiad@apiad.net

@alepiad Professor, I came up with a theoretical "stacked" LLM model, and discussed w/ ChatGPT, who thinks I have a PhD... If you have time let me know your thoughts (I used Turing as inspiration in the prompt): chatgpt.com/share/671ab5f1…

👋 Hi! I'm a computer science researcher and full-time professor. 🔥 I write about algorithms, AI, education, philosophy, technical writing, and computer science history. techwriters.info/apiad Introduce yourself and let's connect! 👇

@ogrisinger @YiMaTweets @OpenAI @AnthropicAI @sama @controlai Cool! This is the kind of research I love to read. Thanks for sharing :)

@YiMaTweets @alepiad @OpenAI @AnthropicAI @sama @controlai (EXPERIMENTAL section of README) #RIPrompt now encouraging the generation of internal 'AoT Symbolic Plans' to later follow and guide critical thought and output sections 🤩 for Claude projects riprompt.com

@DamienTeney Haha true, I've been off Twitter for more than a year for that same reason, came back and twitted this and never expected to blow up. I'll probably go back to my quiet blog again, this intensity just isn't worth it.

@alepiad Thanks for the pointer. Realising that you had something to sell kinda put me off TBH.

Apparently LLMs cannot reason for 3 reasons.🤔 Anyhow one defines "reason", all three are wrong. 🧵⬇️ 1. Stochastic Sampling: a model can simply assign most of its predicted probability mass to correct answers.