There Are No New Ideas in AI… Only New Datasets
blog.jxmo.io · 1,791 words · saved by 12 readers
LLMs were invented in four major developments... all of which were datasets
There Are No New Ideas in AI… Only New Datasets LLMs were invented in four major developments... all of which were datasets Jack Morris Apr 09, 2025 307 27 33 Share Most people know that AI has made unbelievable progress over the last fifteen years– especially in the last five. It might feel like that progress is *inevitable* – although large paradigm-shift-level breakthroughs are uncommon, we march on anyway through a stream of slow & steady progress. In fact, some researchers have recently declared a “Moore’s Law for AI” where the computer’s ability to do certain things (in this case, certai
saved by
- Anson Yu
- Rajan Agarwal
- Aryan Naik
- s j
- Nhật Nguyễn Thế
- Rikard Saqe
- Ada Ge
- Rohan Kanti
- Devin Plumb
- Ayman
- Aayush Gupta
- SUJASH AGRAWAL
related reading
- The Only Important Technology Is The Internet - Kevin Lukevinlu.ai
- As Rocks May Think | Eric Jangevjang.com
- A Short History of Artificial Intelligenceevery.to
- Most Algorithmic Progress is Data Progressberen.io
- GenAI Handbookgenai-handbook.github.io
- Andrej Karpathy — AGI is still a decade awaydwarkesh.com
- AI in 2025: gestalt — LessWronglesswrong.com
- Will scaling work?dwarkeshpatel.com
- AI Revolution - Transformers and Large Language Models (LLMs)blog.eladgil.com
- The real data wall is billions of years of evolutiondynomight.net
- GitHub - adam-maj/deep-learning: A deep-dive on the entire history of deep-learninggithub.com
- The Extreme Inefficiency of RL for Frontier Models - Toby Ordtobyord.com