anton
I think about this paper from last year often lately, the concept is intuitive and simple

Models in the near future will improve themselves...

"We present the Self-Taught Reasoner, which iteratively improves a model’s ability to generate rationales to solve problems" https://t.co/FlAF5Tlxwp
tweet

PromptBase | Prompt Marketplace
RT @nickismeta: I just found out you can buy/sell your prompts for ChatGPT, Midjourney, DALL-E, and Stable Diffusion https://t.co/3LBAZWd9wI
tweet

Robert Scoble
RT @Lynncorrigible: https://t.co/8j52UvujT2
tweet

2000s
paris hilton wearing an anti-paparazzi scarf that obscures flash photography and ruin anyone’s picture https://t.co/tdl0Byk933
tweet

Paul Roetzer
RT @elonmusk: Master Plan 3, the path to a fully sustainable energy future for Earth will be presented on March 1.

The future is bright! https://t.co/11ug0LRlbD
tweet

Robert Scoble
RT @bilawalsidhu: Bringing back that vintage midjourney v3 aesthetic 🤌🏽

This time imagining India 🇮🇳 in a post-apocalyptic sci-fi setting, brought to life with a depth map animation 🔁 Elysium vibes?

🔇 Be sure to unmute for the 🎶 it def adds to the vibe ✨

#midjourney #depth2image #LeiaPix https://t.co/v10swgAahY
tweet

Amjad Masad ⠕
It’s surprising and super unfortunate how vulnerable humans still are to nature. Seems like we should not be at this point in technology.
tweet

Sudharsan ST
Use GPT-3 to generate training samples + Train a few shot model = Zero shot model
tweet