Your Undivided Attention cover image

Synthetic Humanity: AI & What’s At Stake

Your Undivided Attention

CHAPTER

Chat GPT

In 2017, OpenAI discovered this incredibly surprising thing. They trained a neural net to predict the next character of product reviews on Amazon. But what was very surprising is that they found one neuron inside of this neural net that did the very best in the world job of predicting sentiment. In order to do something as seemingly simple as prediction the next character, the AI has to start inferring things about the human. What gender are they? What political leaning are they? Are they feeling positively sentimental or negatively sentimental or positively valenced? That idea is a fundamental one called self-supervised learning.

00:00
Transcript
Play full episode

Remember Everything You Learn from Podcasts

Save insights instantly, chat with episodes, and build lasting knowledge - all powered by AI.
App store bannerPlay store banner