Question: How is AGI different from current AI?
Answer: Current narrow systems are much more domain-specific than AGI. We don’t know what the first AGI will look like, some people think the ［https://www.lesswrong.com/posts/6Hee7w2paEzHsD6mn/collection-of-gpt-3-results GPT-3］ architecture but scaled up a lot may get us there (GPT-3 is a giant prediction model which when trained on a vast amount of text seems to ［https://www.lesswrong.com/posts/D3hP47pZwXNPRByj8/an-102-meta-learning-by-gpt-3-and-a-list-of-full-proposals learn how to learn］ and do ［https://gpt3examples.com/ all sorts of crazy-impressive things］, a related model can ［https://openai.com/blog/dall-e/ generate pictures from text］), some people don’t think scaling this kind of model will get us all the way.