ChatGPT and other generative ML systems represent a step change in AI use cases, prompting questions on content for pattern generation and tolerances for error
A decade or so ago, systems based on something called ‘machine learning’ started producing really good results in Imagenet, a contest for computer vision researchers. Tweets: @benedictevans , @damianburns , @pkafka , @erickschonfeld , and @risj_oxford Tweets: Benedict Evans / @benedictevans : ChatGPT and the Imagenet moment - I wrote something more about generative ML. Where are the humans in the loop, and what's the point of leverage? https://www.ben-evans.com/... Damian Burns / @damianburns : 'It's a 10-year-old that's read every book in the library and can repeat stuff back to you, but a little garbled and with no idea that Jonathan Swift wasn't actually proposing a new source of income for the poor' Good primer on ChatGPT by @benedictevans https://www.ben-evans.com/... Peter Kafka / @pkafka : Very helpful primer that balances hype, hope, reality. https://twitter.com/... Erick Schonfeld / @erickschonfeld : “ ChatGPT . . . often looks like an undergraduate confidently answering a question for which it didn't attend any lectures. It looks like a confident bullshitter, that can write very convincing nonsense.” https://www.ben-evans.com/... @risj_oxford : 🤖 “These systems can be used to make something original but the originality is in the prompt. But if the advance from chatbots to ChatGPT is in automating answers, can we automate the questions as well? Can we automate the prompt?” writes @benedictevans https://www.ben-evans.com/...