
Summary
Hoffman uses conversations with GPT-4 to explore how AI could affect education, creative work, journalism and daily work. He examines useful responses alongside mistakes and limitations. The book is an early experiment in working with a language model and an argument for human amplification, rather than proof that its answers are reliable.
Why read it Our view
Use the examples to design small AI trials with a person in control. Compare a model’s draft with the task’s real requirements and check claims that matter. Record where it saves effort and where errors create extra work. Keep those results separate from broad promises about AI.
Written about
- Reid HoffmanAuthored with GPT-4An exploration of AI developed through conversations with GPT-4.Source