Capability needs a direction.
A faster answer or a more precise system is only part of the question. We also have to decide which results are worth pursuing, who bears the cost, and what would count as evidence that things are improving.
AI: The Struggle for Perfection brings that question into the human conversation around AI. It examines cooperation when trust is incomplete, competition that leaves everyone worse off, and the part citizens can play in decisions about technology and public life.
Questions you will encounter
- When does success at an assigned task make the larger situation worse?
- What changes when cooperation involves more than two parties?
- How do media incentives shape the arguments we spend time on?
- What makes a promise credible when intentions are hard to verify?
- How can ordinary people turn concern into a specific, testable public demand?
For readers who want room to think.
This is accessible nonfiction for the AI community and curious general readers. It makes room for disagreement, defense, unequal responsibility, and uncertainty. It is not a technical manual or a promise that AI can settle human conflict.
The book includes 34 research and historical sources. Its authorship note explains the substantial use of AI under human direction.
Begin the conversation
Choose a task you would like AI to do better. Describe a case where completing that task perfectly could still harm someone. Which part of the goal would you change, and whose experience would you use to judge the result?
Bring a question to your reading group · Practice checking an AI answer · How the books were made
Part of The Human Side of AI
Each book approaches a different human question about AI. You can read them in any order.