The Thoughtful Machine

di , Letizia (Autore)

Letizia, 2026

Everyone knows the ChatGPT story. This is the other one. In November 2021, a group of eleven researchers resigned from OpenAI — the world's leading AI laboratory — and founded a company built on a single, radical premise: that the most important thing about artificial intelligence is not how powerful it is, but whether it is safe. That company is Anthropic. That AI is Claude. And that premise is now one of the most consequential intellectual bets in the history of technology. THE THOUGHTFUL MACHINE is the first full narrative account of Anthropic and the search for safe AI — a story of brilliant, anxious researchers, an unprecedented technical approach, and the philosophical questions about consciousness, values, and responsibility that have never mattered more. WHAT THIS BOOK COVERS The founding story: Who left OpenAI, why they left, and what they believed that their colleagues did not. The personalities, the disagreements, and the vision that became Anthropic. Constitutional AI: The groundbreaking method Anthropic invented to teach AI systems to have values — not just follow rules, but understand principles. Why it works differently from every other approach, and why it matters. Claude's character: The extraordinary effort to specify, in writing, what kind of entity Claude should be — its values, its intellectual interests, its sense of humour, its relationship to its own existence. What it reveals about what it means to build a mind. The alignment problem: Why getting AI to do what we actually want — as opposed to what we literally said — is one of the hardest problems in computer science. How Anthropic is approaching it, and what remains unsolved. The human stakes: How AI is reshaping professional work, education, democratic institutions, and the trust we place in the information we receive. What Claude's deployment tells us about what we value and who we are. The global race: How the United States, China, and Europe are pursuing incompatible visions of what safe AI looks like, and what that means for the future. FOR READERS OF Empire of AI by Karen Hao · The Code Breaker by Walter Isaacson · Human Compatible by Stuart Russell · The Alignment Problem by Brian Christian · Genius Makers by Cade Metz If you have wondered what Anthropic is actually doing, why it matters that some AI companies care about safety, or what the difference is between building AI that follows rules and building AI that has values — this book is for you.

• Autore: , Letizia
• Anno edizione: 2026
• Editore: Letizia
• Formato: EPUB2 con DRMFREE
• Lingua: Inglese
• Compatibilità: Tutti i dispositivi (eccetto Kindle)
• Layout: Reflowable
• EAN: 9798233968204