Name: The Idea That Survives Contact: What the AI Demo Hides and How to Build Past It (The Honest Build) (en Inglés)
SKU: 68980513
Price: 20730.00 CLP
Availability: InStock
Author: McShane, Thomas
ISBN: 9798173283597

portada The Idea That Survives Contact: What the AI Demo Hides and How to Build Past It (The Honest Build) (en Inglés)

The Idea That Survives Contact: What the AI Demo Hides and How to Build Past It (The Honest Build) (en Inglés)

Formato

Libro Físico

Editorial

Independently published

Autor

McShane, Thomas

Idioma

Inglés

N° páginas

Encuadernación

Tapa Blanda

ISBN13

9798173283597

The Idea That Survives Contact: What the AI Demo Hides and How to Build Past It (The Honest Build) (en Inglés)

Mcshane, Thomas (Autor) · Independently published · Tapa Blanda

Libro Nuevo Importado

Envío: 14 a 19 días háb.

$ 41.450$ 20.730

-50%

Costos de importación incluídos en el precio ✅

Libro Nuevo

Quedan más de 100 unidades

$ 20.730

Llega entre el 05 Nov y el 12 Nov a Santiago, Región Metropolitana. Seleccionar ubicación

Agregar a lista de deseos

Reseña del libro "The Idea That Survives Contact: What the AI Demo Hides and How to Build Past It (The Honest Build) (en Inglés)"

The demo always works. The product usually does not. Ninety seconds to produce two pages of confident, professional prose. Forty minutes to establish that one factual claim inside it was wrong. That arithmetic is where this book starts, and it is the arithmetic almost nobody runs before they decide they have a product. This is the first of six short books on building AI systems honestly. It is an argument, not a manual. It assumes you are already building something, that you did not come up through computer science, and that you are not asking permission. You are asking what you have missed. It covers what the model is actually doing when it predicts, why it fabricates through the same mechanism it uses to be right, and why instructing it to stop makes the failure rarer and more dangerous at once. It covers the line between decisions the model can own and decisions your system must, the difference between AI as the product and AI as a finish applied to something that already had to exist, and how to write down an idea in a form that survives week three. The evidence is a real build: Dearbh.io, an appeals system working on medical denial letters, from the original idea to the version that was worth building. Including the parts that did not work. No tutorials. No screenshots that rot. No pitch. If you finish this book and never visit any of the sites in it, you got the whole thing.