256 pages
english language
Published April 2, 2024 by Random House N.Y..
Ethan Mollick, professeur à Wharton et auteur de la populaire newsletter One Useful Thing Substack, nous livre le guide ultime pour travailler, apprendre et vivre à l’ère de l’IA.
En novembre 2022, une innovation est apparue : la première IA polyvalente capable de se faire passer pour un humain et d’accomplir des tâches créatives et innovantes que seuls les humains pouvaient auparavant réaliser. Ethan Mollick, professeur à Wharton, a immédiatement compris ce que signifiait ChatGPT : après des millions d’années d’autonomie, les humains avaient développé une forme de co-intelligence capable d’améliorer, voire de remplacer, la pensée humaine. Par ses écrits, ses conférences et son enseignement, Mollick est devenu l’un des vulgarisateurs les plus éminents et les plus provocateurs de l’IA, se concentrant sur les aspects pratiques de la manière dont ces nouveaux outils de pensée peuvent transformer notre monde.
Dans Co-Intelligence, Mollick nous invite à interagir avec l’IA …
Ethan Mollick, professeur à Wharton et auteur de la populaire newsletter One Useful Thing Substack, nous livre le guide ultime pour travailler, apprendre et vivre à l’ère de l’IA.
En novembre 2022, une innovation est apparue : la première IA polyvalente capable de se faire passer pour un humain et d’accomplir des tâches créatives et innovantes que seuls les humains pouvaient auparavant réaliser. Ethan Mollick, professeur à Wharton, a immédiatement compris ce que signifiait ChatGPT : après des millions d’années d’autonomie, les humains avaient développé une forme de co-intelligence capable d’améliorer, voire de remplacer, la pensée humaine. Par ses écrits, ses conférences et son enseignement, Mollick est devenu l’un des vulgarisateurs les plus éminents et les plus provocateurs de l’IA, se concentrant sur les aspects pratiques de la manière dont ces nouveaux outils de pensée peuvent transformer notre monde.
Dans Co-Intelligence, Mollick nous invite à interagir avec l’IA en tant que collaborateur, co-enseignant et coach. Il évalue son impact profond sur les entreprises et l’éducation, en s’appuyant sur des dizaines d’exemples d’IA en action. Co-Intelligence montre ce que signifie penser et collaborer avec des machines intelligentes, et pourquoi il est impératif de maîtriser cette compétence.
Mollick nous met au défi d'exploiter l'immense pouvoir de l'IA sans perdre notre identité, d'en tirer des leçons sans nous laisser tromper et d'exploiter ses dons pour créer un avenir meilleur pour l'humanité. Large, stimulant, optimiste et lucide, Co-Intelligence révèle les promesses et le pouvoir de cette nouvelle ère.
I liked the ideas on how to use GenAI but wish he had talked more about the negative consequences, but they were just mentioned and not really considered in any way.
It already feels a little dated but Ethan is legitimately thoughtful when drawing out the good and bad uses of modern LLMs for knowledge work.
It'll only take a couple of hours to read through, very breezy, but not dumbed down or inaccurate
It already feels a little dated but Ethan is legitimately thoughtful when drawing out the good and bad uses of modern LLMs for knowledge work.
It'll only take a couple of hours to read through, very breezy, but not dumbed down or inaccurate
Solid basic intro book. I should have read this when it first came out. I probably would have given it more stars then. It is very short, and touches on a lot of very deep issues, which can be useful for the complete novice, but will be unsatisfying for anyone looking for real answers.
A generally optimistic take on what could be: Generative AI won't destroy education, it will be how everyone will have their own personal tutor. Smart executives won't lay off all of their human workers seeking cost savings but will use the productivity gains of AI to allow their workforce to do so much more than they could have previously.
Mollick covers some history of AI, current capabilities and limitations, and guidelines for working with it. But even a future rushing at us so quickly is difficult to predict, as the author acknowledges.
The book could use an index -- it is pretty hard to go back and find things later. For example, the jagged frontier is an insightful idea -- that it is really hard to know what kinds of things an AI is good at and what kinds of things it is bad at -- but in …
A generally optimistic take on what could be: Generative AI won't destroy education, it will be how everyone will have their own personal tutor. Smart executives won't lay off all of their human workers seeking cost savings but will use the productivity gains of AI to allow their workforce to do so much more than they could have previously.
Mollick covers some history of AI, current capabilities and limitations, and guidelines for working with it. But even a future rushing at us so quickly is difficult to predict, as the author acknowledges.
The book could use an index -- it is pretty hard to go back and find things later. For example, the jagged frontier is an insightful idea -- that it is really hard to know what kinds of things an AI is good at and what kinds of things it is bad at -- but in wanting to refresh my memory of the idea, there's not much to do except scan the pages visually. Maybe I should have gotten an electronic edition.
Mollick gives us a "helicopter view" of generative AI. Readers come away with an understanding of the technology behind ChatGPT and the other large language models. Educators will find this to be a particularly useful read as Mollick describes how AI can help folks to be more effective workers.
Mollick gives us a "helicopter view" of generative AI. Readers come away with an understanding of the technology behind ChatGPT and the other large language models. Educators will find this to be a particularly useful read as Mollick describes how AI can help folks to be more effective workers.
