[Download kindle] Rebooting AI. Building Artificial Intelligence We Can Trust Author Gary Marcus – Andy-palmer.co.uk

Rebooting AI. Building Artificial Intelligence We Can Trust I ve been waiting for this book for awhile Gary Marcus has always been on the forefront calling out the sensationalism and hype of modern AI trends Many of us have forgotten or were not born when this all happened before in the 1960 s and 1980 s This AI rhetoric and sensationalism has happened before, and was always followed by an AI winter.What Gary and Ernest do well is to not to leave readers without a possible solution Sure, they take a critical appraisal of deep learning and how limited it is in practical use cases But they also offer a path and possible research areas where AI may be better realized However, it is going to be long and hard, and general AI seems unlikely to happen in our lifetimes. Almost all the computer science researchers and programmers unanimously adopt machine learning in general and deep learning in particular as the algorithm that will bring AI and AGI artificial general intelligence This book begs to differ It bravely criticizes the prospects of deep learning to deliver AGI, by presenting cases where this approach fails and suggests some half baked ideas in regards to other approaches such as using Predicate Calculus in Logic in order to implement common sense knowledge Some of the text is repetitive. I currently work as an engineer in the autonomous vehicle space, and I find that this book sheds light on the current state of AI.Basically, I found this book because at work I ran in to a couple of the current fundamental problems of deep learning One, deep learning struggles to adapt with a low number of human labeled examples Two, deep learning also struggles to adapt when the distribution of the input shifts or the task changes.I was poring through recent deep learning research papers attempting to solve these problems, and I found that I was mostly unsatisfied by their results and the directions that they were going.I think, at the time of this writing, AI is often conflated with deep learning, and after reading this book I have a greater appreciation about the broader scope of AI, both in its problem and solution space We have a long way to go in creating truly intelligent systems. I have been waiting for the industry to finally get critical with itself Thanks for this thoughtful treatment of a subject that has had misleading and frothy treatment from the media, unrealistic and misguided expectation setting from software vendors, and too little serious self examination from practitioners The authors have set out a pretty good set of functional requirements for a generalized AI here They ve done a good job of articulating 1 Why the current data driven deep learning cannot progress beyond simple and very narrow tasks, 2 A common sense understanding of the world is missing from these approaches 3 a conceptual faculty able to learn without 10,000 high quality labeled examples beforehand is needed and 4 What potential corrective actions might be taken.Importantly, they have suggested a reunification of the two divergent AI traditions the original, in the Minsky tradition, and the current data driven, tabulated rasa deep learning approach is ultimately what is needed to progress and fix the current course Additionally they implied or I read into it that a trans disciplinary approach is needed, given the stated need for understanding the functionality of the brain, not just at a biological chemical level, but at a fundamental metaphysical philosophical level.Probably the biggest contribution is the authors recipe for achieving common sense, and ultimately general intelligence, too long to quote here but I couldn t agree.The book paints a clear, yet challenging road forward, but as they argue, tough love for this young teenager AI is what is needed if there is any needed hope of it achieving its potential. Two Leaders In The Field Offer A Compelling Analysis Of The Current State Of The Art And Reveal The Steps We Must Take To Achieve A Truly Robust Artificial IntelligenceDespite The Hype Surrounding AI, Creating An Intelligence That Rivals Or Exceeds Human Levels Is Far Complicated Than We Have Been Led To Believe Professors Gary Marcus And Ernest Davis Have Spent Their Careers At The Forefront Of AI Research And Have Witnessed Some Of The Greatest Milestones In The Field, But They Argue That A Computer Beating A Human In Jeopardy Does Not Signal That We Are On The Doorstep Of Fully Autonomous Cars Or Superintelligent Machines The Achievements In The Field Thus Far Have Occurred In Closed Systems With Fixed Sets Of Rules, And These Approaches Are Too Narrow To Achieve Genuine Intelligence The Real World, In Contrast, Is Wildly Complex And Open Ended How Can We Bridge This Gap What Will The Consequences Be When We Do Taking Inspiration From The Human Mind, Marcus And Davis Explain What We Need To Advance AI To The Next Level, And Suggest That If We Are Wise Along The Way, We Won T Need To Worry About A Future Of Machine Overlords If We Focus On Endowing Machines With Common Sense And Deep Understanding, Rather Than Simply Focusing On Statistical Analysis And Gatherine Ever Larger Collections Of Data, We Will Be Able To Create An AI We Can Trustin Our Homes, Our Cars, And Our Doctors Offices Rebooting AI Provides A Lucid, Clear Eyed Assessment Of The Current Science And Offers An Inspiring Vision Of How A New Generation Of AI Can Make Our Lives Better


Leave a Reply

Your email address will not be published. Required fields are marked *