It's Alive!: Artificial Intelligence from the Logic Piano to Killer Robots

by Toby Walsh

There are few other human inventions that are likely to have as large an impact on our lives as machines that can think . . . The steam engine liberated our muscles. The computer is set to liberate our minds.

The development of thinking machines is an adventure as bold and ambitious as any that humans have attempted. And the truth is that Artificial Intelligence is already an indispensable part of our daily lives. Without it, Google wouldn’t find out whatever you need to know. Your smartphone would be . . . just a phone. In countless ways AI has made the world what it is today.

But where will AI technologies take us in the future? We know they will continue to change society, but how? Will AI destroy our jobs? Could it even pose an existential threat? What should we be doing now to prepare for the future?

In this new book, Toby Walsh provides a fascinating survey of Artificial Intelligence for the general reader: where it came from, the state of the art today, and where it will take us tomorrow. His ten predictions of what AI will have achieved by 2050 will surprise you! Walsh discusses how AI will transform our societies, our economies and even ourselves, and what we can do about this.

Genres: Nonfiction, Science, Technology, Artificial Intelligence, Popular Science

320 pages, Paperback

First published August 1, 2017

About the author

Toby Walsh is one of the world’s leading researchers in Artificial Intelligence. He is a Professor of Artificial Intelligence at the University of New South Wales and leads a research group at Data61, Australia’s Centre of Excellence for ICT Research. He has been elected a fellow of the Association for the Advancement of AI for his contributions to AI research, and has won the prestigious Humboldt research award. He has previously held research positions in England, Scotland, France, Germany, Italy, Ireland and Sweden.

Reviews

[Log · Rating 3 out of 5]

3.5

Quite outdated, but it's interesting to get insights on how closely the predictions of researchers play out. LLMs definitely moved much faster than this book predicted (well, he said it could take decades. Does one decade count?). I tested an AI chatbot (using OpenAI) and it could answer all the semantically challenging sample questions posed in this book.
Otherwise, this book has predicted quite closely what's been happening in AI.

Honestly such a depressing field to read about, even when looking at all the positives.

The book was also personally depressing - I feel like boo boo the fool for just living my life as a JS developer (outdated...withering...dying off), instead of doing actually interesting research and development in emerging tech.

I did go on a tangent looking into speedcubing. The current world record is a 7 year old who solved a 3x3 cube in 3.05 seconds?? Can't even fathom moving that fast, let alone thinking. The video is legit mesmerising. Good to know kids now are built different and nerdier than ever.

[Richard Farnworth · Rating 5 out of 5]

Very good overview of AI's past, future and ethics. Occasionally felt like the writer had a bone to pick as he meticulously explains why a particular survey or paper is wrong. This sometimes leads to strange choices of subject e.g. the section on the future of vocations included watch repairer but not doctor, police officer or nurse. But overall a great read.

[Bruce McNair · Rating 4 out of 5]

This is a non-fiction book that provides a good background about the history, current status, and potential future of artificial intelligence. It is an interesting tour of all things artificial intelligence. I am looking forward to read the sequel, 2062: The World That AI Made. I gave the book 4 stars.

[Al Bità · Rating 2 out of 5]

Over the past 50 years or so, digital technology has subtly insinuated itself into our lives — to the extent that we may not even be aware of it. More seductively, it has so entrenched itself that we have become dependent on it for just about everything, and in general we love its charms and alluring promises of making things easier and faster by putting us in greater control over our busy lives than ever before.

This is achieved through the use of incredibly ingenious and extensive programming codes to process algorithms at electronic speeds to arrive at some required conclusion, judgement, product or activity we might desire. This process has been continuously refined and improved on, and these desirable outcomes have been able to be reconfigured and included within the overall programs we use. The same applies to special programs which can examine the previous results of our electronic activities, which then can be understood to have a predictive quality to them, and which can then be used to further refine and adjust as required — and the process repeats itself, potentially endlessly. We are so delighted wth our success at this activity that we have called it “Artificial Intelligence” (AI).

Proponents of AI are highly enthusiastic at the potentially unlimited applications for its use. Walsh is one such enthusiast, and his book is essentially advocating for its many real advantages which can be used for the betterment of humanity worldwide. Of course, Walsh is also very aware that, like any tool, AI can be used both for good and for evil. His book lists some of the “problems” that could arise societally, economically, and so forth with simply allowing for unfettered AI to be permitted free range. He accepts that we should make sure that we only permit it to be used for good, and specifically to be used by humans for humans. To this end he believes that appropriate government interference and control will be required; and that within these constraints, any further problems that might arise will be more than adequately addressed and fixed by more AI.

This approach strikes me as being rather naive and dangerous. Given the state of the political world at present, for example, it is naive to think that world governments will agree as to what is “best” for their subjects. If anything, each capable organisation could very well establish their own AI teams in direct competition with others, in the hope of subverting too much influence. In that scenario, AI will find itself at war with itself, possibly multiplied by new entrants in the field. Humans caught within opposing camps will, I would suggest, find themselves in the wrong location at the wrong time… Joblessness, homelessness, economic downturns, health concerns, terrorism, anarchy, rebelliousness, criminality, etc. will pop up all over the place, all the time, and always in different geographical locations, at different times and seasons of the year. Presumably, for AI to be universally beneficial to all, it would need to be limited to just one, monolithic, global system for all. Good luck with that.

The title of this book also suggests that a more fully advanced AI will be so efficient and effective that a real human being would find it almost impossible to distinguish AI automatons from ordinary human beings, implying that such “robots” or automatons are actually “alive”, raising moral, ethical and other social problems to be somehow dealt with — but there is very little about that in this book. It appears to prefer to be all bright eyed and bushy tailed about just how wonderful a world of AI will be…

At the end of the book Walsh dares to make ten predictions about the future of AI for the year 2050. In my opinion they provide a cross-section of repulsive and/or jejune outcomes, each of which raises more questions than answers: they certainly do not inspire confidence one way or another for the future of humanity.

[Mel Campbell]

This book is written in agreeably plain English and is structured in easy-to-follow chunks, with little summaries at the end of each chapter so you don't get lost. But it helps to have a background knowledge or interest in artificial intelligence, and/or an interest in mathematics, because the author is a scientist, not a sociologist or cultural commentator.

I found my eyes glazing over during the descriptions of how certain programs and algorithms work and was far more interested in the historical, philosophical and cultural aspects, but that is just the way my brain works (or doesn't). Others may find the science stuff gripping.

A lot of the book's content wasn't new to me because I follow media coverage of some of the key developments in AI such as robotics, natural language processing, deep learning and big data, artificial general intelligence and singularity panics, the impact of AI on the future of work, and autonomous lethal weapons (or as we philistines outside the field insist on calling them, 'killer robots').

As an academic authority in the field, Young has the advantage of not being breathless in the way of journos and tech entrepreneurs, but at times he's more dry and sober than I would have liked. But Young's professorial authority is the point of this book: it adds a sense of reassurance, a counterbalance to the often hyperbolic claims made about the present and future capabilities of AI.

I read this book in preparation for interviewing Young at a City of Melbourne event. I was really worried about embarrassing myself with my helpless Terminator fandom in front of an IRL brainiac, but I think it went okay in the end. Anyway, now I shall return to my copy of 'I'll Be Back, Therefore I Am: Terminator and Philosophy'.

[Lippes · Rating 3 out of 5]

A little outdated by now to be honest which in this fast developing field isn't much of a surprise but leads to an awkward feeling at times and made me give up the book

However basic concepts and thoughts are explained real well

[Carmel-by-the-Sea · Rating 4 out of 5]

Sztuczna inteligencja (AI) to jedno z tych dociekań nauki i techniki, którego praktyczna emanacja wprost przekłada się na jakość naszego życie, coraz bardziej nas oplata i uzależnia. „To żyje! Sztuczna inteligencja” profesora AI Toby’ego Walsha jest świetnym przewodnikiem po nadziejach i zagrożeniach związanych z technologiami cyfrowymi. Dawno nie czytałem tak wciągająco napisanej książki popularnej na trudne w sumie tematy, w dodatku podanych bez cienia slangu akademickiego i z świetnie dydaktycznie rozłożonymi akcentami. Bilans cywilizacyjnych zmian, pytania o umysł i świadomość, sfery dominacji AI – to sedno dociekań autora.

Książka klarownie podzielona jest na trzy części. W pierwszej dowiadujemy się o historii formowania się podstaw AI, w drugiej krótko poznajemy aktualny stan jej rozwoju, w trzeciej dostajemy kilka możliwych prognoz i kierunków ewolucji w przyszłości. Ponieważ Walsh zaangażowany jest w badania nad społecznym wymiarem AI, to sporo miejsca poświęcił na ciekawe dywagacje socjologicznych konsekwencji interakcji człowieka z światem cyfrowym. Przytacza garść świetnych przykładów na poparcie stawianych tez. Pokazuje zarówno zdumiewające osiągnięcia AI, jak i jej wielkie wpadki, niespełnione oczekiwania.

Zupełny laik, który chciałby poznać istotę testu Turinga, ograniczenia obliczalności czy główne gałęzie AI – będzie zachwycony. Jaśniej chyba nie można! Walsh omawiając przykłady systemów uczących się uświadamia nam, że właściwie niemal ‘od zera’ AI może opanować bardzo zaawansowane umiejętności – analizę gier, tłumaczenie tekstu czy rozpoznawanie mowy. Jej zdumiewające osiągnięcia sugerują wręcz, że może już dysponować twórczymi możliwościami w sferach kultury zarezerwowanych dotychczas dla człowieka. Coraz trudniej przychodzi nam utwierdzać się w przekonaniu, że ‘umysł’ może być domeną wyłącznie wyewoluowanych biologicznie mózgów. Systemy uczące się, dzięki ogromnej bazie wiedzy i szybkości obliczeniowej, symulują złożone zjawiska. Ostatnio (w 2016 roku) AI udało się zdać test Turinga. Przytoczone przykłady rozmów człowieka z AI dają próbkę możliwości komputerów XXI wieku, które muszą zdumiewać.

Autor jest umiarkowanym optymistą co do przyszłości i własne rozważania puentuj stwierdzeniem, że AI jest nadzieją i jednocześnie zagrożeniem. Przykłady, które przytacza dla poparcia tych tez, przekonały mnie jasnością i bezstronnością. Najważniejsze, to wiedzieć czego oczekujemy i starać się minimalizować negatywne konsekwencje przemian społecznych, które AI aktywuje. Walsh pokazuje, że przed nami sporo pracy w sferze społecznej, psychologicznej czy prawnej. Sam techniczny postęp w pracach nad pojazdami bez kierowców stawia liczne formalne wymagania natury moralnej, zaś autonomiczne systemy wojskowe to przykład szarej strefy wymagającej kontroli.

Dużo uwagi profesor poświęcił na przedyskutowanie zmian w zawodach, które bądź się pojawią, bądź znikną w wyniku rozwoju AI. Część argumentacji w tym temacie jest spekulatywna, ale nadchodzące zmiany naszego stylu życia to fakt - nieuchronność wielopoziomowych rewolucji od biotechnologii po interaktywne rozrywki świata wirtualnego odcisną nieusuwalne piętno na naszym gatunku.
Ogromną zaletą książki jest jej aktualność. Sporo przykładów rozwiązań informatycznych, ustaleń potentatów rynku AI, ma wprost przełożenie na naszą codzienność. Dużo wysiłku Walsh włożył w krytykę wizji przyszłości z ‘osobliwością’ (promowaną przez Kurzweila), czyli wieszczonym przez niektórych nagłym przyspieszeniem możliwości AI, przejęciem kontroli nad systemami informatycznymi i marginalizacją (albo likwidacją) Homo sapiens. Takie katastroficzne wizje, według autora, są raczej mało prawdopodobne w porównaniu z innymi zagrożeniami dla cywilizacji (klimatycznymi, epidemiologicznymi, itd.).

Podsumowując. Świetna książka dla każdego. To po prostu trzeba przeczytać, nim świat bitów nam umknie, albo stanie się zupełnie niezrozumiały w swej nieprzewidywalności.

BARDZO DOBRA - 8/10

[majoringinliterature · Rating 5 out of 5]

Although I'm sure there's a lot of people who are fascinated by the many nuances of the field of Artificial Intelligence - the myriad applications of this new technology, the ethical issues, the many ways in which AI has already begun infiltrating our lives in subtle ways - I think I'm not wrong when I say that most people will pick up this book with one simple question on their minds: viz., what is the likelihood that we will be wiped out by evil killer robots within the next ten to fifty years; and, if we are, will they at least be sexy killer robots à la Battlestar Galactica, or will they be more like Arnie in Terminator (i.e., terrifying)?

This book does discuss the possibility of killer robots, of course. Its conclusions on the topic will not leave the technologically paranoid quaking in their boots, but the more sensitive reader may nevertheless finish this book with a slight sense of diffuse dread, because Walsh makes clear that even though robots probably won't want to kill us out of hatred or spite (yay), they might do it... well, accidentally (erm... yikes?). In the sense that their programming may simply dictate that killing humans is the logical thing to do. Or we may simply be in the way when a robot begins trying to fulfill the function it was programmed to do (Walsh uses the classic example of a robot programmed to make paperclips, who promptly destroys the entire Earth converting every available resource into materials that can be used to make more paperclips. Which is great news for paperclip salesmen, although their market will probably be severely diminished on account of the robot having killed everybody, so maybe it's not such great news, on the whole).

So this book does tackle the big questions. But it also outlines a whole host of AI applications that are incredibly interesting, as well as outlining the ways in which AI has already begun to be implemented in our towns and cities - there's a significant discussion of self-driving cars, for instance, which I had no idea were being tested to such an extent.

This book is definitely layman-friendly. There are a few daunting formulae thrown in, but it's easy to read and gives a really good overview of the history of AI. Walsh steers clear of making too many concrete predictions, but he does give indications of the directions in which AI research may be going. One of the things that I would have been very interested to read about would be the ethical issues, but although he touches on this, there's not a huge discussion, mostly because Walsh is more interested in the issue of whether we will actually ever get to the point where our technology might develop consciousness and self-awareness.

His biggest argument, however, is that we are hurtling towards the future faster than our laws and social norms can catch up. Walsh's argument that AI may arrive before we, as a society, are politically, legally, and culturally prepared for it is perhaps more worrying that the vague possibility that our creations might one day turn on us and convert us into office stationery. As a result of this preoccupation, this book, despite outlining the possibilities of a technology that has not yet been fully developed, is very much a book for the present.

[Rebecca Fung · Rating 5 out of 5]

I've read a few books on the AI/robots theme and this was one of the more accessible to the ordinary person. I really liked how it set out its ideas in very plain English and tried to relate them to the world we live in today. I found the headings very logical and simple, and the examples interesting.

I saw Toby Walsh on a panel at the Sydney Writers Festival discussing technological futures. I have no doubt he knows what he's talking about but also that there are very different perspectives on this subject (it was a great panel of diverse views). I was very interested in Walsh's book and found his book explained some of his ideas quite well - but I'm aware that there are plenty of other ideas about the future of AI as well.

This was a great book that balanced an interest in the possibilities for AI with many concerns for it. It's hard to know what the future of AI is. As Walsh says, we're very bad at predicting the future - but then he goes on to attempt to predict some of it anyway. We tend to overestimate what we can accomplish in the short term.

We also are never quite sure what competing technologies are going to "win" and often we think something that is going to be a hit isn't and vice versa. Look at Back to the Future with all those hoverboards and flying cars ... on the other hand, they still had fax machines. I remember over ten years ago people thought radio would be completely replaced by the internet very soon (it hasn't been, there is plenty of traditional broadcasting out there) and there were constant ads for internet fridges which somehow never took off.

But now we are obsessed with driverless cars - will they or won't they take off? And how soon?

Important, though, is thinking about regulation and that we are at a point now where we can make many of our decisions before AI is making them for us - so we should ensure they are good ones.

[Anna Baillie-Karas · Rating 5 out of 5]

My friend's autonomous vacuum 'Robo' recently ran into an artwork she'd left on the floor and broke the glass. We know that robots will get better at vacuuming, but will they ever be able to create a work of art? This is one of the many questions raised in this great book about artificial intelligence.

(The answer is that machines can already write poetry and make music, so there is no reason they cannot learn to be more creative. We don't know how conscious they can become though. My brain starts to hurt when I think about this).

Artificial intelligence & the possibility of 'thinking machines' is fascinating - & happening faster than we think (eg autonomous cars).

I tried a couple of books about technology for the Read Harder challenge which were not for me - the author of one was an arrogant schmuk and the other was informative too dry.

Thank goodness for Toby Walsh! He cuts through the hype to explore the likely advances, the benefits & dangers - in particular he warns against autonomous weapons, or 'killer robots', in war. Wonderful engaging style, clear writing & he brings expertise & thoughtfulness to the topic. Recommended.

I think it has an alternative title in the US 'Machines that Think.'

#readharder

[PeterBlackCoach · Rating 5 out of 5]

Great book by a real expert in the field of machine learning and artificial intelligence - covering the past, present and future applications and implications and the need to ensure we do not lose our humanity. It is a terrific introduction to AI and provides some context as to this subject area that we cannot escape.

[Katerina · Rating 4 out of 5]

4.5 stars

Brilliant. Loved it. Fascinating.

[Daniel · Rating 4 out of 5]

Pretty interesting layman's overview of the AI field. If you already have more in depth knowledge of the field you'll find it a bit shallow, but with interesting philosophical points about the field.

[Michael Lawrence · Rating 1 out of 5]

This book was narrow minded. A book called "it's alive" should surely be about the future of living sentient machines? which will emerge from from A-life. Not .ai.

little to no mention of A-life. it should be called 'brief history of logic and automation'. The book is a waste of tree given it's just wiki content. huge text, massive spacing. I got through 100 pages in about an hour.

People don't care about .ai. they care about A-life. They just don't know the terminology. This book mentions that Turing delved into shapes generated by chemistry. but only for half a paragraph. the rest of the book covers automation.

A typical. start from the beginning with Aristotle and logic up to Google today computing book with not much else going on. It's not alive. It's dead.