Author(s): Pedro Domingos
Report Published: On Launch
Algorithms are everywhere, all around, quietly dictating to us what we buy, what we watch and what information we have access to. Ultimately, they exist to give us what we think we need to make our life easier and more entertaining. Yet, so few of us could state clearly what algorithms do, how they evolved or how they work.
And, this is where Pedro Domingos’ comes in, by laying down the foundations for a meaningful debate in what is likely to become a key, if not THE, inflexion point for technological singularity (at least in my eyes the more conservative expression of the term, as defined by Robin Hanson, who introduced the concept of a quake view, which evolved into QUAKE Books, and not as futurist as Ray Kurzweil’s).
And what is that inflection point exactly? For Domingos it would be the unveiling of the master algorithm what else, if not the Omega of the current world and the Alpha of the next. Ok I exaggerate his view slightly, but I do so in the tone and style of his witty candid writings, which deserve praise in the way they sculpture the subject.
The Master Algorithm is in theory a book written for everyone. In fact, Domingos explicitly states that the book’s purpose is to open up the floor to the non-initiated so we can all shed light on and contribute to machine learning. As commendable an aim that is, this book is NOT for everyone and I thoroughly recommend that you watch Pedro’s Google Talk before investing in the read because, for most Field Agents, the time needed to do his work justice is a real investment that may just lead you onto a new career trajectory… or not.
“People worry that computers will get too smart and take over the world, but the real problem is that they’re too stupid and they’ve already taken over the world.”
“The Master Algorithm is the last thing we’ll ever have to invent because, once we let it loose, it will go on to invent everything else that can be invented.”
“The power of a theory lies in how much it simplifies our description of the world.”
This book has certainly challenged my notions of what constitutes a “robot”. Apart from that, I was especially intrigued by the history and present direction of machine learning and computer science generally. Being an academic myself, and currently doing my PhD at Pedro’s Alma Mater in Lisbon, I can certainly relate to the epistemological tensions that exist between “opposing” camps, and it is always interesting to note the subtle differences between my field and somebody else’s.
Certain reservations I have concerning the conclusions drawn in The Master Algorithm prevent this book from resonating further. To give an example, I do think it naïve to speculate that should artificial intelligence one day generate mass unemployment then the increased productivity they help create will go on to provide extra resources for those left unemployed. For one thing, critical levels of unemployment in Spain and Greece hasn’t resulted in the richer EU nations giving them more benefits (just read Yanis Varoufakis’ And the Weak Suffer What They Must?) and for another who is to say that the private corporations that own the robots, their patents and so on, will share their increased profits with those now poorer as a result? Surely, they are more likely to pour such profits back into the research and development stages to make even greater gains!
The other issue that wasn’t mentioned but does need to be considered is the increased economic polarity that may occur as richer countries will develop and frankly own more robots. If this activity generates more economic productivity, less affluent countries will fall further behind, especially given that their main “advantage” in a capitalist system is cheaper human labour, which Domingos admits would, in most cases, devalue.
I really do hope that this review encourages some of you to venture out from your comfort zone and into an enlightened participation regarding the role, prospects and ethics of machine learning. Ignorance is NOT bliss for a real learner (robot or otherwise), and this issue is far too important to dismiss as someone else’s problem.
That aside, I really hope Pedro reads this review and explains to me which algorithm(s) was used to create Ava in the futurist Oscar-Winning 2015 film X Machina.