(Image Credit: http://www.penguin.co.uk)

Any debate involving the topic of Artificial Intelligence (AI), in contemporaneous times, more likely than not leads to an intensely splintered outcome. On one side of the debate, stand avowed optimists going to great lengths to extoll the Panglossian prospects of AI. Noted proponents of this notion, such as American computer scientist and futurist Ray Kurzweil, for example even dwell at length, and in all seriousness, about concepts such as Singularity, the advent of which would blur the distinction between man and machine in terms of intellect and attendant faculties. At the other end of the continuum hold forth the very Cassandras of doom. Warning about indiscriminate belief in AI and expending monetary and human capital in an untrammeled fashion in AI research, these pessimists fear about the time when machines would take over mankind and reduce humans to mere lab rats. The author James Barrat, the former theoretical physicist Stephen Hawking and one of the world’s richest individuals, Elon Musk form part of this latter brigade.

So, what exactly is the future of humanity vis-à-vis AI? Are we careening towards a Terminator-like scenario where a Skynet in future would send us into oblivion? Or are these fears an unfortunate figment of an overworked imagination running riot? Noted computer scientist Michael Wooldridge attempts to hack away at the cobwebs of confusion and provide a balanced and nuanced perspective on the theme of AI in his extremely accessible book, The Road to Conscious Machines.

In a painstaking yet, compelling fashion, Wooldridge traces the trajectory of the domain of AI, guiding his readers through the peaks and troughs of developments in AI, AI winters (a period characterised by drought in the funding for AI related research and a lack of confident innovations) and golden ages, before finally concluding in a quasi-philosophical manner about the terrifying prospects of machines lording over men.

At the heart of AI research, the name of Alan Turing stands out like a beacon of hope and ingenuity. This brilliant mathematician, whose life was as tragic as it was productive – he was found dead in his bed after suspected of consuming cyanide following a conviction in March 1952 of “gross indecency”, that is to say, homosexuality, and a 12 month sentencing for hormone “therapy” that would result in chemical castration – was a standout genius who worked at Bletchley Park and assisted in decoding the German Enigma encryptions, thereby paving the way for an Allied victory in World War II. Alan Turing also took it upon himself to solve the Entscheidungsproblem (“decision problem”) posed by the German mathematician David Hilbert. This problem asked whether every question in mathematics can be “decided” – solved with a “yes” or “no” answer. Turing employed theoretical computers, to demonstrate there existed problems for which calculation alone could not provide a solution.

The first public conference on AI was held in 1955 at Dartmouth. Pioneered by John McCarthy (the man who also coined the word Artificial Intelligence), the delegate list included future Nobel laureate John Nash and soon to be stalwarts in the sphere of AI such as Alan Newell, Marvin Minsky, and Herb Simon.

The period between 1956-74 is commonly referred to as the Golden Age of AI. The first serious attempt to build a robot led to the unveiling of SHAKEY, a robot capable of perceiving its environment, understand where it was and what was around it, receive tasks from users, plan means to execute that the concerned tasks before finally proceeding to complete them. However, the Golden Age came to an unfortunate end following the publication of the Lighthill Report. Lucasian Professor of Mathematics at Cambridge University, Joseph Lighthill penned a report expressing disdain for mainstream AI thereby leading to a turning off the funding spigot.

AI made a resurgent comeback a couple of decades following the above ‘Winter.’ IBM’s supercomputer Deep Blue defeated the then reigning Chess world champion Garry Kasparov in 1997. The world of AI attained dizzying proportions following the acquisition of a London based AI firm called DeepMind by Google in 2014. A start-up founded by Demis Hassabis, Shane Legg and Mustafa Suleyman in September 2010, DeepMind copied the way neurons communicate in the brain, in virtual structures called “neural nets. DeepMind has performed some amazing tasks such as recognising images and game-playing. In March 2016, AlphaGo a DeepMind programme beat Lee Sedol—a 9th dan Go player and one of the highest ranked players in the world—with a score of 4–1 in a five-game match. In 2017, an improved version, AlphaGo Zero, defeated AlphaGo 100 games to zero. AlphaGo Zero’s strategies were self-taught. AlphaGo Zero was able to beat its predecessor after just three days with less processing power than AlphaGo; in comparison, the original AlphaGo needed months to learn how to play.

But as Wooldridge explains, even the most sophisticated of these systems remain many orders of magnitude less complex than a human brain. As Wooldridge illustrates, life hides manifold complexities that put to shame the intricacies present in a 19×19 Go grid. While neural networks may have heralded the promise of self-learning systems there is still no comparing artificial neural nets to the structure of the brain.

Mankind – yet – does not have a theory of the mind that would conclusively claim to dissect and decipher the working of the mind. Mind cannot be the outcome of a loose agglomeration or coalescing of handy reductionist theories. This one fact alone is sufficient to provide reassurance that the gloom and doom Terminator scenarios continue to remain urban legends.

Hence, we would do better to address instead some real perniciousness that are unintended consequences of AI – displacement of jobs and dislocations of workforce, deepfakes, sock puppets and a whole swirly assortment of technology induced dangerous propaganda machines.