The idea of artificial intelligence, particularly the
application of artificial intelligence algorithms in the service of human
activities that have always been human based both scare and fascinates us, that
idea being a fecund field to harvest for popular literature and films. We humans, with some of us relying on overdeveloped
imaginations, dreams of AI as the solution for everything that ails us. While for
other humans, those with an overdramatic sense of pessimism, have nightmares
about how human society will be conquered by droids who will eventually destroy
us.
The truth I believe, is somewhere in between. One of the
problems is that the general layperson has no idea what AI entails and how it
works, not how well or how badly it works in real life. The popular media is of
no help as they are wont to lean towards the sensational, in both the
optimistic and pessimistic directions to enable the selling of papers or website
subscriptions.
Fortunately for us, there are people like Hannah Fry, a
mathematician, and a hell of a good writer to explain it to us, if not the nuts
and bolts of AI, then the results of existing experimental results and how the
algorithms are applied to real world problems. To get Prof. Fry’s credentials
out of the way, she is an associate professor in mathematics at the University College
of London. I am a fan of her writing in the New Yorker, as she has a way of
explaining the details and nuances of mathematical topics with great clarity
and ease.
The subtitle of the book is: Being Human in the Age of Algorithms.
It is both somewhat comforting and a touch menacing at the same time. I took it
to be comforting. The title Hello World comes from an example in rudimentary
programming classes, the very first program any neophyte programmer writes are programs
that outputs: ‘Hello World’ onto the screen. I too, have had the excitement of
having those two words present themselves in my computers. So, it is a
welcoming sign, it is also a foreshadowing of what is in store for the reader
in the succeeding pages.
The book is divided simply into nine total chapters; an
Introduction and Conclusion bookends the middle chapters named after seven
distinct parts of the human existence as we know it in the 21st
century. They are: Power, Data, Justice, Medicine, Cars, Crime, and Art. While
the structure of the book is simple enough, the intent of the book is quite ambitious.
Prof. Fry lays out the present and past excursions we humans have made into the
realm of using artificial intelligence to alleviate human based computational
efforts. Some reasoning which drove us to evolving our decision-making advances
along this route involves the perceived and many times a real need for faster
and more accurate computations. The faster part is won handily by computers,
and most of the time the accuracy part is also won by the computers. What
people forget is that first, the computer’s cogitations is only as good as the
data and to a much larger degree, the algorithm that it is given. It is garbage in, and you get garbage out. The
parallel effect is that if you have garbage logic cooked into the algorithm,
then garbage out as well. The more egregious result then is that garbage
analysis and interpretation of the results mean even worse garbage out.
Prof. Fry goes through each of the seven topics and
demonstrates where the human propensity for bias creates disastrous errors in
inference and in computing the wrong numbers or asking the wrong questions. On
the other hand, she also takes great pains to explain why computers are much
better suited to not just doing the computations quickly but to also make decisions
quickly and at times more accurately. One would think that the main arguments in
a book such as this are all along the lines of: it is game over, let the silicon-based
lifeform govern our existence, but that is not the case. Prof. Fry explores and
negotiates the complex and nonlinear landscape of what we humans have done in
experimentation with designing and allowing algorithms to make decisions for us
in order to get at the clearest picture yet of what AI can do for and against
us.
She tells us stories of how Gary Kasparov, chess master, the
very epitome of human decision-making prowess, became seduced by the idea of
the AI, after having been beaten by Big Blue. She tells us about how a self-driving
car is supposed to navigate our highways and byways, but still can not do so
safely. She, most disturbingly, tells us how our government, in their attempts to
simplify and creating accurate decision-making processes had wreaked havoc in
our society, thereby creating equality issues in how justice is dealt out to
us. Indeed, I found the chapters on Data, Justice, Medicine, and Crime the scariest
and the most fascinating because those chapters hit the closest to home. The
idea that our faceless bureaucracy places their trust on unrealistic, biased,
and logically error ridden algorithms to handle our privacy data, decide on long
term guilt and innocence of our fellow humans, cure what ails us, and solve problems
due to human proclivity to trespass on our fellow beings is decidedly unsettling
to say the least.
In every chapter, however, Prof. Fry collects and organizes
the stories in easily digestible and logically intuitive chunks, giving us
cogent arguments for her opinions. In
the Conclusion, she lays out her case, buttressed by the facts and gave me
quite a bit to think about, after of course, educating me on the nuances of the
intricate and logically confounding sequence of action, reaction, and
unintended consequences, which we are not very able to predict a priori.
My belief is that this is a must read for all cogent human
beings who live in todays’ world of technological abundance. We can not live
without fully understanding how decisions are made by algorithms, most
importantly, we need to understand how those decisions can be wrong. In
addition, we must also learn how we can leverage the algorithms so that the computational
tools can be used in conjunction with those areas of cogitation where we human have
an advantage and succeed in creating a
more perfect society.
