Artificial Intelligence, progress and happiness
May 6, 2016
I had the pleasure of participating in Milan’s FI-WARE VIP Bootcamp (4–6th May, 2016), under invitation by EBAN. There, along with several other investors and entrepreneurs, we assisted 15 startups from all across Europe to improve the structure and the presentation of their VC pitches.
AI as a danger
One of the people who actively participated, giving their full support to the event (and as a volunteer, nonetheless) was Federico Travella, founder and CEO of NoviCap, an extremely interesting venture in the “fintech” sector that came up with an innovative way of providing working capital for every kind of business. In discussion with Federico and other participants, the topic of AI was brought up, seen as a contributing factor to increasing unemployment, and through that, to social discontent and political instability.
It’s surely an interesting subject and is connected to the broader issue of technological progress and its social impact.
I think it is clear that a major factor in the advancement of human civilization, the very thing we define as our most important achievement, has been technological progress.
From the crude stone tools of our ancestors ten thousands of years ago, to today’s amazing technical advancements, we have countless tools are our disposal that make our life easier (and even extend it significantly) — while, at the same time, can also make it more dangerous (if they fall into the wrong hands, or are used for the wrong purposes).
The nuclear bombing of Hiroshima and Nagasaki, perhaps the greatest technologically-assisted war crime in the history of humanity, made possible by the advancement of technology, teaches us, most explicitly, how technological progress can become a danger.
Even so, and despite the (always present) dangers, there can be no doubt that human civilization is inherently tied to technological progress. Which makes technological progress not just a necessary evil, but a basic requirement for the very existence of civilization.
21st century Luddites
A phenomenon as old as the Industrial Revolution (or even older), is that of the fear of progress, especially when it comes to its effects on the division of labor. Technological advancements often result in changing the ways in which certain tasks are performed, with the new way being more productive, but also requiring less, or even none, human resources.
We’ve seen, for example, farmers to protest the development of advanced farming tools, fearing that they will make them redundant, and the story has repeated itself with factory workers protesting against automation and factory robots.
The term “Luddite”, which today is leveled against anybody speaks against technological progress has its etymological origins in the struggles of 19th-century British weavers, who, under the leadership of the (mythical) Ned Ludd figure, destroyed the newly deployed mechanical weaving machines to save their jobs.
And while it is easy to laugh at 19th century protests against technology, the same phenomenon is possible to arise again in the near future, this time in white collar and office jobs. Only instead of mechanical weaving machines, it will be Artificial Intelligence that performs the same tasks faster, better, and more efficiently.
A categorisation of professions and tasks
In the following diagram we can see an interesting categorisation of professions and tasks over two dimensions:
1. Whether they concern manual or mental labor.
2. Whether they require repetition or creativity
The above categorization is especially interesting. Manual and repetitive tasks were the first in which people started to be replaced by machines — and mostly concern the classic factory and assembly line type jobs. For many decades, the other three categories remained strictly human affairs.
What we are witnessing today, are machines finally conquering two more categories: first those office jobs that don’t really require any inventiveness, through various software systems, and then the manual jobs that require some (but limited) inventiveness, with the introduction of “smart” robotic systems.
In any case, professions that are mostly based on intellectual capacity (but sometimes physical dexterity too), that require strong inventiveness and creativity, are still quite safe from automation. In fact, my personal belief is that some of them will never be performed better by machines, or at least not in the foreseeable future.
A recent revolutionary development
An impressive development that might point against my prediction above, has been the recent win of DeepMind’s AlphaGo software against the Korean worldwide Go champion, Lee Sedol(DeepMind, btw, an AI startup, was acquired by Google in 2014).
Why is that impressive?
Because the Chinese game of Go is so complex that it’s impossible for any computer to calculate in advance all possible moves in order to pick the best one (something that is still possible, after a little pruning of obviously bad moves, in Chess).
This means that a software playing Go has to “think” similar to the way a human player would, trying to figure which move appears to be better given its playing strategy. In a feat previously relegated to the realm of science-fiction, AlphaGo managed to “teach” itself Go, more or less like a human player would, but at a much larger scale and speed — by playing millions of games and studying countless previous Go matches.
In other words, it did what we would have done, but at CPU speeds — and the results were quite impressive: it managed to beat the world champion in 4 to 1, in a game that requires great mental effort and creativity.
So, if a program that can learn like we do, only faster, is possible, does that mean that we’ll soon be having software systems taking over creative jobs?
I wouldn’t hold my breath.
While the urge to generalize from that one feat is great, I believe that we’re still quite a long way from going from software beating humans in a specific board game (however complicated) to it being able to take over any intellectual / creative task.
That said, it’s absolutely certain that people working in highly intellectual and/or creative jobs, will soon have very powerful tools at their disposal — digital assistants that would help them do their job faster and better.
What about (un)employment?
As we’ve seen, despite all current impressive technological advancements, jobs and professions that call for increased inventiveness and creativity, will remain under human control in the foreseeable future. Even more so for jobs where emotions and empathy play a great role.
The same cannot be said for the kind of job, whether manual or mental, that requires little or no creativity. Software and machines will sooner or later take over those jobs, which opens the question of what will that mean for employment (or its lack thereof).
The example of a startup called SMACC is particularly apt here — they’ve recently received funding to develop their software for automating accounting for SMEs, something that, if successful, will help put millions of accountants (or at least assistant accountants) out of a job.
We live, and will continue to live for the next few decades, in a great transition. And while nobody can deny the impact on unemployment of such transitional periods (when millions of people discover that the thing they’ve studied or worked on for all their lives is suddenly obsolete), nobody can seriously propose slowing down progress as a solution either.
Besides, we’re not that far from the era when “computers” were actually humans, usually women, that were called that way because they were tasked to perform (using pen and paper) thousands of calculations every day for big enterprises, research centers, and similar organisations.
If the fear of automation had prevailed at that time, those human “computers” would still have a job, but you would have not had the chance to read this post on your PC (since it would not exist), or do anything else on the web (which would also not exist). And considered how dull their job must have been, and how many people were freed from it to take over something more challenging and ambitious, you will probably agree that the net result has been a positive one.
Even geniuses have the right to be wrong
Many great men and women, including notable intellectuals like the Stephen Hawking, have expressed their concerns about the advancements in AI, suggesting that such research much be performed with the utmost care, be heavy regulated, or even prohibited altogether.
I could not disagree more.
And I’m not saying that there is no danger, just that the danger is not qualitatively different than the danger from fire, tens of thousands of years ago, from gunpowder, a few centuries back, or from nuclear energy, a few decades back. There is, of course, a quantitative difference, concerning the possible catastrophic impact of new technologies — but that doesn’t mean that society cannot take precautions and sufficiently protect itself.
We have been playing with fire (pun intended) for a long time, and as always advance technology requires and advanced society, and an advanced civilization. That’s the real challenge, then, and that’s what we should be striving to build, instead of trying to stifle technology.
Let me be clearer on how I view the upcoming economic transformation kick-started by advanced AI (a transformation that is already underway, to be precise):
In the modern enterprise, there is a number of relatively trivial tasks that are performed by humans. I refer to tasks such as data entry, invoicing, answering and forwarding calls, and even some more advanced ones, like stock market analysis. All of these tasks, and a whole lot more besides, will soon be performed by programs that are (more or less) “intelligent”. Obviously, this will make a great number of office jobs obsolete.
That’s nothing to be afraid of. In the same way that there’s nothing wrong for a car to be assembled by machines, there’s nothing wrong for eventually having machines perform trivial office tasks (or even not at all trivial tasks, like surgical operations, somewhere further down the line). It just means that those tasks will be performed better, faster and cheaper — and this will benefit everybody.
The difficult transitional period
There’s of course the question of finding new employment for all those people that are today employed to perform such tasks, and who, at least initially (and until the labor market is able to absorb and compensate for each new innovation) will be left without a job — with all the resulting economic, social and psychological issues that will result from that.
Special care should be taken for these people (from re-education programs to financial and psychological support), especially those that are not prepared to start a new and different career in the post-automation business landscape.
That will be even more pressing for people that, whether due to advanced age or limited educational background, will be left, like those 19th-century weavers, on the wrong side of those who’ll benefit from automation.
Numerous solutions have been proposed (from back-to-work programs to “guaranteed minimum income” schemes), and while the optimal solution remains elusive, what’s clear is that it would be a tragic mistake (with enormous social impact) to just leave all those people unaided.
At the same time, our educational system should change drastically in order to prepare the new generations for the upcoming (and, in some ways, already here), labor market.
During our discussion in Milan, Federico asked me “how many secretaries” do I know “with a university degree”, and what do I reckon is their overall percentage in the market.
I said that I know quite a few, and I was surprised to learn that the percentage of secretaries with university education is well over 85%. What does that mean? Several things: for one, the role of a secretary is more demanding now that it has ever been. In most SMEs, the secretary is at the epicenter of many workflows. But, even more importantly, it means that there’s a huge disparity between people studying and available job positions.
Today’s university fails to prepare young people for the new labor market. It’s based on the conditions met by the previous generation — if that. “Information workers”, i.e. office workers that performed some heavily repetitive tasks in front of a PC are already relics of another era. They will soon be replaced by AI-enhanced software systems that will perform the same tasks cheaper, faster and better.
I’m willing to bet that soon enough (perhaps after a few decades) the only line of work in an advanced economy that will still make sense for humans to undertake will be that of Research and Development (whether in a scientific or an entrepreneurial sense) ― that, and anything related to arts and culture (a great reminder that science and entrepreneurialism comprise, along with the arts, the main outlets for human creativity).
Investing in education, research and development
Man’s capacity for innovation and radical thinking, for setting new priorities and for pursuing Truth (whether scientific, artistic or philosophical), Knowledge and Happiness, are characteristics that will take a long time for AI systems to tackle — if they ever will, that is. And those should be the basic priorities for the current and upcoming generations.
We should focus our efforts on further developing our educational system, together with our conception of society’s functioning and purpose, towards what this new reality demands from us: to build citizens with strong theoretical foundations, clear minds, team spirit and great internal motivation towards creativity, adaptivity, and personal development.
It’s only this capacity for adaptiveness and constant re-education that can help us be prepared for the next stage of technological development.
In practical terms, this means that we need to rapidly increase the number of students, researchers, and professionals working in R&D positions.
This, along with a re-evaluation of a well-rounded Education, one that goes beyond the needs of the Industrial Era for human cogs and the needs of the Information Era for hyper-specialized scientists, towards creative spirits and citizens in the ancient Greek sense.
What’s especially interesting is that this necessity will lead to an ever faster technological progress, combining the optimism for our achievements with the concern for the breadth and depth of the dangers that we will face because of a misuse of our new tools.
This makes the need for real education, for a deep and true appreciation of the values of civilization, respect, humanism, and of Life itself, greater than ever.
My message is an optimistic one: the age of boring menial labor is over. In other words, we’re moving from grunt work to real work.
In contrast to the so-called Golden Age of Ancient Athens, in this new Golden Age there will be creators, philosophers, mathematicians, scientists, artists, entrepreneurs, and such, but no slaves.
And knowing how many great works (artistic, intellectual, scientific, philosophical, etc) were achieved in that era, when a relatively small number of people had the luxury to think and create free from menial labor (taking advantage, of course, of its slaves’ work), we can’t help but feel enthusiastic for a future society where such luxury will be within everybody’s reach.
From technological to social progress
Through technological progress we can achieve the abundance of goods and the availability of free time that will allow everybody to feel the joy of creativity, will simultaneously eliminating myriads of issues that plague billions of people.
To achieve this without getting burned by the immense power of our own technical arsenal, we will need to re-establish the fundamental values of our civilization as the most important virtues of our society.
(I like to think that the fact that a lot of these values have strong historical ties in ancient Greece, lends special importance to our little country).
Recently, I happened to read the book, “Well, You Were Killed Early”, by the great writer and freedom fighter, Chronis Missios. Somewhere, Missios writes about the value of freedom, referring to a dialogue he had with one of his companions who taught him about communism:
“In the communist society work will be a game, it will be joy, it will be a sport. People won’t be working to feed themselves, but because of an urge to create and enjoy”.
We should not think too short of Chronis Missios. There was simply no way for him to even imagine that a few decades after he wrote his book, the dream of freeing people from toil and giving them the gift of creativity would be made possible by a hodgepodge of new technologies and business models, with AI playing the most important role in all this.
In this sense, AI offers us the opportunity of an unprecedented creative leap for mankind, and for human civilization in its entirety.
These developments, of course, are not without caveats, and it’s up to the future generations to take advantage of AI and all the other immensely powerful technological tools at their disposal for the general good. Our one and only safety pin will be, once again, Knowledge and Wisdom, evenly distributed throughout society. In the end, freedom, and maybe happiness itself, amounts to the uninhibited access to knowledge for everybody.
From the layman to Stephen Hawking, we will be better of striving to understand and fully leverage Artificial Intelligence, this unprecedented and immensely powerful technical development, instead of being afraid of it.
An Answer to Federico
I can finally answer to Federico that, not only am I not afraid that AI will condemn us in unemployment, social exclusion, and civil unrest, but I am certain that, as soon as a critical mass of people finally understands its potential, we will not only witness new, life-changing technologies, but also new ways of thinking and new theories, that will lead humanity to a better, brighter and happier future, where freedom will finally be made into a concrete reality for everybody.
To recap, the most important question brought forward by the advances in AI, is this:
Do we (as a society) have the right to burden people with tasks that don’t require thinking and creativity, while knowing fully well that we could easily automate them?
My response is clear:
No, we don’t have that right.
In the short and mid-term future, menial labor will be relegated to machines. This will free more people to feel the joy of creativity and increased job satisfaction. In this way, Artificial Intelligence will be a great leap forward for our civilisation.
Dimitris Tsingos is the Founder and Head of Entrepreneurship at Starttech Ventures. He also is a Board Member at EBAN — The European Trade Association for Business Angels, Seed Funds and other Early Stage Market Players, as well as a co-founder at HeBAN — The Hellenic Business Angels Network. Dimitris has been the President at YES for Europe — European Confederation of Young Entrepreneurs, the Founder of the Hellenic Startup Association and the President of the Hellenic Association of Computer Scientists and Engineers. Being an undergrad computer science student, in September 2000 Dimitris co-founded Virtual Trip, one of the first tech startups in Greece.