Being human in the age of Artificial Intelligence
Liv 3.0 Att vara människa I den artificiella intelligensens tid. Volante, Stockholm 2018
When we talk about artificial intelligence (AI), many researchers are weary of longer-viewpoint texts around Artificial General Intelligence (AGI) the stronger form of AI, which would potentially mean that a machine could successfully perform any intellectual task that a human being can (for a critical approach see for example Lee in an earlier blogpost). As we haven’t achieved AGI yet (and don’t know for sure if we ever will), these texts tend to be either highly optimistic and sci-fi spirited in sharing all the wonders of what the world then would be, or equally dystopian in seeing the arrival of AGI as the end of humans.
Tegmark’s text tries to give us viewpoints and approaches so that we could have a more open discussion of what these future opportunities of AI and humanity would be, what we would like them to be and what we should try to avoid. His point is that we all have different value sets and hence approach questions of future AI differently, and the ability to affect in which direction we would like the world to go is based on these preferences.
Tegmark starts his book by telling a fictive story of the Prometheus, the first future AGI, built by a team called Omega. Prometheus development goes through different stages; first it learns from performing more well-defined tasks. The team then goes on to develop a media company that produces optimized films and animations for any taste. Through the income of these products the team is able to continue developing virtually all known technological challenges from cancer treatments to better energy solutions. These innovations flood the patent agencies globally and slowly lead to the overtake of most governments forming a new global Alliance, with a joint political agenda.
The story is told as a provocation – could this kind of development happen? Would we like it to happen? Where would we want the story of our future to go?
In order to simplify the question, and also derived from the title of his book, Tegmark defines different stages of life, 1.0, 2.0 and 3.0.
Life 1.0, like bakteria, are born with their hardware and software. Development happens through evolution. Life 2.0 is when we have the same hardware but are able to alter our software. Software in his terms are all the algorithms, all our knowledge, everything we use to process information around us. He simplifies the difference between these two as life 1.0 being simple and biological, and life 2.0 cultural. Humans, currently on level 2 according to Tegmark, are born as babies, that do not know how to speak, read or write. We learn these skills. We can also consciously choose to upgrade our software, and for example learn an additional language. It is up to us if we learn French or Spanish, none of these are pre-programmed in us.
The ability to learn makes life 2.0 superior to life 1.0.
Despite this ability to learn and upgrade, we as humans are still restricted by our biological hardware. Our bodies die, and we cannot spend time in space or under water without extra equipment. Life 3.0 would then be the time when we can alter both our software and our hardware, and become free of the restrictions of evolution. Tegmark calls this the technical stadium.
We do not know what life 3.0 would mean for humans. Science fiction typically provides two scenarios for this; either we would improve our bodies with technology to become cyborgs or just upload our brains to a machine and avoid the distractions of a physical body altogether. Or maybe AGI could be created in a different way, leaving the human out of the equation altogether?
Tegmark questions where the difference is between us and cyborgs if we start by gradually replacing our digestion system, blood or heart, and then over time in the early 2030’s replace skin, skeleton or brain? The body might keep its emotional value to us, but all parts could be renewed. This same point of gradual transformation was lifted in “Are we human?”by Colomina & Wiley that was reviewed in the previous blogpost.
In order to be able to discuss the future we might want to have, Tegmark provides us 12 scenarios; Libertarian Utopia, Benevolent Utopia, Gatekeeper, Protector God, Enslaved God, Conquerors, Descendants, Zookeeper, 1984, Reversion, Self-Destruction.
In the Libertian Utopia humans co-exist in friendly ways with the new technology. The world is divided into three zones. The machine zone consists of large robot-driven factories and autonomous datacenters. The mixed zone is where computers, robots, humans and hybrids thereof coexist. Many humans have upgraded their bodies, and the question of identity becomes increasingly fluid. The thirld zone is for humans only, and in this zone technically improved biological organisms are not allowed, nor are computers with a higher intelligence. The humans in this zone live in an isolated world, using old technologies. In a later chapter Tegmark does a comparison to the Shakers, who choose to live a life with lower levels of technology than the rest of society.
Tegmark also draws this idea of humans continuing with their current level of knowledge in the scenario of Zookeeper, where AI has surpassed humans and run the world, but has chosen to keep a few in zoolike circumstances, where humans continue living their lives, potentially oblivious about the AI’s achievements around them. Some of the scenarios are also purely dystopian, where humans themselves destroy the world, either through biological or technological warfare.
Tegmark’s book might not bring much new information about AI as a technology, but it does initiate discussion of some of the main topics this development puts in front of us. In my opinion, the main topics we should talk about are are ethics, which Tegmark talks about but in a fairly top-level way, and the change of future work and the society it creates. The latter of course also relates to the kind of education we should have in the future, and how we give people the sense of purposefulness and identity in a changed workplace.
The arrival of new AI-based solutions puts in front of us so many ethical questions. As an example, Tegmark talks about warfare with AI. He paints the picture of small, bee-like drones that would be able to identify who they should kill, and shoot a tiny missile through the eye. How can we assure that future AI is only used for the good? The comparison to how the atom warfare was regulated in Kissinger’s time sets a good discussion point: agreeing on usage of new warfare weapons is always better for the currently strong. Chemical warfare could have been more lethal if it wouldn’t have been collectively stigmatized, so international convention is needed.
This is an example of a question where we should consider if there are things that machines are able of doing but that we should stop them from doing. Likewise, we could say the same about humans, should we design our systems in a way that doesn’t make it possible for humans to do certain things? Colomina & Wiley were talking about the human as the only species capable of designing something leading to their own extinction.
So should it be possible for a human to drive a car over someone with the intention to kill? What about an autonomous car? Although Tegmark doesn’t mention it, this comes close to the classical ponderings in ethics and product development on whether we should only, more in the vain of virtue ethics, design solutions we know are good, or should we, in a more consequentialist or utilitarian way, be able to invent anything and then trust the users that they only use the innovations for the good? Ethics will probably be one of the most important fields to do research on in relation to AI, preferably before we have reached AGI. Who’s and what kind of targets do we want our new technology to be built on? And how do we for example avoid biases in our algorithms?
The role of new technology in creating or destroying jobs is also a widely discussed topic. Tegmark quotes Erik BrynjoIf, an academic who has himself published widely on AI related topics, and his book The Digital Athens. The inhabitants of classical Athens could enjoy democracy, art and sport because they had slaves that did all the work. Digital technology similarly drives inequality in three different directions:
- replacing old jobs with jobs that need more knowledge, towards those with a higher education.
- revenues increasingly go to those who own the companies, rather than those who work for them- with increased automation it is increasingly those who own the machines that win
- the digital economy favours Superstars. J. K. Rowling became a billionaire, and first movers’ and inventors like Scott Cook inventing the self-tax- declaration software TurboTax likewise.
It is quite obvious that many of our current jobs are changing through the arrival of better AI -solutions. Technology pessimists predict that all or most of our jobs will disappear. They see us like the two horses discussing what the combustion engine and cars can do to help them. Tegmark notes that some job optimists predict that the next big shift will be into the creative professions, whereas the pessimists claim that creativity is a mental process as well and therefore it too will eventually be mastered by AI. Whichever the solution, we might well end up in a situation where we have many people who do no longer find their place and self-esteem through a classical job.
One solution Tegmark, akin to many other visionaries currently in Silicon Valley, provides to this question is universal income. It is seen as the solution if we do not need a job to get a living. The main question in this discussion has been, in addition to the monetary aspects of added taxation or wealth distribution, what do we as humans need to feel happy?
Finland has been one of the few countries this far were the concept of universal income has actually been tested. The test wasn’t very long and all the results haven’t come in yet but what has already been quite obvious is that the experiment has caused a flood of debate all over the world. Below are the links to just a few of these stories.
I am far from convinced that universal income would actually work, but to me this shows how much we really need these showcasing prototypes in today’s changing word, and how much easier it is to talk about something concrete, like the Finnish experiment in this case, than about an abstract principle. I think there are many more of the potential societal implications that would benefit from a more open prototyping and discussion. It takes time to get big global solutions in place so starting small and light might well be a way forward.
If the future of work and the society around us changes, it will also change the kind of education that we would need in the future. Tegmark ponders about what advice we should give our children to educate for the future and what jobs should they aim for? His answer is jobs which machines are bad at, and which won’t become automated in the near future. He gives our children questions to ask:
- does the job require that one interacts with humans and use social intelligence?
- does it include creativity and need you to find smart solutions?
- does it require you to work in an unpredictable environment?
The more of these questions you would answer “yes” to, the better your career choice.
In addition to everyday experiments on near-time societal implications we also need a longer view of our potential future, in order to react and prepare for it. The future doesn’t just “appear”, we choose how we want to develop it. In science, this future imagining is done by those who have the scientific ability connected to a visionary mind, but not all scientists (in fact, rather few) have this ability or will to vision far and beyond what we can achieve today.
In design, to be able to show what the potential future could be, is core. William Morris imagined the (not always so positive) future of industrialism and utopian socialism in the mid 1800’s, The Bauhaus Manifesto from 1919 wanted to create a future (modern) society and Le Corbusier cities and buildings as “living machines” in the 1920s, just to mention a few. If you are interested in this discourse, a good discussion of the development up to the year 2000 can be found by Victor Margolin in the chapter “The Politics of the Artificial” (pages 106-123 in his book with the same name from 2002). Since 2000, this future driven design debate has only expanded, and taken many forms. Areas such as speculative design (see for example Speculative Everything. Design, Fiction and Social Dreamingby Anthony Dunne and Fiona Raby from 2013), have become increasingly popular. The discussion of what our future should be, how we should design for it, and what the ethical considerations thereof is, is now everywhere.
A good example of this is the Milan Triennal. Founded in 1923 as an every-three-years exhibition on architecture and industrial design it was for a long time the place where nations put forward the most dazzling design objects they could produce. For Finland, the years of 1951 and 1954, when our pavilion was designed by Tapio Wirkkala, have by many been seen as the epitome of Finnish Modern Design and the Grand Prizes given there as the pinnacle of good design. Since then the ethos of presenting beautiful objects has dwindled, and in 2019 the Triennal has a very different stance.
In 2019, the theme of the Triennal is Broken Nature: Design Takes on Human Survival. The main exhibition, curated by Paola Antonelli with Ala Tannir, Laura Maeran, and Erica Petrillo, aims to:
“celebrate design’s ability to offer powerful insight into the key issues of our age, moving beyond pious deference and inconclusive anxiety. By turning its attention to human existence and persistence, the Triennale will promote the importance of creative practices in surveying our species’ bonds with the complex systems in the world, and designing reparations when necessary, through objects, concepts, and new systems.”
In the introductory text when you enter the Broken Nature exhibition they state: “Broken Nature celebrates the revolutionary power of imagination and ingenuity. Even those who believe that the human species will become extinct at some point in the (near? far?) future, design presents the means to plan a more elegant ending. It can ensure that the next dominant species will remember us with a modicum of respect: as dignified and caring, if not intelligent, beings.”
Each country in the Triennale then takes on the challenge of what we could do, or maybe should, to take on the issue of our currently broken nature and the future question of human survival.
The future opportunities and challenges that AI brings us are huge and right in front of us. Most people agree that we are currently facing a big change; not only technical but also societal and human. We see so many possibilities that this might create and at the same time fear for all the (often ethical) questions that arise. Despite this we still do not know how the future will unfold.
Facing this future shift, I think there are at least two things we can do. Firstly, we need more research, experimentation, discussion, design, visioning and potential worldview-painting to give us all something more substantial to talk about – to get joint images of what the world could be. If we get better material and concepts of what the world can be we also get a higher-level discussion (and maybe even some decision making) on what we would want the world to be in the future.
Secondly, even a brilliant vision doesn’t take us far if we don’t do something about it, or are able to relate to it. This is where I think this book by Tegmark is trying to play a part. It is not really a book about AI, despite its title. The style the book is written in, with this “airport-bestseller-non-academic-boast a little – namedropping” -approach can also be a little annoying at times. However, it does provide us with tools to think about how we relate to these issues, and how we would like to see them unfold. I think this is very timely – we can all choose how we approach our future, and affect the discussion of where the world is going. Or as Colomina & Wiley put it – not to do anything is also a choice – designing neglect, as they formulated it.
How should we then relate to the future and these opportunities and challenges? In the end of his book Tegmark provides his own answer to this question. He again quotes Erik Brynjolfsson, who has spoken about two kinds of optimism. There is the unconditional kind, where you trust that the sun will rise the next morning anyway, but also something he calls “mindful optimism”, which according to him is the expectation that good things will happen if you plan carefully and work hard for them. I tend to agree – no matter what we think about these questions the world will most likely become a better place if we engage in an informed discussion about them, and actively pursue a world and a worldview we would like to see happen – whether it is through our everyday actions like consumption or voting, or through grander scale visioning and new research initiatives.
What do you think – what are the areas we should plan carefully and work hard for – in order to pursue “mindful optimism”?