Updated: Apr 20, 2020
In recent years, people from the likes of Stephen Hawking and Elon Musk have publicly expressed their concern about the potential existential risks of "super intelligent artificial intelligence (AI) systems" for us, human beings. In this blog series, I want to share my two main reasons for why we should not be afraid of super intelligent AI or "robots taking over our world" at all. Firstly, human stupidity poses a much more immediate threat to ourselves. Secondly, why should we assume an extremely intelligent entity to be destructive to its environment? In this first part, I will focus on the former reason.
Humans are the biggest threat to themselves
Singularity is used in the context of AI to refer to a hypothetical state of reality in which a system beyond human intelligence exists as a result of uncontrolled and irreversible technological advances. Whenever we hear about this through the media, it is often discussed in the light of its risks to us, human beings. The main one is simply that a super intelligent machine will be able to do things that we cannot understand anymore, which could result in a disadvantage to us. Such a system, smarter than the smartest biological organism we know, would essentially be a big black box to us. Well, spoiler alert: there is another black box that has existed since our evolutionary ancestors - the mammalian brain.
Before I get into that, I want to make a note about our response to hyped stories like "robots taking over the world because of AI". We, human beings, are typically afraid of what we are not familiar with yet. Especially to a layperson's ear, such narratives may sound exciting and scary at the same time. Particularly when they imply that technology will replace our jobs - how would we survive without our way to generate income? As a short-term relief: there are still too many fundamental problems in AI research for us to be worried about actual large-scale replacement in the next decade of that which we all derive a part of our dignity from - the work we do for a living. However, it is important to continuously improve our skills and knowledge to keep up with the pace of science and innovation. Hence, it is in everyone's interest to keep learning new things. Specifically during the current corona crisis, there should be no reason to be bored, as there will always be something you can learn more about.
This is not to say that we should not be worried about the (side) effects of technological advancement at all. I will talk more about this in the second part of this blog. What I first pose to be a more important threat to our existence now is ourselves. Human stupidity, manifested through shortsightedness, brutality, greed, indifference and hypocrisy, especially when observed in those in power, can directly affect human lives today and tomorrow for certain. A striking example is the way rich, developed countries such as the USA and Sweden were not prepared to deal with pandemics in early stages, because governments ignored early warning signs for looming pandemics such as the current corona crisis. Sadly, it requires philanthropists like Bill Gates and Jack Dorsey in the US to help public (healthcare) services cope with the increased pressure on their limited care resources, because the American healthcare system on its own would not be able to.
You can find numerous examples of human stupidity ranging from micro to macro scales throughout history, often with devastating effects for both people and planet. From clicking on suspicious URLs in scam e-mails, foolish civil engineering (e.g. Grenfell tower, Rana plaza) to xenophobic indoctrination leading to international wars. At the root of all this lies the workings of our mammalian brain.
The root of (relative) stupidity: Biases and heuristics
The systematic study of the black box that is our "mind", our brains, is relatively new. It was not until the 19th century when German physician Wilhelm Wundt and American philosopher William James laid the foundation for the research field we know today as psychology or cognitive sciences. Through empirical experiments on human subjects, we try to infer what processes are actually running in people's black boxes - their brains. Along with technological developments in electrophysiology and brain imaging we are slowly gaining more insights in how we perceive, interpret and decide - why we behave the way we do. An essential finding from the past decades of research is how behaviour is shaped by biases and heuristics.
Our behaviour is the outcome of the way our brains process information. The processing capacity of our brains however, is limited. Hence, to cope with the large amount of information we constantly receive from the places we find ourselves in, nature has evolved biases and heuristics to help us more efficiently interpret what is going on. They inform our assumptions about how to process the world. Some of these biases and heuristics are "built-in" - we are simply born with them and everyone has them to certain extents. Think of the way we typically value short-term rewards over potential rewards in the future, termed present bias. Or how we tend to value information that confirm our existing beliefs more than information that disagree with our beliefs, known as confirmation bias. Similarly, we could regard personality traits as biased innate tendencies to respond to situations in a particular way and our predisposition to learn our mother tongue as another biased susceptibility.
Other biases and heuristics are acquired through life experience. By learning from previous mistakes, for instance, we eventually learn how to ride a bicycle. An unconsciously coordinated pattern of body movements. Thus, biases and heuristics can be regarded as unconscious patterns of thoughts about certain situations, certain objects and certain types of people that we encounter. These thoughts inform us what to do - how to behave. In this light, learning is simply the process of processing information in more efficient ways that result in adapted and novel biases and heuristics.
This process already starts from birth when we start to explore the world around us, with our caregivers helping us in doing so by pointing out what is what. What we learn to value, what we consider "good" and "bad" and what we learn to be "normal" and "abnormal" is primed during these precious early years of our lives. This upbringing is informed by the cultural standards and life views that your parents pass on to you - either consciously or unconsciously. As we live on, the biases and heuristics we learnt during the first two to three decades of our lives are left as a seal for who we are and why we behave the way we do. So we mature, naivety and mental flexibility decline.
What is the big deal then about these life developments? We know from statistical research that growing up in certain types of environments yield a higher probability of being mentally and socioeconomically disadvantaged as such children grow into adult life. We also see that children growing up without a father, not necessarily from "bad neighbourhoods", are also more likely to have psychological issues. Only a small proportion of children from such households are resilient enough to grow into perfectly healthy adults in the longer run. A different example is the overly results-oriented and emotionally lacking household that may unconsciously form a seedbed for narcissists. Or growing up in an area without any coloured people that makes white skin the norm and dark skin scary. These sometimes mentally straining environments lead people to develop unhealthy, inappropriate and maladaptive biases and heuristics.
Thus, early life circumstances have a profound effect on the biases and heuristics we develop over time. I go to the extent of saying that these inform the tiniest behaviours up to the factors you take into consideration when you make big decisions in life. From the way you shop: do you tend to maximize or satisfy? To your sense of solidarity: you first or your team first? To your moral standards in making important business decisions: how would you manage the chemical waste from your production sites? Can you combine a full-time working career with raising a child as a woman? Should you use social media data to target your political ads? Should you choose the cheapest or the more expensive sustainable paint for your next building project?
Once you understand how people's behaviours are formed, you also understand why stupidity, or intelligence for that matter, is relative as well. We cannot choose where and to whom we are born - we simply are. Yet, when it comes to those in power at this very moment, their biases and heuristics can have immediate effects on a large scale of people. One ignored fact, one miscommunication between world leaders, one blind spot or one failure in policy execution could literally cost people's lives. This is why humans are their own biggest threat, why they have been so in the past and why it is more immediately devastating than emergent properties of new technologies.
Concluding remarks... for now
So far, we talked about the unquestionable direct effects of human stupidity and its origins in our innate and acquired biases and heuristics. These properties may induce irrational fears of AI and certainly pose more immediate threats to our existence than technological singularity can within the next decade. As long as we remain biological creatures, we can best reduce stupidity by staying open to different perspectives, being aware of and reflect on our own and other's behaviours, and being considerate of all the information accessible to us (including the non-favourable).
An interesting next question is to what level these characteristics of human cognition can be extrapolated to artificial intelligence. One reasoning could compare more and less intelligent examples of the same species to derive what it means to be intelligent. However, this analogy may not necessarily apply when we regard machine intelligence as that of a different type of species. Instead, we may need to follow the comparison of human intelligence to other "lower level" species, such as fish or plants, to look for common signs of intelligent behaviour. In all cases, we should keep in mind that intelligence exists only relative to the entity's objective. I will elaborate on this in my next post.
Stay safe and until then!
I am one
I am all
I am hollow and bigger than myself
I am ever-expanding
I put the universe in a box
and throw it further back in time
Disclaimer: none of my writings are sponsored nor have any other purpose than to share my wonders about the world with you. All of my writings published on syzheng.com are original works. I look forward to opening new discussions and am curious to hear your thoughts. I started the habit of writing a small poem for each blog I post since 2020.