Some good information sources on Artificial Intelligence

We are at a very special point in history. You might call it a tipping point at different fronts:

1. Artificial Intelligence (AI) is being applied at massive scale and developing at tremendous speed. Two countries are in a race against the time.
2. Climate change has come to a point of visible impact and a 'climate change polarisation' even exists, not contributing to solutions
3. China is taking over from the US as a 'global power'
4. People have never been so polarised and the wealth gap has never been so high since the 1930s (Dalio, 2019)

In this article I will zooming in on AI. I have been watching some documentaries and reading some books on the topic (see further). I'm interested in it because I believe it will not only drastically change the industries I'm working in, but our society as a whole. And as a European, I am currently mainly an 'observer'.

The difference between the industrial revolution and AI revolution

The difference is that the industrial revolution helped us to overcome our limitations in muscle power. The AI revolution (i.e. 'fourth industrial revolution') will help us overcoming our limitations in cognitive power. Let's start with this great figure of the ''landscape of human competence" by Hans Movarec. The further AI develops, the higher the water rises, and the less 'only humans' can do:

As you can see, playing chess was already 'flooded' in 1997. But that was a programmed 'regular computer algorihm'. The more complex game of Go was flooded in 2016. For China, the latter has been an important trigger to start investing hugely in AI (because as opposed to chess, Go can't be programmed, as Go has more possible combinations than atoms in the universe. It can only be 'learned'). But as you can see, AI is not only disrupting games. It will disrupt a lot of jobs. Another observation is that social competences or 'art' are rather high in the hills or mountains. The World Economic Forum list of 'skills you need to thrive in 2020' contains many of these 'typical human properties'. That red word on the grey mountain ('AI design') indicates a point at which AI can design itself. This is the so-called 'intelligence explosion'. It is not unlikely to occur, but we're still 'quite far' from that. However, the water really rises fast in the last 5 years.

The two superpowers

Here's a great documentary I would definetely recommend you to watch. Came out only in December 2019. It covers 5 main topics:

1. The plans of China
2. The plans of the US
3. The future of work
4. Surveillance capitalism (i.e. businesses relying on your personal data)
5. Surveillance state (i.e. states relying on your personal data)

Here's an important assumption: being economic rivals, China and the US don't exchange structurally on AI developments and are hence driven 'to be the first' in different applications. Second, what is most exciting and will attract money: doing research on AI applications, or on AI regulation? Quoting Tegmark: "Many of the safety problems are so hard that they may take decades to solve". Giving one little example: if your self driving car is crashing, will hit a person and should decide which person to hit: which etic rule will you rely on to decide whom it will be?

Here's an important consequence: therefore, in the absence of a 'global state', also no exchange on threats and regulation will occur. While we know that AI can be used in many ways that are beneficial for society, we have to 'hope' that it will be used wisely in a climate of rivalrly.

A computer beating a human in Go was compared to the 'Sputnik moment' in the cold war. Machine learning scientists would now be the equiavalent of physicists a few decades ago. As expressed by it's president, China wants to be the global leader in AI by 2030. From that documentary I learned that data will be the oil of the 21st century. Given the fact that China is now collecting huge amounts of data, it might be the equivalent to Saudi Arabia in the 20th century. How the government and economy in China are currently organised (i.e. kind of run like a company) creates an interesting ecosystem.

Jobs and society

Here I relate to this excellent article by Ray Dalio, describing some current risks in our society. Populism has not been so strong and the wealth/opportunity gap has not been so large since the 1930s almost 100 years ago. Also similar to that period in time, one world power (China) competes with an existing dominant one (US). While Dalio offers clear solutions, his concern is if leaders will be capable of handling this situation well. For capitalism to work, the pie has to keep growing (i.e. productivity growth). For democracy to work, the pie must be divided better.

AI has huge economic potential for those who understand it and can handle it. Challenge one will be to grow and divide this big pie smartly. Challenge two will be to find a way to handle the massive amounts of jobs that might disappear and keep giving meaning to people's lives. Past revolutions have replaced jobs, but apparently this time the scale might be different and the pace is hard to anticipate. According to the documentary, 50% of the jobs in the US can potentially be displaced in the coming 15 years. And apparently only 20% of the current job loss was caused by offshoring (i.e. producing goods in other countries). I'm really curious how this evolution will be handled by leaders. To date, I have not found any single source that gave me solid suggestions on how to handle the future of jobs and meaningful lives. I'm in contact with a Belgian professor in 'work economy' and apparently his group is preparing a paper on this. Looking forward to reading that.

The Virgin podcast 'future visions' and the documentary and book mentioned gave some good suggestions of where AI can help (e.g. pre-curing deseases such as cancer before they manifest). Well known examples are self-driving and other repetitive tasks. However, what I learned from the documentary is that not necessarily the 'blue collar jobs' (i.e. manual labor) would disappear first. 'White collar jobs' (i.e. desk jobs) and businesses that have high potential to disrupt with AI are for example law firms, bookkeepers and insurance agencies. It makes a bit of sense. The digital and automation revolutions we faced had focused on these 'blue collar jobs'.

Apart from the economy and jobs, privacy is another complex topic. The topics 'surveillance capitalism' and 'surveillance state' are touched upon briefly in the documentary. And combined with that, the potential impact on democracy (for example voting). So AI not only has potential to predict the future. It is definetely impacting the future.

The future

You have global challenges such as climate change and global opportunities such as AI. I always find it interesting to think about how things will work out.

• Without a global policy level for example, climate change for example is a very difficult problem to solve. The problem is so massive that you need serious action. But if you are the only country or region taking drastic measures, it might lead to negative economic impact. This is where we're currently struggling with. Further, climate change cannot be decoupled from economy and society. For example, the current populism puts it lower on the agenda (e.g. 'immigration' is more important than 'climate' in Europe in many people's minds).
• I see parallels with AI: AI has massive opportunities. But if you are the only country or region implementing drastic regulation, it might lead to negative economic impact and slow down AI's development. Further, AI cannot be decoupled from economy and society. AI impacts jobs, the growth of the pie and the wealth/opportunity gap.

This article was more of a braindump than a solid analysis. Writing this down was part of my strategy to digest, so I did not aim to predict the future. I'm however very curious how the world will look like in 2030 in many aspects. I'll be watching.

Recommended reading and watching:

• Life 3.0 - being human in the age of AI, Max Tegmark
• Virgin's podcast 'Future visions' (https://www.virgin.com/in-focus/future-visions)
• Why and how capitalism needs to be reformed, Ray Dalio (LinkedIn article)
• Documentary 'The Age of AI' (December 2019) (YouTube video)