Most of my thoughts in this blog post come from a book called "AI Superpowers" by Kai-Fu Lee. I encourage you to read it as it's fairly accessible and it contains a ton of knowledge around AI.
While it's true that AI, on a 10 year perspective, has the capability to replace around 40% of the US workforce, I don't think it's something negative. At the end of the day, life isn't about working, it's about living. And having a world based on AI algorithms would allow us to spend more time with our families, to spend more time with our friends,... Some people argue that AI is just like the internet a few decades ago. They think that the job market will simply change, nothing more. I agree with that because yes, the job market will change.
Most jobs will switch to a more human-centric way of doing things, where they will give much more attention to how people feel. These include health experts, teachers, and even judges. The reason being that, there is literally no point for a health expert to take years to learn everything by heart about how a brain works, while a machine can learn that in a fraction of that time. Besides, like I already argued in another blog post, AI algorithms have the advantage of not being impacted by emotions, nor feelings.
An example I used to illustrate this in that previous blog post, is an economist in a bank.
Imagine that you are Goldman Sachs.
Do you prefer to hire an economist, who can be (massively) impacted by his feelings, whose knowledge is limited, and whose decisions are actually mostly made by hormones, genes, and synapses, instead of what we call "himself"?
Or do you prefer to hire an AI, which has no feelings, whose, knowledge is in theory unlimited, and whose decisions are rational, as they are based on actual data?
Logically, Goldman Sachs would hire the AI in this case. What you can read above was my vision of AI until I read AI Superpowers. While I still think that banks will use (and likely already partly use) AI algorithms to determine if you are worthy of a loan, I think that the economist in this situation will not be fired. Why? It's incredibly simple. If you are a business owner, you don't want to talk to an AI when covering your financial situation. You want to see if you can trust the bank, if you can trust them with your money. Let's put that in a table.
|Impacted by emotions||What are emotions?|
|Limited knowledge||Unlimited knowledge (in theory)|
|Decisions based on hormones, genes and synapses||Decisions based on actual data|
|Understands human feelings, can express empathy||No understanding of human feelings, can't express "real" empathy|
Like you can see, AI wins on nearly all of the above aspects, except the last one. As an AI can't express feelings, and can't express empathy, there is no way for an AI algorithm to fully replace humans.
When you are a client, you want empathy. In fact, empathy is one of the key parts of running a successful business. How can you become successful, if you don't care about your customers?
And this is precisely the reason why I think that most of the people constantly saying on social media that AI is going to destroy everyone's lives are completely wrong. Essentially, AI is simply a shift in how society works, nothing more. Besides, we are still decades a way of having the so-called "Artificial General Intelligence" (AGI) which scares so much people.
While it's true that an AI called AlphaGo has already won in a tournament agains the world champion of the game of Go, which is basically an Asian version of chess, the team behind AlphaGo worked during years before their AI became truly good at Go. On top of that, AlphaGo can only play Go. If you ask it what is the capital of France, it won't know the answer. And it needed to play thousands, and thousands of games, before arriving at such a level. Because yes, as a human, an AI needs to learn.
To conclude, I think AI is going to have a tremendous impact on society. It's going to radically change how people live, and how they interact with the digital world. However, it's not going to destroy humanity, at least not yet. And it's not going to cause massive unemployment. AI is simply going to change how our society works.