Setting up the room…

Skip to content
horasanli.devEnter the room

What if something very good for humanity is bad for the system we've built?

Ya insanlık için çok iyi olan bir şey, kurduğumuz sistem için kötü olursa?

· translated from Turkish

technology
  • #ai
  • #tv-series

In Common Side Effects, Marshall, a mushroom expert, finds a mushroom in Peru called Blue Angel that can cure almost every illness and injury. He wants to use it to give everyone free healthcare. But the drug company Reutical and the government don't want anyone to even hear about it. So Marshall tries to get the mushroom to people while running from them.

In one episode, one of the people at the top of the drug industry, Reutical board member Jonas Backstein, makes the company shut down its project to sell the mushroom. If it reaches everyone, he predicts, the result will be economic collapse, cartels taking control, and war. What makes the mushroom dangerous is that it works too well.

Watching that scene, I thought Blue Angel could be a good analogy for AI. I don't think the show means anything like that. It's a similarity I noticed.

As with Blue Angel, there are two main views on AI. One side is optimistic, expecting productivity to go up and people to be freed from a lot of work. The other side is pessimistic: AI will take our jobs, get out of control, be used in wars and for harm, even bring about the end of humanity.

In my view, AI will take over our work in many areas and do much of what people do faster, cheaper and better. We already see examples of this: in software engineering and tech it does many tasks better than experts, and the same goes for research. As robots are trained in the physical world, AI will reach there too and we'll feel its effects even more. Robots will be doing many of the jobs that need a person to be physically there.

That looks like where we want to get to. But today's economic order will struggle with such a fast shift, because it rests on people producing, earning an income and spending it. If people step back from producing and working, the economy won't do well either. As unemployment rises, other social problems follow.

Like the Blue Angel in Common Side Effects, a tool that solves every kind of problem and every time-consuming task may not be that good for society if it arrives this fast. So I think the right approach is to adopt AI gradually, thinking about the problems it could cause, and to take cybersecurity seriously while we let it develop. A tool that does this much work will have access to many systems, from banks to hospitals. If it's hijacked or misused, the damage spreads just as fast.

I think both the optimists and the pessimists who are afraid have a point. But I think the question we should ask is whether we're ready for a tool this good and this useful.