Why you’re underestimating AI

AI has been “coming” for four decades. But today is different. AI is already here, and we are underestimating its value today.

The founder of DeepMind, Mustafa Suleyman provide four reasons for this, for attributes AI has, that makes it so underestimated.

AI is a general-purpose technology. That’s a strong statement! Previous general-purpose technologies were the steam engine, electricity, and the internet. All of them were undervalued at the time, even though they had a clear and visible impact. AI is the same; you’re not sure how to apply it to your specific problem, and yet, it will creep into every single one of them. You just shouldn’t expect it to fit exactly like a special-purpose tool because it is not. Like electricity, you’ll need to build tools on top of it to make it work for every single special purpose.

AI hyper-evolves in the sense that it evolves faster than we’re used to. One reason for this is that it improves itself. The implications of hyper-evolution are that it’s impossible to forecast the progress, the only thing you know is: It’s only gonna get better.

AI has asymmetric impacts. This means that while it is a general purpose technology, the effects aren’t distributed evenly. Just like with the steam engine and electricity, some places are quicker to benefit from it. But that just means, in some place it takes longer for this progress to accumulate. It will happen though.

AI is increasingly autonomous. We’re human, and we’re just not used to 24/7 working autonomous machines. Again, this makes it hard to forecast and understand the value of AI as it is today. But think about it: a 24/7 hour machine that replaces a human at any task, even if it does it 80% worse, is still better because it can do it 24/7. It’s a matter of how you use the tool.

caption...

Created with my https://www.memealchemist.com/ ;)