Artificial intelligence won’t solve all our problems

We also need communities.

Hello readers! I missed a newsletter this weekend due to general busyness. While I’ll try to write weekly I don’t want to be dogmatic about it. However, my thoughts were provoked so I thought I’d put fingers to keyboard this morning and share them with you. Thanks for reading and, as ever, if you find this interesting please do forward to someone or share on social media so others can subscribe.

Ive been thinking about Artificial Intelligence. Not least because of the news this week about a Google engineer being convinced a language AI they worked on is sentient.

Yesterday I read a fascinating article by Thomas Pueyo on ageing. It’s well worth your time. His broad argument as I saw it is that people today are likely to be in one of two states. We’re either the last generation to live what we think of as normal lifespans, or the first to live radically longer. He’s talking 150 years or more.

There are lots of interesting points in this article, which I note is much more comprehensively researched than my writing here. One point in particular really got me thinking. Pueyo makes the point that if you can hold out until we manage to create Artificial General Intelligence, or AI that is at least as intelligent as humans, then the problem of holding back ageing and extending life is likely to be solved very quickly.

I don’t necessarily disagree with this point, but it got me thinking. I’ve been reading a lot about AI recently. One argument seems to be that if we can create an AI with our human intelligence that at least matches us in this capacity then it will be able to continue to develop the technology to take its intelligence well beyond this. Given the way technology often develops, this improvement could happen exponentially.

It follows that we would quite quickly then have artificial intelligence that is magnitudes more intelligent than we are. We can apply this to our problems and solve them very quickly.

We can apply it to the problems where intelligence is holding us back and solve those very quickly. But is the bottleneck to all our problems intelligence?

I’m not sure that intelligence is the only thing we need to solve many of the big problems we face. In fact, many large global problems such as climate change have multiple capacities needed to address them.

There’s certainly a place for the sort of scientific intelligence you might also use to develop anti-ageing technology. Some new science-based technologies to reduce emissions and address the high carbon levels in the atmosphere would be incredibly useful.

But technologies aren’t usually used in isolation. They need to be adopted, people have to be persuaded to use them, other choices have to be made in order to manage how they interact with our values. ‘Our values’ are not homogenous, not across any one society and certainly not across the globe. There’s different values to understand and reconcile, compromises to be negotiated.

There is a whole social and community, dare I say political, aspect to all of this. Scientific intelligence could get us so far, but as humans we work in groups. We already have some technologies that could help immensely with climate change, but implementing them is as much about communities and social dynamics as it is having the workable science.

Perhaps this is an argument for multiple types of intelligence. Discussions in this area quickly get back to fundamental definitions. I could be talking about ‘social intelligence’, which could be one aspect of ‘general intelligence’ alongside what I have slipped into defining as ‘scientific intelligence’ above.

What I am sure of is that this discussion is revealing of my world view. This is true just as much as I find the argument that AI will solve our problems through technology development reveals a world view that all problems can be broken down into problems of science and technology development.

But for me, my view stands. Artificial Intelligence could be very useful to develop new technologies, but unless it also directly helps us with the social aspects of making the best use of those, then it’s hard for me to imagine enormous leaps in really solving the big problems we face as societies.






Leave a Reply

Your email address will not be published. Required fields are marked *