There’s a lot of fear around what will happen with AI, not so much of extinction but rather of two things: fear of losing income, and arguably more importantly, fear of losing identity.
People often are invested in what they do to the point that it’s who they are. That being replaced or eliminated might be a bigger psychological threat than lack of income, at least to those of us fortunate enough to be well off right now.
However, these threats are outweighed by the benefits that AI can eventually bring. Medical advances, power generation, manufacturing capability. Our systems for running society have a lot of problems, economically, politically, epistemologically. These can also be improved with AI assistance.
The real problem is the transition, it’s such a huge shift, and it will happen all at once to everyone, uprooting our idea of the world and our place in it.
What we need is to embrace AI and find a way to make sure that the transition and benefits of AI are distributed instead of concentrated.
For me this looks like the following: companies must commit to retaining some minimum number of employees in every currently existing function, to be determined proportional to their profit taking. This sets a floor on the job losses that can come later when AI really comes on stream.
The justification for this is three fold: firstly, it’s a safety mechanism, it ensures that regardless of the capabilities of an AI system, there are multiple humans working with it to verify its results. If they aren’t verifying diligently, then they’re not doing their jobs.
Secondly, jobs aren’t just a way of making income, they’re wrapped up in identity and meaning for at least some people, and this helps to maintain that existing identity structure across a meaningful cross section of society.
Third, it keeps the economy running, money circulating. You can’t have a market economy without consumers. UBI is one component of this too, but this is both more direct, more useful and more meaningful.