Not just jobs. AI potentially replaces industries.
When we consider activities by the big guns (Google, FB, Apple, Amazon) and activities by disparate devs/researchers, I feel the general strategy will be more toward big organizations buying out any niche/narrow domain innovation, and not primarily on a global niche AI services market. This means that AI technology that truly tips the scales might come with one company having intellectual property rights over what redefines industry.
‘Safety and convenience’ of AI based offerings are at the initial outset of AI market disruption. What is disturbing is the potential to replace large swathes of jobs and industries. The reason why the world’s manufacturing hubs are mostly in China is because of cost and competitive edge.(All the other industry players are right there too, you can’t survive in the market or have an edge without it.) This cost driven free-market movement over the last 15 odd years sits as a success story (where the money is) and a failure to be hidden in local economies.
Inevitably, when AI costs are competitive in a over a 5–10 year scenario over human roles, we will see a mass migration effect that might only be contained with legislation and policy. If we leave it to free market dynamics then we necessarily will have to go through that route.
There is already a need felt to redefine economic spending and the inclusion of UBI (Universal Basic Income) to supplant job replacement at scale. The only issue is, trying to define and have policies ready to be implemented to counter any spurtive growth advances in AI is still a murky area.
With the advent of AI, our notion of capitalism as it stands might drive itself to destruction, because today’s economic drivers are revenue and spending driven.
Agglomeration(and not distribution ) of wealth is a classic at-scale characteristic of capitalism. The rich will get richer, more than that, now the keepers of this wealth will have to spend, not out of desire, but out of need to churn the wheels of the economy.
The problem with AI is that it is an abstract element (software) that has the potential to replace a non-abstract or material element (human).
Among other techniques, it seeks to mimic biological neural networks which are how our brains work. That means, given a greater than natural functioning brain mass using the same principles of biological neurons, we humans (having ~90 billion biological neurons) won’t be physically able to adapt and compete to a machine sitting with 200 billion synthetic neurons on a specialized task. Whether human consciousness and the defining principles of intelligence is embodied only in functioning neurons is debatable. But specialized functions of neurons congruently replicated by synthetic neurons seem to be comparably equivalent.
Among the many sub-domains that this will create: job automation robots (x-robots, if one would call them that , focusing on ‘x’ domains), cyborgs, augmentation, implanting memories, etc
There will be potentially tough competition in specific domains, at one point, AI might be able to see better, hear better and judge on statistical data better than gut feel.
General AI at a human level is not so clear. The markers for it are not yet prevalent or visible. However, scalability and growth on the long term seems to be on AI’s side, so we ought to definitely think if this might be mankind’s undoing.
The ‘Son of Man’ in this context might just be the biblical beast.