1531 sats \ 2 replies \ @BTC_LN 16 Feb 2023 \ on: Are you an AI dystopian or an AI utopian? bitcoin
TL;DR; The road to utopia will be paved with hardships.
The device called calculator killed the profession called calculator. A the same time the calculator also created a lot of new jobs, like data analysis and programming. In the near future we will have an AI system called programmer which will kill the profession called programmer.
But fear not, we will use the programmer AI like we use a calculator Today. Once the programmer AI is here, the job of a programmer will shift the focus of programming work towards more creative and complex problem-solving, rather than mechanical recreation of existing patterns. Today's programmers will become more like conductors of an orchestra, or directors of a movie, providing vision and general directions for a project.
Today, virtually everything is an information problem and AI will be able to help with most if not all. We will be using our new AI tools to unlock new efficiencies in the economy and make new discoveries in most areas of life. AI will allow us take on ambitious projects that currently defy imagination.
For example, we will build satellite factories and farms providing us with all kinds of goods in a safe, sustainable and extremely efficient manner in space. We will recover extinct species of animals and plants or design entirely new ones that will enrich life on earth. We will have personalized healthcare that will focus on prevention and maintenance of good health rather than curing illnesses. We will have immersive education that will make it fun to acquire new knowledge. Imagine a history / art class where you can walk across a realistic 16th century French Paris to get a deep understanding of the budding Renassaince era.
Widespread AI brings change comparable to the industrial and information revolutions. This latest revolution, the intelligence revolution, is just as scary and we have to be vigilant. We have to watch out for negative effects so we can reduce the amount of short term damage and suffering it will inevitably cause in everywhere. We didn't do quite well in this regard during the first two revolutions, hopefully we can do much better this time.
For example, we will have massive job loss in many fields. No job will be safe from the advances in robotics and automation which will be the direct result of the AI revolution. We will need significantly fewer people who currently process data in various capacities. Most government employees and the entire middle management in corporations will be eliminated. We will need less people doing repetitive jobs like driving, waiting, retail and construction work. Even most creative professions are going to be hugely affected.
Heavy reliance on AI systems will also bring bias against certain groups in ways that will be often hard to detect.
But with sufficient care and compassion towards those who are on the losing side, just like the first two technological revolutions, this third one will also bring more long term equality, prosperity, and peace.
In practice this means we have to do many things at once:
-
We have to make sure the public is educated and understands AI systems we have now and where all this is headed. We have to have people dedicated to the subject on every level of society (family, company, town, country). These people should aid others to keep up and plan for the future.
-
We have to actively monitor and help those who need to retrain to find new jobs. Don't give financial aid long term as it will create the wrong incentives and will ultimately destroy social structures. UBI may be a tempting solution, but I worry it's not going to solve the problem. Give a man a fish, and you feed him for a day. Teach a man to fish, and you feed him for a lifetime.
-
We may have to artificially slow down progress so society has time to catch up. This may not be needed right now, but we should be prepared to put breaks on when becomes necessary.
-
We should keep AI advances as democratic, open and decentralized as possible so no one entity gains insurmountable intelligence advantage over everyone else.
-
Do embrace the positive effects of AI and ensure the benefits are spread widely.
bias against certain groups.... do you think this issue would be helped by analysis of realistic differences and directing aid to the proper places or would we be better off continuing to ignore and pretend diversity of skill and aptitude does not exist?
reply
I'm not equipped to solve this problem. It's extremely complex. The difficulty lies in the fact that groups of people sliced in different ways do have real differences of all kinds and AI can uncover them in order to make better predictions. But this also means individuals within those groups may be treated differently.
For example, AI uncovers that people driving red cars suffer more accidents therefore they will raise the insurance for all of them. I drive a red car, but I'm extremely careful and my accident rate may be way lower than the average, yet I will have to suffer from the red car tax. No big issue one may say, as you can simply buy white cars instead. That's true, but what if the red car tax is a hidden tax for a group of people who can be identified by a race, age, gender or trait we should not be biased against? If we don't do the red car tax, we're punishing everyone else for the red car driver accident rates or go out of business. If we do the red car tax, we indirectly force the group of change their choices and behaviours and lose culture bit by bit. Imagine when everything is permeated by AI. There could be massive cultural shifts where groups are marginilized by AI indirectly, even if we try to manually control for direct discrimination by not allowing a religion, political affiliation or other obvious groups to be judged differently.
Not to mention that such manual overrides can backfire in spectacular ways. This happens without AI already. For example, we force hiring from a certain group despite their real world statistical lack of competence. A truly expert hire from this group will suffer, because everyone in the company will be aware of their group's preferential treatment and thus will not trust their real skills. This individual despite their real skill will be marginalized because of this artificial situation, which is unfair because they didn't need the forced hire policy in the first place.
reply