Artificial intelligence and the new normal

So that is it. I have been, like many others following what is the biggest paradigm shift of the past hundred years. In my opinion, the true next big step in human history that is being taken right now, as some people might seem to realize it, but also as many others simply ignore.

In the early 20th century, the invention of the combustion motor changed everything. From the way people moved to the way people generated electricity, to how they worked. Basically, artificial intelligence is here to do the same, with unforeseen consequences.

I have a history of being burnt by technology, so much that when I was 20-25, I started to be very wary of new tech. 3D glasses, 3D printing, virtual reality and augmented reality. I saw all of those things rise and most of them fail. 3D printing is the only one that really stuck out, but we are far from the Star Trek replicators that was promised back then. Virtual reality proved to be nauseating in most cases, not because of the quality of the display which really improved in the past couple of years, but mostly because of biological reasons that we can’t simply solve using technology. As it turns out, the brain uses both signals from the eyes and from the inner ear to create a model of the position of the body in space, and if the signal from the eyes is not “matching” the signals from the inner ear, the brain assumes that you have been poisoned and goes in the simplest reaction: nausea. Augmented reality is somewhat OK, but not that much helpful for the general consumer, and it did find some niche uses… Still far from being a game changer technology.

Now, since last November and the apparition of the like of ChatGPT, Midjourney and Stablediffusion, I saw something entirely different. Those are new technology that got me excited like never before. At the same time, it also got me worried.

For a long time, robots have been lurking at the corner to replace workers. It happened in many industries where simple repetitive tasks could be easily automatized by dumb machines. Singular purpose machines. And back then, when people were wondering if their jobs was future-proof and automatization-proof, the common answer was : “anything creative and intellectual can’t be automatized” Well, it was 10 years ago, and now even those jobs are at risk of being automatized within a few years.

I would even go further and argue that 90% of jobs are at risk of being fully automatized within 5-10 years. The world will soon be taken by storm by these artificial intelligence, and if openAI is correct by artificial general intelligence (AGI) that will perform better than a human brain. Creative jobs and intellectual jobs will just vanish overnight, as most companies will simply use data centres with AI to generate more output they could dream with untrustworthy, lazy people that can only work 5 days a week, 8 hours a day, and demand 5 weeks of vacations…. I mean, if you were a CEO, what would you do? Pay a team of 10 engineers to do a project, or install a general artificial intelligence in your R&D server to do more output, a perfect slave that will do exactly as you tell, and will never complain about overwork, or about co-workers, or about vacations, or whatnot? Optical engineering for instance is something that could be performed by a trained AI. Creating a lens given specs in a prompt like manner is just around the corner, and I expect Zemax, Code V and other lens design programs to implement such a feature in the future versions soon… Now the talent of an optical engineer will be to write those prompts, and a single engineer will be able to replace 10 “old school” engineers easy.

Oh, and even if you are a CEO, I would be scared right now, as you job is very likely to disappear sooner than later, replaced by one of those AGI. Even today a company in China has an artificial intelligence as a CEO that is doing better than a human CEO… It must be because of this threat that suddenly, CEO’s and other executives are signing the ban against research on AI recently. [1]

And this got me wondering. Why would CEOs be scared? What is the true value of work? Why do we work in the first place? Shouldn’t we be happy that we can have custodian artificial intelligence to take care of us, as we live hedonistic lives? I mean we could direct collectively our society to be like the one in the Culture books from Iain M Banks, where planet sized starships were governed by artificial intelligence, while the humans in the ships were simply living as they pleased, all watched over by machines of loving grace.

After all, with AGI governed ships could sail the sea without anyone on board. Harbours could use robotic workers created by companies like Boston dynamics to load things and carry out maintenance on the ships. Construction of those robots could be done almost automatically, with shapes that we could not imagine, but would most likely resemble crabs (see the carcinisation idea, where many times, evolution just leads crustaceans to crab like life forms on earth) and would just go and build everything we need, from skyscrapers to single family houses, to parks, tables, TV, cars, etc… AGI could direct power plants, construct solar plants automatically in the desert, mountain, even in orbit and take care of millions of panels like it is nothing. Energy prices would fall to nothing, and suddenly, we are free…

This is the Utopia from the Culture books. We live and enjoy our freedom from work. Some might work for “fun” but most will just learn, travel, do parties, etc.. No more suffering through a 9-5 to make dubious contribution to society at best and just waiting for the timer to end at worst. But can we do that?

If anything, the ultimate goal of our specie should be to get rid of work so we can enjoy our life without having to fear famine or the cold, while keeping our standards of living stable or improving. If the work is entirely done by AI, this is perfect. This it the end goal of any civilization I assume. We can then take to the stars in our AGI generated and directed spaceships with custodian looking robot crabs doing the heavy work for us.

In my opinion, the biggest issue is the transition from our state right now, where work takes a central place in our lives to the glorious jobless utopia. Because obviously, not every jobs is going to be automatized immediately, some are going to be automatized faster than others. It has been going for years actually, with the automatization of simple tasks, and the workers that are automatized just found jobs elsewhere. Plus it was uneducated workers being replaced, so who cares (this is sarcasm by the way, always hard to transcribe in written form). But what happens when there is a massive amount of jobs being automatized? What happens when it is every work class that gets replaced. Worst, what happens when the robots do better than humans the humans they replaced? What are all these people going to to next? Enjoy their free time while others are still required to work because it is harder to automatize their jobs? Should the people being automatized get paid a Universal Basic Income (UBI) while others “non-trivial” jobs performed by people get paid more?If a lot of people get paid the same, how do we distribute luxuries? If everyone wants a nice golden gate in their 1000 ha property, how does it work? wages and prices were used to create a limit to what people could ask for. If everyone gets the same wage, UBI, how do we keep people from all getting a Porsche? a Jaguar? 100’s of Rolex watches? Should the people hardest to automatized be paid more to keep them working? Should UBI be different from people to people based on education? based on something else? Who gets to have the nice flat with the perfect view over Lake Geneva and the mountain? Not everyone can have that view, there is only one… so to speak.

AI is moving fast, very fast and in my opinion, it is going to replace a lot of work. If nothing is done, we are ripe for dystopian future where corporations will own everything, even the AI workers, you are going to output more than any human could do. Actually, I am sad to see that the statement from openAI from September, which was basically that AI is for everyone’s benefit and to be free (hence the “open” in openAI’s name) quickly became a for profit company once they saw that everyone in the world would be interested in paying for their tech… Who could resist?

So, risky jobs will be done by robots. War will be done by robots as well. I am not afraid of terminator style robots, like some names of AI are [2]. In a way, it is easy to think that AI will copy us given that they are trained on our data. What would people do given undying bodies and super intellect if not trying to kill those who are lesser than them? I believe that AI would bypass these trivial warlike thoughts and be able to think more moves ahead than simply using nukes and guns to solve the earth’s problems.

But if we think about it and act smart, we will have the doors of a new golden era opened for us, where jobs are going to have a different meaning, where status will come from somewhere else than business achievements. And to be fair, I think that deep down, this is why most CEO’s are afraid of AI. Because suddenly, the whole of society gets layered flat and now, how do we gauge our respective values? How can an ex-CEO justifies a big flat with a beautiful mountain view in that new world? At one time, I would have said creative output maybe, but Midjourney is taking care of that as I write those lines…


Sources

[1] https://techcrunch.com/2023/03/28/1100-notable-signatories-just-signed-an-open-letter-asking-all-ai-labs-to-immediately-pause-for-at-least-6-months/

[2] https://time.com/6266923/ai-eliezer-yudkowsky-open-letter-not-enough/


PS: Maybe I need to mention: ChatGPT was not used to write this…