Iwas reading an article by Ezra Klein and he starts off saying that, a pool from the Quinnipiac University (It began as a localized marketing research class project led by professor Maurice “Mickey” Carroll and grew into a nationally recognized, independent polling institute.) found that 70% of American, believe that AI will be disruptive for the job market and 30% say they are worried for their own jobs. Then he adds different opinions given by leading AI executive like Dario Amodei and Mustafa Suleyman that most of the job for withe collar will be taken over AI in the next near to medium term future. Then after citing OpenAI suggested policy for company and government saying that the work hours should be reduced, he goes on saying that if all of this is true there is nothing preventing human like machine to heavily substitute human. AI doesn’t ask for salary raise, or doesn’t form union, it doesn’t need to sleep, it’s always productive without highs and lows.
What the opinions say are far from the reality.
in 2020 the unemployment rate in march was 4.4% in march 2026 is 4.3%.
Claude code is crazy cool still the demand for software engineer was never so high.
The author of the article goes on and says that, talking with a prominent economist Alex Imas, he believes that we are looking at AI the wrong way.
The answer to any question regarding the future of AI comes from identifying what will become scarce. “Imas Alex”
For most of human history, calories were scarce, and our energy went into finding or growing food. Agriculture came and calories were abundant and goods become scarce, then manufacturing came and goods became cheaper and technical knowledge became scarce and now comes AI. What people are scared off is that knowledge will become a commodity like shirts or apple. So now the question is what do we want AI to not substitute and which services AI will make us want more of.
The author then shows a poetic finding from econometrics. When people get richer they want more human touch: Clothing with an history, service that involves people experience and professionalism, foods with a personal story behind it. When something becomes scarce we want it more the author takes coffee as an example. When coffee machine were invented it was believed that people would not go to coffee shops ever again still coffee shops open every day and people want to have the human interaction and coffee more as an experience.
There is a famous paradox called the “Jevons Paradox” (after a British economist called William Stanley Jevons) in 1865 he was interested in the in Britain use of coal. Then the famous James Watt created the an improved steam engine which would use half the amount of coal need to run a normal engine and people believed that the consumption of coal would have been less and slowly o down. During that years there was a boom on coal consumption because coal was now used for more task then ever before.
A professor at Arizona State University “Maksymov” believes that AI is likely to do the same with knowledge. The use of AI will create so many more task that there will be the need to have more knowledge then ever. The author does an example, he says that during the year he had an editor that had to sleep a therapist that he could talk to for an hour a week and a doctor that would take work to see them and he thought AI would be able to substitute them well it did the opposite it created more question to discuss with the editor it create more topic to talk with the therapist and it gave the willingness of creating more appointments with the doctor because what AI said was concerning.
I know what you are thinking what if AI actually becomes better than the editor at giving and discussing more ideas? well think about it there will be more people writing and the editor becomes writer with more discussions to give to AI what if AI substitute writer well personality I don’t like a story that is written by an AI, like the coffee it becomes more of an experience. Now same thing with analyst AI will give more information will give them more stuff to thing off then analyzing that stuff will be problematic and more analysts should be hired and what if AI can analyze that stuff and give different outcome I need more people to understand and explain this analysis and make sure that they are coherent with the ever changing world and more opinions from different mind I have the better it is. I have more stuff to do and I need more people to take on this multiple tasks.
When he started the podcast he had just himself to look for the content, now he has a team and one would think that the job became easier. The team brings more information, news and ideas that before he could never think or be able to look for which brings him to create much more complex and difficult episodes rather than still have the same kind of episode done faster. And this is reflected in every kind of jobs. Because AI is defiantly a different technology, that previous one and it can fully take on task that before were just complemented by other technologies but still that is the possibility we are least prepared for. Everybody will be displace at work and the economy will need to reshape it’s self and we had Covid that its a good example are and did people die yes did people lose money yes but we are adaptive we find new way of shaping what’s around us and is definitely not going to be easy
He does an example that I found very notable, he compares the “China shock” would happen if we treat AI similar to how the society treated worker from 2000 to 2020. Because the job that was affected by the cheap manufacturing was just a part of the economy the fault was given to the people that were doing this jobs and not to an advancement therefore the help from the government was very little. If taken seriously like Covid the all economy will be very important for the government to protect this jobs and create a different economic system.
To conclude I find this paper in line with other research and I believe that AI will indeed create more prosperity than will create disruption but it will be up to how quickly we has human can adapt. I have some criticism to the paper as for example that like many other is highly speculative and it clearly is more a opinion rather then certain he is clearly optimistic and maybe right but there are some aspect that I find strangely provocative like rich people search for more human interaction well not everyone is rich so that doesn’t hold for the all economy and most of the people that are doing basic jobs still feel threaten by AI because most of the routine jobs that they were doing will be early substitute and another thing is that he does take into account current people that will lose its jobs but doesn’t mention how difficult will be for this people to learn and get different jobs or enter new job markets even because most of this people are old and have a high friction to change and therefore would be very disruptive and this are the same people that are reading the next generation that will grow with the idea that will happen to them too.
Lastly he would say that he took that into account talking about how China and Covid differ and that every technology has had disruption for certain community, but now not everybody hates the internet or the typewriter. I could answer by saying that as he mentioned in the article AI may is not the same technology we faced before and that it could likely take all over an entire economy, in some case scenario, and there are forces like OpenAI and Anthropic trying to reach that scenario, maybe, even involuntarily. Because guided by market competition or AI races. Creating AI agents, that can truly not need human interaction is probably feasible. We are fully at the start of this technological revolution and maybe we should instead of looking at the consequences and react, start doing something to make sure this cases do not have the chance to become real.