You Ask, I Answer: More Work for Humans in the Age of AI?

In today's episode, KS asks, it's a very interesting what's going to happen. I think AI simply consuming jobs is way too human, pessimistic way to look at things. At the same time, AI is rising. We're gonna have millions of individual professions having access to AIs, not just companies, and that is a very interesting scenario. I have a gut feeling there might be a lot of work for humans because everything will just get supercharged.

What do you think? Okay, so in the long term, in the long-term horizon, it will be humans plus AI. We will have, we will generate entirely new professions that do not exist today to manage machines, to provide value above and beyond what machines can provide, to provide those human connections. Real simple example. No, did not exist.

Today it exists, and there are people make and bank on it. OnlyFans professional did not exist 10 years ago. You know, today there are people like Ayala who are earning hundreds of thousands of dollars a month. Those are our new professions that are that are coming out of the woodwork. Same is true for things like climate change.

Climate change is going to create entirely new professions because it's a major disruption. My issue with AI is not that it's gonna change or eliminate some professions, because that's inevitable. My issue is the timing, right? In past generations, generally progress has been measured in terms of decades or sometimes even centuries. Right?

Now there are exceptions, and folks have pointed out, having this discussion on threads, folks have pointed out things like, yeah, the car was a real fast innovation, right? You had the 10-year period where horse and buggies went away and cars were the thing. But in a lot of cases, there's a lateral transfer of labor there. The car example in particular, people could learn how to drive cars pretty quickly, and the person required to drive a horse and buggy, and the person required to drive a car, that same human being is still employed, right? They've changed their method of employment.

The analogy I like to use is agriculture to manufacturing in an industry, right? Where you have things like a spinning jenny or a combine that removes hundreds of people from the fields, and you have to do something else with them. Now there's other parts of the process that still need human intervention. Someone has to do QA, uh, like, hey, that's not a corn stalk, that's you know, a piece of wood that's fine, it's going into the combine. But generally, a lot of the people displaced by industrial machinery in farming made their way into industry.

And that happens over generations. If you are hand picking corn, your kid might be operating the combine, or your kid might work in the combine factory. You, the person picking corn, don't have to reskill because the change is slow enough that in life, or if you do have to reskill, it's it's relatively slow. With AI, we are seeing the labor is not getting shifted to another human, to a human in a different job. It's not going from person copying and pasting spreadsheet data to person telling AI to copy and paste the spreadsheet data.

What's happening is that AI is capable of doing the task entirely. And so there's less person time needed. In many cases, dramatically less. And again, this is not a problem in the long term, right? In the span of a hundred years, we could easily adapt.

You know, if I if I'm in my 50s and my kids are in their 20s and they're learning how to use AI, and I don't know how to use it, that's fine. I'm close enough to retirement that I'll finish out my my next 15 years in upper management or whatever, and that's that. And the kids have made already made that jump. But if that happens in a five-year time span, suddenly there's a problem because many people struggle to reskill. Many people struggle to do something substantially different.

There was an effort in the 2010s to reskill coal miners as coders. It didn't go well. It didn't go well for a couple reasons. One, the program was really badly run by the US government, go government. Um, those are radically different skill sets.

Yes, they both have require attention to detail. Yes, they both require perseverance, and yes, they both require determination. But they are radically different skill sets and radically different ways of thinking. And so a lot of the people they were able to get some of the basics down, but it was a hard transition, and many of them did not stick with it. Even though there were promises of things like higher wages.

There's also an issue with wages. So, in a gradual rollout of AI across multiple professions, you would have people have the time to reskill, have the time to adapt generationally. But if it's fast, if in pursuit of higher and higher profits, you're like, we're firing everybody, and machines are gonna run it, and you know, we're gonna have five five PhDs who are chief AI officers, and they're gonna run all the machines. That's not enough time for that those people that you've laid off to adapt to something different. And this is this is likely to happen in a lot of industries.

Not because AI technology is so amazing, I mean it's pretty cool what it can do, but because, at least in the country where I am, I'm in America, in the United States of America, the United States of America is kind of a hyper capitalist nation in that we pursue profits really, really, really assertively. And not necessarily with a long-term perspective. A lot of folks, the way they run their business, they try and make a quick buck and get you know cash out and move on. Or they try to build systems that help them retain their wealth. And wealth isn't bad.

I mean, I certainly would like more wealth, but it's what you do with it and how you manage it that matters. And so a likely outcome for a lot of the use of AI is to reduce cost substantially, reduce headcount substantially within companies, and replace that with machinery. And yes, with people who are highly skilled in the use of AI. That's one of the reasons why so strongly emphasizing that everyone who is capable of doing so should become minimally proficient with AI. You should learn how to use generative AI in particular.

How do you prompt? How do you get good results? How do you get repeatable results out of machines? How do you proactively identify things that the machines can do and you get ahead of your company? You be the pioneer in your company and your organization for the use of the AI so that you are the one in charge of the machines because that is the best insurance in the short term to be one of the leaders within your own organization.

Are there long-term jobs that will come from AI? Yes, absolutely. Um real simple example. There is the possibility to use generative AI, uh, chatbots in particular, to provide mental health care services. Now, that is going to require a lot of supervision.

Could conceivably, it might be a good idea to have it require a doctor's prescription, right? And have a qualified therapist monitoring to make sure the machines are doing what they're supposed to do. Therapists are so hard to come by, at least again in the country where I am, that having machines be sort of the first line, not the uh if you think about it in like almost a customer service aspect, you have that front line of machines that can help people having a bad day, but not being in danger. And then a triage, right? Uh prioritization so that the machines, the language models would identify, hey, this person seems to be having a harder time than normal.

Let's elevate this to a human. And then you bring in your human therapist who comes in and says, Hey, it looks like you're really having a hard time today. Let's talk, right? By providing more access to mental health care, we can maximize the value of the very few people we do have, right? So, and those therapists would be, they would have a lot of work, right?

To your question. Yeah, there might be a lot of work for humans. Yes. In anything that is high risk, there absolutely should be human beings managing the process, overseeing the results, and tuning and training retraining the machines to say, like, hey, that response that you gave there, that was not effective or appropriate. Let's train retrain you this way.

But it's all a question of time. If we can take measures like universal basic income in the short term to bridge that gap for new generations to be able to take on these new positions for those positions to flourish and grow, then I think we'll be fine. If we don't do that, then in the short term, you're going to have some pretty severe economic dislocations. Software development is a really good example. Software development is a field that is highly automatable.

And if you look at the data from Indeed.com, it is that field has not recovered from the pandemic, right? Hiring, the number of open positions has been low and has remained low for years now because machines are capable, it turns out, of being really good software developers. So that's my point of view on this. AI will consume jobs, A will create AI will create jobs, but the pace at which it consumes jobs has to be slowed down long enough for it to also create the new jobs as well. Thanks for asking.

Talk to you next time. If you enjoyed this video, please hit the like button. Subscribe to my channel if you haven't already. And if you want to know when new videos are available, hit the bell button to be notified as soon as new content is live.