Abhas Gupta argues that retraining displaced workers for new jobs is largely a fantasy. Jobs cluster by skill complexity, and moving someone from retail to molecular modeling isn't realistic. Automation creates new roles, but they'll be high-complexity positions unlikely filled by low-skill workers.
Chapters
The next point, have you guys heard of like retraining, reeducation?
Oh.
So a lot of politicians kind of talk about, they're like, oh, we lost our jobs in Detroit in the car sector, in the manufacturing sector. We need to retrain those workers to go do something else. And essentially, if you're going to take the argument down and say like, I use this idea of center of mass. So I showed these, this is what a doctor does. So you could say that their center of mass is right there for that job. And this is a retail worker, the center of mass is probably right here, right?
Mm-hmm.
You're asking for a center of mass of someone's job who exists here to be retrained up here, because automation's gonna take over all the jobs over here. So can you actually take people from all the jobs that exist here and shift them into the top right? So who thinks that you can? If they need to, to survive, then they will. OK. And who thinks that it's not going to happen? OK. So let's just— whoever thinks that it can happen, tell me why. Go ahead.
I think there will be a lot of new jobs, new kinds of jobs, not like doctor or something like, uh, something else, maybe some developers in their home, and they can just model the molecule of the health and translate the data to anyone else's 3D printer. Yeah, they can print this kind of stuff. There will be a new job that we cannot think about.
Sure.
Imagine nowadays.
I totally agree that those jobs, new jobs, will emerge. But think about the person that's running a cash register. Do you think that person is going to be doing the creative work of building a molecular model, understanding how the model interacts, doing the 3-dimensional—
And the need is actually— will be part of the job in the future. You mean the mouth that's going to eat something.
Demand in the economy? Yeah, yeah. That's a great point. We're going to come to that in a second. Hold on to that thought.
Yeah.
I think in that part, the black— the white part, it's mainly dealing with people's emotion and relationship. So at the end of the day, when machines occupy those parts, those parts can be automated, human beings might shift to those works that deal more with relationships?
Definitely. That's definitely going to happen. So you could— but can you retrain someone to be more empathetic? Maybe, I don't know. But those are the kind of jobs— I agree, they're going to have to shift into those types of jobs. But you're talking about these are graduates, so I think there's an intellectual limitation to go from something that's low complexity to go to something that's high complexity, right? And so when a lot of politicians talk about retraining, I think that that's kind of misguided. I think you're just not going to be able to move people who are down here into this sort of highly complex, highly non-routine world. Yeah, go ahead.
Isn't this a process that's going to take a really long time, though? I mean, it wouldn't be just moving those people there. It's a gradual process. Fields will start getting eliminated or the cashier job. So I don't think it's about retraining, it's about the new education of the future.
We're starting those programs. Yeah, so there's a kind of consistent theme around all these questions that maybe I'm not making very clear, which is there will be new jobs that are created, there are new economies that will emerge, new sectors that will emerge. I don't disagree with that for a second. My point is that the jobs within those new economies and new sectors, they have to be up here. Because if they're anywhere in this area, they're going to be automated.
Right?
Like the companies that are emerging, the startups that are emerging today, they're doing the work of what took 500 people even 10 years ago. And they're doing it with 40 people. Because all the shit work, they're kind of automating. Right? They're using AI to do a lot of this work, or they're using analytics to do a lot of the work. They're using workflow automation to do a lot of the work. Go ahead.
So I agree with you that there's going to be a lot of unemployment. What do you think should be done for all this free time humans are going to have?
Yeah, okay, perfect. So that's the next kind of piece. So we're talking about a world where 50% of the jobs in the country don't exist. New job creation has to come from this top right corner. This is also why I think job retraining is not a viable solution. Let's just try to imagine, I think to your point, so if 50% of the people in this country do not have a job, that means like we can't— in the most recent crisis we had about 11 or 12% was the kind of peak of the unemployment, 13%, and we were straining our finances as a country. Like, we can't handle 50% of people being unemployed. To your point, demand collapses, right? So 60% of demand in the country comes from the bottom 90% of income generators, right? So the country's economy is dependent on people buying things. You cannot have a capitalist society continue to persist in a world where 50% of people are unemployed. So these are really, really big problems. So my kind of closing point before we open up for more questions is that this is going to be, in my mind, the biggest social political issue of the next decade. It is going to start by impacting, I think, countries abroad faster. A lot of what we outsourced thinking that we're looking at wage arbitrage— outsourcing is the first step in the path to automation. A lot of emerging economies are gonna get hit with this first, and then it's gonna come here, and we're seeing it already in our portfolio companies and in the sectors that we're going after. Like, it's happening. So, really big deal. I have not seen any significant research dollars being committed to researching solutions to this problem. You know, people brought up like a basic income. That's sort of one of the more thoughtful solutions I've seen, but it's still super early. Like, it's, it's, you know, people in the Valley, technophiles essentially, talking about basic income. It's not economists at major institutions. It's not governments doing huge pilots to understand what the impact of basic income is. That's the kind of work that has to be done for this to be truly tested and viable solutions when this day actually comes. So thank you. Thank you.
