I can see both sides of the automation discussion to be honest.
On the one hand, automation is an absolutely crucial tool to maintaining our standards of living. It keeps costs down for consumers, it maintains profitability for businesses which in turn keeps tax revenues up for the government to in theory use to maintain infrastructure (and in reality squander). In that way it is essential to the viability of the nation. Plus of course, most of the products we buy are only available at all because of automation, and certainly they're only affordable to the average Joe because of it. Food, cars, technology, clothes, a multitude of others. I can't fault any business for investing in it, even at the loss of jobs. It's a necessary evil in that respect.
However, there are disadvantages. The simple fact is that, even if automation DOES generate new jobs, and it does of course, they're not jobs that the person being replaced on a production line, or in a grocery store, or in a warehouse are realistically able to 'just pick up'. The skill sets aren't always there, they're much more technically complicated jobs, they require a great deal of extensive training, and often they're a lot easier to offshore.
As an example, I did a case study during my MBA on a project run by a FMCG company to optimize one of their production facilities, integrating IoT, centralized data collection, and Machine Learning based optimisation tools. Once it was complete, they'd increased capacity from that site by 30% with no nominal increase in head count and a substantial reduction in COGs. The dirty secret (not reported in the 'official' case study I might add) was that the business case was contingent on being able to close two of their smaller, less efficient facilities as a result. The project was successful, and 500 people lost their jobs. Of course, the new, more automated facility with its IoT integrated equipment DID create new jobs. In the end, they hired 4 optimization engineers, 8 coders, a statistician, and a machine learning specialist (at the corporate location), plus 4 additional maintenance specialists and 4 additional technical personnel (on site). Their suppliers probably had to hire a couple of technical support personnel as well.
But that's no comfort to the 500 people out of work, especially as they'd need to move across the country at the minimum, and to the Netherlands for most, to take those positions. Plus of course, a guy with 10 years on the job experience running a packaging line, or a spray dryer, isn't realistically going to take a machine learning position.
But, the efficiency went up, and the cost of goods went down enough that it allowed them to keep price hikes to a minimum during COVID, soaking up a lot of raw materials inflation through increased production efficiency. It also went some way to keeping the business commercially viable during that period, which maintained many more jobs in country and gives consumers more choices on shelf.
This is a pretty typical example of automation in manufacturing in 2024. Some pros, some cons, probably necessary for a business to be competitive, but coming with a net head count reduction.
Roll this up to the national level, and especially when you consider the influence of Big Tech in this space, and I think it becomes problematic. Take Amazon. It's a great benefit to consumers. Very cheap goods, delivered right to you door very quickly. But, also consider that every $1 in revenue that goes to amazon is coming straight out of the pockets of traditional retail. Amazon delivers that service with substantially less people. The true COST of Amazon's business model is a loss of retail jobs, warehouse jobs, delivery and courier jobs. Millions of them across the nation. Plus Amazon taking that order doesn't increase the size of that order (much). You buy the same bog roll, or the same jacket, or the same TV as you did before, just for less money, so upstream benefits to manufacturers are fairly minimal.
It's the same story with Tesla's self driving tech. If they really can manage self driving semi-trucks, the benefits are evident, but the COST, is again, millions of well paid jobs disappearing. If AI truly takes off, we'll see the same thing with paralegals, and accountants, and coders, and many other white collar jobs. Social media has already done this for many traditional news outlets, and it will continue to eliminate those jobs, and marketing jobs, and consumer insights jobs, and sales jobs, and broadcasting jobs as it grows. I can't see any of those jobs magically creating more jobs than they take. After all, cheaper trucks doesn't really mean we will move more goods, nor does AI tools mean we'll door do more tax returns, or need more legal services, or read more news. Affordable mechanisms are already in place to serve these needs, these are just more profitable.
This takes me on to the overall point of this post. This wave of automation isn't (IMO) like the last industrial revolution, or even the slow creeping industrial automation of the past few decades. Those worked to create jobs because standards of living were low enough, and export markets were numerous enough that making more things more efficiently allowed more of them to be sold. Henry Ford and the car is a good example of that, as is industrial glass blowing back in my hometown in the UK in the 1800's, or even the proliferation of affordable home computing that the advances of the 80's and 90's enabled. Automation provided entirely new markets, opening up those goods to a slew of new consumers, which in turn lead to increased prosperity.
But today, I don't think that'll happen in nearly the same way. Much of this wave is about either selling the same amount of stuff to the same consumers cheaper (eg Amazon, automated self check out, automated warehouses, self driving trucks), or providing the same intangible goods more cheaply (advertising on FB vs cable, AI driven accounting services, or legal services, or market research). It's all bottom line focused, not growth focused, at least within western nations, and let's be real, if the market your AI guided micro chip design tool opens up is affordable smartphones for Indians, Nigerians, or the Chinese, you aren't adding much headcount in America as a result, nor are you drastically increasing living standards for Americans.
As such, I don't see a world where the current wave of automation CREATES more jobs than it destroys, at least in the West. Therefore it's my opinion that the conversation around a universal basic income is going to become a genuine policy point in the US within my lifetime as a result.
It'll be a fantastic thing for developing nations though.