Like most people, I’ve been thinking a lot about how AI will affect our long-term wellbeing and whether we will see an increase or decrease in unemployment because of it. This is especially important to me for a couple of reasons:

• I'm a software engineer: it's one of the fields that has seen the largest amount of disruption due to AI, given the undeniable success of coding agents.
• I'm a father: by the time my (young) children enter the workforce, their reality will likely be vastly different to mine, and I'd like to set them up as best as possible for success.

This post is my attempt to lay out my own thoughts on the topic.

Background

If you haven't kept up with the debate, there are two main camps that try to answer the following question: "With the current state of AI and advancements made in the next few years what effect will this have on people's wellbeing?"

This question actually takes many forms, which sometimes focus more on employment, but what it's mostly getting at is will humans survive, thrive and be better off with AI or will AI result in some dystopian future.

There are two main camps here, which will be outlined below.

1. Mass unemployment and economic hardship

The pessimistic view is that AI is already showing promise of being able to automate many types of tasks (not jobs, yet) e.g. coding, routine customer service, etc.. Furthermore, the pace at which AI is advancing means that it is only single digit years until we have AGI or at the very least AI that can replace whole jobs. For this reason, we'll have mass unemployment, as companies realize that they can replace humans for machines at a fraction of the cost (and in many cases that cost saving is orders of magnitudes cheaper). This will lead to economic hardship for a huge proportion of the population, as a few AI labs (and their shareholders) become incredibly rich, leading to an even more unequal society.

It's easy to sympathise with this camp. Just in 2026 so far, there have apparently been ~120k layoffs in tech alone with many companies explicitly stating AI as one of the major reasons the layoffs are happening e.g. Block.

2. Economic abundance for all

The second camp is much more optimistic. They basically state that, as AI leads to greater efficiency and lowers the barrier to entry for many people to enter fields they may have not been able to previously e.g. a non-technical person starting a tech company, productivity as a whole will soar and we will essentially lack nothing. They also refer to Jevons Paradox extensively, saying that as people become more efficient at doing their jobs, we will likely need even more people, as goods/services which were previously prohibitively expensive will now be possible to make. Although some jobs might be replaced, they argue that many, many more will be created due to AI.

Some even suggest that people may not even need to work or, at the very least, the work week will be reduced to 2-3 days (like Bill Gates).

As with the other camp, it's not farfetched to imagine this future either. I know personally now that I can more easily spend 1 hour building something meaningful with AI, I'm much more likely to do so after work, instead of watching Netflix, than I was previously.

Who's right?

That's the part that nobody can really answer, right now. The issue isn't just that we have this new technology and it's already leading to major disruptions. It kind of is, but it's unclear whether some of the recent events like layoffs and major re-orgs have been premature without proven value. It's that it's very hard to predict how much more it will advance in the next 1-5 years. If it stays as is (which it won't), we will have a problem, but it will be manageable. But what we don't yet know is to what extent it will replace whole job categories, not just tasks needed for a job.

For example, I'm a software engineer. I used to code quite a bit. Now I still code, but it takes a fraction of the time. But... I'm still very busy. There are still a lot of higher level decisions to make e.g. system design choices, I still need to define objectives for AI to do the right thing (and in the right way), I still need to align with people on what needs to be built and which problems we need to solve and I still need to plan for the future. These used to take the lion's share of my effort and now they still do, but I can focus on them more and do a better job at them. So, I think I'm still useful. But that's not to say that I still will be useful as a software engineer in 5 years time, because I really don't know whether AI agents will be able to do all these other tasks by then.

There are comprehensive reports (like this one) that go to great lengths to predict what the effects will be, but I don't think it's actually possible right now, because it's a dynamic system and nobody knows what the capabilities will be like in the next few years.

My take

As I've said, I don't know who's right.

That being said, I'd probably say that I'm a short-term pessimist, long-term optimist.

1. Short-term pessimist

The biggest issue in the next few years for me is what I've state above. Given that we don't know what the future of AI will look like, it's leading to a lot of uncertainty everywhere. This manifests itself in companies doing mass layoffs, people being unsure they'll be relevant in the future, and generally fear being quite palpable. In other words, my major concern is not around the technology, but human psychology and how people are responding to this uncertainty, which is causing very real hardship to many.

It's also very clear that, no matter what the future is, during a major transition happening as quickly as the changes being made by AI, there will be people caught off-guard. For instance, studying Computer Science at university used to almost guarantee you one of the highest starting salaries in the UK and a fairly stable career. However, with AI automating many of the tasks typically done by more junior software engineers, people graduating today find it much more difficult to find a job. If they had known that AI would have advanced as it has, perhaps they would have pursued something else. Fortunately, future generations will likely have that foresight (personally, I still think it's a great option, even with the advancements in AI, but I'll save that for another post).

So all in all, I do expect that the next 3-5 years will be pretty chaotic, as we see whether AGI comes to be and what it can do and people stop speculating so much and acting on these speculations.

2. Long-term optimist

Even with the above though, once everything starts to settle, I am a firm believer that it will be a net positive for the world and the wellbeing of most people will improve.

I'm not basing this on a prediction of a given future e.g. AGI happens and can do everything for us at a fraction of the cost and benevolent companies share the compute required for free. I'm just betting that the future effects of technological advancements will look more like the past than people expect.

What I mean by this is that, for the most part, technology has led to vast improvements in the standard of living globally (as outlined in Factfulness: Ten Reasons We're Wrong About the World – and Why Things Are Better Than You Think). This has been almost universally true for everyone i.e. for almost every person today, they are better off than their ancestors were 100 years ago. Note that, this is not mutually exclusive with inequality potentially rising in some countries. Since economic output is not a zero sum game, you can have everyone becoming better off, with some people becoming more better off than others, such that relative inequality increases. For example, if AI makes productivity increase 1000x, but 80% of that gain goes to the richest 20%, whilst the rest has to be distributed amongst everyone else, it still means that everyone is better off relative to their past selves. I'm not advocating for such a future, I'm just acknowledging that in the worst case, people should still be better off than they are today.

I do believe that a few jobs will disappear or the demand for them will decrease a lot. For instance, customer service phone lines will likely be automated for the vast majority of cases, with humans only being needed to solve issues that are truly nuanced. But, I believe these will be a minority. I think for the vast majority of jobs, the title will remain the same or change slightly, but the work will fundamentally be transformed. This has happened a lot in history. From meteorologists to bankers, technology has made it such that their jobs are vastly different to what people with the same title did 50 years ago. In addition, I also believe that there will be new jobs that come to be, as many have also alluded to.

The way I see it, as long as there are still problems to solve e.g. healthcare, poverty, entertainment, etc. there will still be demand for people to do something to contribute. And I believe that people find creative ways to provide value.

Another reason I'm optimistic for the long term is that, until now one of the major threats to the future workforce was the fact that falling birthrates in many developed countries will lead to a much smaller workforce for an increasingly aging population. AI efficiency is one way to bridge that gap in the future.

Conclusion

I am a bit concerned about the immediate effects of AI on jobs and how we can limit the blast radius for those affected in the next few years. But, with the possible improvements to the wellbeing of many people due to AI, I'm quite excited about the future and the world that my children will grow up in.