The Case for Active Optimism

Exploring the innate human addiction to catastrophe, and what it means for the future of work

Humans are extraordinary at one thing above all else: finding something to worry about.

We have, throughout our entire recorded history, been absolutely certain the end was near. The early Christians believed the Kingdom of Heaven would arrive within a generation, Martin Luther predicted the apocalypse, and Christopher Columbus calculated the world would end in 1658. The Cold War gave us a literal Doomsday Clock which, in its 70-something years, has never been more than 17 minutes from midnight. When the Soviet Union collapsed and nuclear apocalypse faded from the headlines, you might have expected a collective exhale. Within a decade we had Y2K, then climate catastrophe, then the Singularity.

This pattern spans cultures and centuries. In 5th century China, Taoist texts prophesied a messianic figure who would destroy society and rebuild it. During the Yellow Turban Rebellion of 184 AD, peasants proclaimed that an age of inequality would end in destruction and rebirth. A 2012 Pew Research poll found that in Afghanistan, 83% of respondents expected the Mahdi to return in their lifetime, and in Iraq it was 72%, both countries having just endured a decade of war. We are wired to scan for threats. Our ancestors who worried survived, and the ones who were relaxed about the rustling in the bushes did not pass on their genes. The snakes and tigers are largely gone, but the anxious wiring remains, so it gets directed at whatever feels most existentially threatening at the time.

Right now, in my world, that threat is AI. I spend most of my time talking to people who are worried about their careers, and the dominant mood in almost every conversation is fear. The story goes like this: AI destroys jobs, without jobs humans lose all sense of purpose, and without purpose society collapses. It’s a tidy narrative, it maps perfectly onto our hardwired appetite for catastrophe, and I think it’s missing a huge part of the picture.

None of this is to say the concerns aren’t legitimate. AI will reshape the labour market in ways we can’t fully predict. But layered on top of every legitimate threat is a thick psychological coating of catastrophism that has been with us since the first human looked at the sky and wondered if the gods were angry. Recognising that coating exists doesn’t make the threats less real, but it should make us more careful about how we respond to them.

We’ve been here before

Before the Industrial Revolution, the overwhelming majority of the British population worked the land. By 1851, for the first time in the country’s history, more people lived in towns and cities than in the countryside. The mechanisation of farming didn’t just change how food was produced, it fundamentally rearranged where people lived, what they did for a living, and what kind of society they belonged to. The transition was brutal in places. Entire rural communities were hollowed out as people migrated to factory towns, and the early decades of industrial work were dangerous, poorly paid, and dehumanising in ways that are hard to overstate.

And then, over time, things got better because people made them better. Labour laws were written, unions formed, education expanded, working hours fell, and living standards rose in ways that would have been unrecognisable to the first generation of factory workers. The agricultural workers who were displaced didn’t disappear from the economy; they and their children became factory workers, then office workers, then service workers. In the US, the agricultural share of employment dropped from 41% in 1900 to under 2% by 2000, while farm output more than doubled.

The same pattern repeated with manufacturing. US manufacturing employment peaked at 38% of the workforce in 1944 and fell to 8.5% by 2019, while the service sector grew to employ over 80% of workers. A study by Autor et al. found that roughly 60% of workers in 2018 held occupations that did not exist in 1940. Sixty percent of modern jobs were literally unimaginable two generations earlier.

When ATMs were introduced in the 1970s, everyone predicted the death of bank tellers. What happened was that the cost of running a branch fell, banks opened more branches, and the number of tellers actually increased for decades. Their job changed, from counting cash to relationship management and sales, but the job didn’t vanish.

I know the counterargument. These transitions played out over decades or centuries, and AI is moving faster. That’s fair, and it’s one of the things that makes this moment different from previous transitions. But the direction of travel has been remarkably consistent across every major technological shift we have data for, and dismissing that track record because this time feels scarier seems like exactly the kind of thinking our catastrophe-wiring would produce.

What we know so far (not a lot)

The World Economic Forum’s Future of Jobs Report 2025, which surveyed over 1,000 employers representing 14 million workers across 55 economies, projected that 92 million jobs will be displaced by 2030 while 170 million new ones will be created, a net gain of 78 million jobs globally. That’s significant, but it’s also just one projection from one set of employers, and employer projections have historically been unreliable in both directions.

What’s more telling is the disagreement among the people who should know best. The CEOs of the companies building AI can’t agree on what’s coming: one warns that half of all entry-level white-collar jobs could disappear within five years, another argues that productivity gains have always created more work throughout history, and a third writes optimistically about an age of unprecedented prosperity. Meanwhile, the Yale Budget Lab analysed US employment data since ChatGPT’s release and found that measures of AI exposure showed no significant relationship to changes in employment at the aggregate level so far. The Dallas Fed found that workers aged 22 to 25 in AI-exposed occupations have seen a 13% decline in employment since 2022, but the overall impact on unemployment has been roughly 0.1 percentage points.

I’ve spoken to hundreds of people about career confusion over the past year, but I would not bet on myself to crack a question that has the people building and studying this technology fundamentally at odds with each other. What I can say, based on what I’ve observed, is that the conversation has become strangely binary. You’re either terrified or dismissive, a doomer or a cheerleader, and there’s almost no space for the people in the middle who think this is going to be a profound shift that we have genuine agency over.

The case that rarely gets made

Almost every conversation I hear about AI and work follows the same script: jobs will disappear, people will lose purpose, society will unravel. The positive scenarios barely get airtime, and when they do, they’re treated as corporate propaganda or wilful naivety.

So let me try to sketch one, with full acknowledgement that I might be wrong.

Research from the London School of Economics found that employees using AI save an average of 7.5 hours per week, roughly a full working day. That’s an extraordinary number if it holds up at scale. Right now, most of that productivity gain is being captured by employers, which is part of why the job displacement anxiety is so acute. But it doesn’t have to stay that way. What if some of that reclaimed time gets redistributed back to workers, in the form of shorter weeks or more meaningful work? What if the tasks AI handles best turn out to be the repetitive, administrative, soul-destroying parts of most jobs, the exact tasks that make people miserable?

What if the entry-level jobs that disappear are replaced by entry-level jobs that are more interesting, the way backbreaking field work was eventually replaced by roles that involved creativity, problem-solving, and human connection? Each transition in history was painful, but each also raised the floor of what work could be.

I’m not guaranteeing this outcome. I’m saying it’s at least as plausible as the catastrophe scenario, and the fact that almost nobody is talking about it seriously is itself a problem, because the future we get will depend heavily on which scenario we prepare for.

How you orient towards this matters

I’ve been thinking about why the conversation about AI and jobs has become so stuck, and I think it comes down to something that most people haven’t properly separated in their own thinking. Picture two axes: one running from passive to active, the other from pessimistic to optimistic. This gives you four positions, and most people occupy one of them without realising it.

The actively pessimistic person goes looking for evidence that things are getting worse. They seek out the most alarming projections and the most catastrophic framing, and they build an identity around sounding the alarm. Every generation has had these voices, from the Luddites smashing textile machinery to the Y2K preppers stockpiling canned goods. Sometimes they’re right about the danger, but they’re almost never right about the scale or permanence of the damage.

The passively pessimistic person has accepted that everything is terrible and checked out. This is nihilism in a comfortable jumper. The world is burning, nothing matters, so why bother engaging with any of it? This orientation produces total disengagement from the systems and decisions that will determine how this transition plays out.

The passively optimistic person assumes everything will work out without any effort. Technology always creates more jobs than it destroys, history always bends towards progress, someone else will sort it out. This is the position of people who cite the agricultural revolution and stop there, without mentioning that the transition took over a century and involved immense human suffering along the way. It’s comfortable, and it’s irresponsible, because every historical precedent shows that transitions go well only when people actively work to make them go well.

Then there’s active optimism, which is where I’m planting my flag. Active optimism recognises that things have historically worked out well because people worked to make them work out well. The agricultural transition wasn’t resolved by market forces alone. Governments intervened, labour laws were written, education systems were overhauled, unions fought for protections that hadn’t previously existed. The shift from manufacturing to services required retraining programmes, new regulations, and decades of deliberate policy work.

Active optimism applied to AI means holding two things at once: that this transition will be deeply disruptive and that we have meaningful agency over the outcome. It means pushing for reskilling infrastructure, safety nets, education reform, and fair distribution of productivity gains, while also recognising that the historical trajectory of technological change has, on balance, expanded the range of what humans can do and be paid for.

What this means for your career

If you’re reading this and feeling anxious about your work, I want to be direct with you.

The evidence so far suggests that the immediate risk is concentrated among entry-level roles where tasks are repetitive and easily codifiable, and the workers most affected are young people finding fewer traditional entry points into the workforce. The LSE research found something encouraging here though: a trained Gen X employee gets more productivity benefit from AI than an untrained Gen Z employee. The gap is about who has bothered to learn how to use these tools and who hasn’t.

The actively optimistic response at an individual level is simple enough. Stop doom-scrolling about whether AI will take your job and start learning what it can and can’t do. Figure out which parts of your work it can handle and which parts require your judgment, creativity, and human understanding. The people who learn to work alongside AI will be more valuable, and the people who refuse to engage with it will be more vulnerable. That’s been true of every technological shift in history.

The broader actively optimistic response is to push for systemic change: better reskilling programmes, stronger safety nets during the transition period, education systems that teach adaptability and critical thinking, and companies and governments that distribute productivity gains broadly.

None of this happens if we spend our time arguing about whether the apocalypse is coming while the change rolls on regardless.

The farmers of 1900 couldn’t have imagined that their great-grandchildren would work in software development, UX design, or content strategy. The factory workers of 1950 couldn’t have predicted data analysts or social media managers. We can’t predict what jobs will exist in 2050 either. But if history is any guide, and it’s the only guide we have, there will be more of them, and the people who navigate the transition with agency will do far better than those who panic or disengage.

That’s the argument for radical optimism. Things will probably work out. But only because people like you decide to make them.

---

I’m building Rumbo, an AI career coaching platform that helps you understand yourself, the market, and the problems you’re drawn to, so you can make a career decision you trust. It’s live now here.