How To hedge against AI stealing your job

By Chris Bryant

Will artificial intelligence take my job? This question is really starting to preoccupy me and millions of other white-collar workers. There’s even a word for it — FOBO, or fear of becoming obsolete — and, regrettably, our apprehension isn’t entirely unfounded.

While I’m not counselling panic, I do think knowledge workers — a catch-all term for people such as administrators, financial analysts, software developers, legal professionals and those in the creative industries whose jobs are most likely to be transformed by AI — should prepare for massive upheaval. This means embracing AI to become more productive, while having a backup plan in case your efforts to resist the march of the machines fail.

Until recently, I was tempted to dismiss chatbots as gimmicky, error-prone stochastic parrots; the latest versions, however, can reason and solve complex problems, and their answers are often indistinguishable from a human’s. Their capabilities keep getting better while the costs of using AI continue to fall.

Previous advances in automation and outsourcing mainly affected blue-collar employees. This time it’s those cossetted professionals who tap away at keyboards in offices (or WFH) who are most at risk of being displaced.

The psychological blow of devaluing brainwork and specialist knowledge is heavy enough — what was the point of all those college degrees or learning a foreign language, one might ask? But the financial toll could be worse. Knowledge workers often marry each other, a phenomenon known as assortative mating. Previously the concern was that this trend increases inequality; now, it leaves those families doubly exposed to the AI chopping block.

Think tank IPPR estimates that knowledge work now accounts for around half of UK labor market activity, and as much as 70% of those tasks could be significantly transformed or replaced by generative AI. Those earning more than $250,000 a year contribute around half of US consumer spending, according to a recent Moody’s Analytics analysis. What if millions of high earners are replaced by algorithms?

While he’s not an unbiased observer, OpenAI co-founder Sam Altman says a software engineering agent will eventually be capable of doing most things that a human engineer with several years of experience can do, albeit with supervision. And “eventually” might be quite soon: Meta Platforms Inc. boss Mark Zuckerberg says these virtual co-workers will have the coding and problem-solving abilities of a mid-level engineer by this year.

Unlike people, AI agents can be deployed at near limitless scale across all kinds of knowledge work. So while AI will certainly augment existing roles and create entirely new occupations, I think lots of job losses are inevitable (and Altman agrees). There’s already talk of a white-collar recession as graduates from top MBA programs struggle to find work, translator commissions are drying up, starting salaries for consultants stagnate and layoffs in information technology increase.

Salesforce Inc. says it won’t hire any software engineers this year thanks to productivity advances from AI tools; Nvidia Corp. will soon require all its software engineers to use AI agents, while more than one quarter of new code at Alphabet Inc.’s Google is now generated by AI.

In banking, AI can draft almost all of an IPO prospectus in minutes, according to Goldman Sachs Group Inc. Chief Executive Officer David Solomon; previously, that might have taken a six-person team two weeks to produce. Last month, Singapore bank DBS Group Holdings Ltd. said it would cut around 4,000 contract and temporary staff in the next three years as AI takes over their roles. If AI were poised to merely free us humans from dull, repetitive tasks I’d probably be less apprehensive. But the latest AI models are capable of PhD level reasoning.

It makes sense to have a back-up plan. The first step is to familiarize oneself with AI tools and become expert at using them; this should hopefully make you more productive and enable work that’s more impactful. Doing so will also give you a better idea of whether your job is in danger, while offering some protection that you won’t be among the first to get laid off. However, we won’t all succeed in becoming AI managers — so keep networking on LinkedIn and elsewhere in case you do need a career lifeboat.

Those yet to join the workforce — or counselling their children — face difficult choices. Nvidia boss Jensen Huang last year sounded less convinced about the merits of learning to code because soon everyone will be able to program using natural language (however, I tend to agree with Microsoft Corp. founder Bill Gates that learning the basics is still important). Similarly, if AI chatbots start taking over all the grunt work in investment banking or management consultancy, getting a coveted internship or graduate program spot could become much harder.

Others further along in their careers should consider whether they have skills that open doors to a job or entrepreneurial opportunity that isn’t as susceptible to AI disruption. In other words, it’s a good moment to perfect that side hustle. While I struggle to imagine doing anything other than journalism, I’m half-convinced I’d make a decent landscape gardener or barber, or failing that a cat café or cheese-shop owner (I’m only half-kidding).

As AI looks like being a lot kinder to capital owners than labor, another sensible move is to acquire more assets. An index fund provides exposure to companies that stand to profit from AI and related productivity gains and a hedge against being laid off (although the nosebleed valuation of some AI-linked firms indicate many folks have fully embraced this idea already).

Unfortunately, there’s no guarantee AI won’t steal your knowledge-worker job. But having a back-up plan can at least help keep the FOBO at bay.