7 jobs that will probably be safe from AI (for a while)

TL;DR - the recurring theme is jobs that require licenses or jobs that make judgments regarding values or morality.

Nursing

Demand for medicine in general is going to uptick in line with our ageing population.

Many hospitals have only just switched from paper copies of files to internet-based systems. So it might be a while until robots are running things.

Even though they’ve started automating some functions in medicine, patients continue to report that, even if an AI is "as good or better than a human," they prefer human care.

Even if some people don’t mind cheaper, AI-assisted healthcare, I suspect a section of society will always pay the premium to have human carers. Particularly in sensitive disciplines like palliative care.

AI-immunity: 8/10 

Not future-proof, but should hold for the next 10-15 years at least!

Complex Repair Work

Think technicians, plumbers, engineers etc.

Sophisticated repairs are fairly AI-proof until the price of robots drops a lot.

Affordable robots are at least a decade off being able to climb fuck-off wind turbines and carry out engineering works with the right level of dexterity.

In the short term, the demand for manual work should rapidly increase when the boomers finally retire.

Until we completely standardise a lot of our infrastructure (this could happen but will take some time), most jobs that require physical defect troubleshooting are fairly safe.

AI-immunity: 8.5/10

Maybe 20 years before affordable robot replacement is commonplace.

Religious Roles

It feels important to us that religious leaders have the capacity to suffer alongside their congregations.

A religious leader who cannot experience pain or consider their mortality with any real stakes seems unsatisfying and hollow.

This is one of about a million theological cans of worms that a de-human-ifcation of religious roles opens.

For example, consider how many traditions emphasise the (human) ‘body’— baptism, anointing the sick, laying on of hands for blessing.

In fact, take the whole idea of the resurrection of Christ, which emphasises divinity manifested in human form.

We inherently privilege human mediation of the divine, and that feels like a pretty immutable barrier for AI to cross.

AI-immunity: 10/10

While a chunk of logistical and support roles may get outsourced, the core of the religious project will remain necessarily human.

Policymakers

The people will demand that political roles remain human-led to keep a check on AI overreach in society.

We’ll still want people "with hands on the wheel"—especially where the stakes are highest.

The visions of I-Robot would be too strong if we allowed AI to run the country, surely?

And again, who is to be held accountable for outcomes influenced by an AI minister?

But there will likely be a new class of tech-savvy politicians who are leveraging AI for data analysis, sentiment tracking etc.

AI-immunity 9.5/10

Leaders will always have jobs. Being "the one riding the horse rather than the one being trampled” is perhaps the single most antifragile skill to AI advancement.

Gamers and Athletes

We’ll still want human vs. human events.

If we believe that part of the reason sport resonates is because we see ourselves in the competitors, AI can never ‘catch up’ to that.

People aren’t as enthralled watching 2 computers play chess against each other, even though they’re technically better than any human.

In fact, most sports are at least in part compelling exactly because of human error. Our capacity for mistakes gives rise to moments of essential drama.

All of that would be lost unless we’re to suppose a future wherein deliberate errors are coded into ‘competitor’ AI, meanwhile robbing the event of genuine competitive intent.

Traditional sports already lucrative, Esports is rapidly ascending as an option for the un-athletically inclined.

AI Immunity: 9/10

As long as we’re human, we’re going to want to win things—and be impressed by people who do.

Therapists

Remember—AI only has access to what we have already learnt, for the moment.

We have barely scratched the surface in terms of how the brain works. We have some useful information, but there is so much we have yet to understand.

Would you trust AI to someone who is suicidal with complex mental health issues? The liability would be insane.

The last thing a paranoid or anxious person needs is a cyber therapist. That would fuel delusions to a level previously unseen.

There’s a counterargument that sees the anonymity benefit of AI therapists, but that still relies on the companies handling your sensitive data to protect against any data breaches/not sell it to advertisers.

So it seems to me AI just pushes that problem back a step, rather than solving it.

AI-Immunity: 7/10 

AI probably opens up easier access to counselling services for those who otherwise wouldn’t use them, but is unlikely to be trusted with anything too atypical in the next 15 years.

Bartender

This makes the list for similar reasons to nursing/therapy. There are some itches that it feels like only a human can scratch, such is our unwillingness to totally ‘suspend our disbelief’ and fully immerse ourselves in AI interactions.

Even in a scenario of technically perfect AI drink-making, certain customers are just looking for a sympathetic ear. And AI’s ‘sympathy’ doesn’t seem to mean much, or lift us with the same utility.

And there’s the ritualistic component. Drinking is often ceremonial—birthdays, reunions et al. We instinctively want these moments to feel as human as possible.

My bet is that consumers will increasingly value certain human-led leisure experiences, and I think The Pub will be one of these

AI-immunity: 6.5/10

Perfect android bar staff are probably coming. But it will be a while before we’re in Mos Eisley Cantina.