Why AI Won’t Steal Your Job. Yet
With the ability to generate human-like responses that may be difficult to distinguish from responses made by a real person, the introduction of ChatGPT in recent weeks has put the spotlight squarely on the strident progress in AI.
Reading the many pieces of commentary praising its abilities, it can be hard not to feel concerned about the technology. Top worries include its possible abuse to spread misinformation, help scammers craft well-written phishing messages, or even produce code for a ransomware.
The million-dollar question though, boils down to this: Will AI take away our jobs?
This has happened before
To be clear, this is hardly the first time that the progress of technology has threatened jobs. From switchboard operators, lift operators, and pin boys at your neighborhood bowling alley, jobs through the ages had been lost in the face of technology’s relentless march.
Even today, automated teller machines, digital order kiosks, and self-service checkout terminals are gradually reducing the need for humans. Yet for every example where technology has proven a better option than humans, there is also another example (or three) where it failed to pan out as expected.
Indeed, a 2015 feature on BBC highlighted five jobs that robots will steal first: taxi drivers, factory workers, journalists, doctors, and the cocktail waiter.
But the last time I checked the news, Apple supplier Foxconn just hired 100,000 new workers to ease production pressure at its high-tech factory. What’s more, every taxi that I’ve taken in Asia and elsewhere is still driven by a human. Now, I’m not saying that the situation will never change. But it’s worth doing a reality check before succumbing to a panic attack about how robots and AI will replace our jobs.
The (limited) power of AI
But isn’t it just a matter of time, you ask? Depends on how much time, I suppose.
Alexey Guzey recently wrote a tongue-in-cheek piece last month about how planes are still “decades” away from displacing most bird jobs such as an eagle’s ability to hunt by throwing live animals off a cliff or a hummingbird’s ability to stay perfectly still against strong winds.
“More than 200 years after hot air balloons became operational and more than 100 years after the first planes flew, it’s clear that building a [general flight] machine is much harder than anticipated and that we are nowhere close to reaching bird-level abilities,” he said as he concluded that the future of bird jobs are in no danger.
Humor aside, it is worth remembering that as impressive as today’s AI might be, they are also extremely specialized and as a result, highly limited. Even Sam Altman, CEO of OpenAI, the research laboratory that produced ChatGPT said so.
“ChatGPT is incredibly limited, but good enough at some things to create a misleading impression of greatness. it's a mistake to be relying on it for anything important right now. it’s a preview of progress; we have lots of work to do on robustness and truthfulness,” he wrote in a tweet.
Moreover, AI also suffers from inherent fragility, which makes them susceptible to failing in completely unexpected ways. As I wrote last week, researchers demonstrated an amateur-level Go-playing program that tricked a top-ranked AI program into losing.
Humans are bad at predicting stuff
Finally, humans are just bad at predicting things. We can envision broad trends, but we fail miserably in articulating the details. As Bill Gates once famously said: “We always overestimate the change that will occur in the next two years and underestimate the change that will occur in the next ten.”
Did you know that Thomas Edison had predicted that the homes of the future would be replete with steel furniture? He was thinking in terms of how durable, ubiquitous, and inexpensive the material is but forgot that most of us don’t like an abundance of the cold, hard material in our homes.
In 2012, Kristian Hammond predicted that 90% of news will be written by an algorithm in 15 years. We are not quite at the 15-year mark yet, but the evidence with ChatGPT is that AI still has some way to go. Sebastian Moss, the editor-in-chief of trade publication DatacenterDynamics attempted to use ChatGPT to write a story earlier this month, and has this to say about it.
“The AI is terrible at choosing sides – its pieces are many pros and cons lists with no payoffs. Its technical chops are lacking, and often confuse things. It doesn't know anything current, and doesn't say anything new… The copy is closer to a school paper than an industry magazine,” wrote Moss.
The world will change
In my mind, there is no question that the world will change, however. And as noted by Harvard Business Review, entry-level positions that include routine or highly-repetitive tasks will start to disappear at a faster clip, while roles focused on complex tasks with competitive salaries will arise in their place.
Elsewhere, the implications will take a while to settle in, but I suspect it will have a broad impact on the way we do many of our jobs. Like the grammar or spell checker, AI-infused tools and AI services will be par for the course in the workplace of the future.
Remember the time before the invention of typewriters or computers? Or the era before smartphones and instantaneous worldwide communication? One day, our children's children will marvel at how we ever worked without using an AI.
Paul Mah is the editor of DSAITrends. A former system administrator, programmer, and IT lecturer, he enjoys writing both code and prose. You can reach him at [email protected].
Image credit: DALL-E 2
