Will AI create more jobs than it replaces? 

Even on the slow route to the intelligence explosion, big tech plays a vital role: they advance AI as fast as they can without safeguards. [Cognitive scientist and AI existential safety researcher] Peter Park said, “Big tech has boatloads of money and boatloads of lobbyists, but big tech also has admitted, either implicitly or explicitly, that their goal is to reduce the economic leverage of humans to zero. And they don’t have a plan to ensure humans’ ability to advocate for their rights and interests, which will go to zero soon after their economic leverage goes to zero.”

And how does our economic leverage go to zero? When we lose our jobs, our money, and with them our ability to influence lobbies and politicians. Despite their seeming innocuousness, chatbots and image generators have already replaced everyone from graphic designers to customer service representatives. Big tech has been heavily hinting at job loss for a long time, and once the AI revolution got underway, the head-rolling started, ironically, at the top. In 2023 tech workers had the hardest twelve months since the dot-com crash of the early 2000s; 260,000 jobs disappeared. In the first part of 2024, nearly one hundred tech companies, including Amazon, Microsoft, Meta, TikTok, Google, and Salesforce, laid off about 25,000 employees. 

Tellingly, Google and Microsoft reduced staff in their respective trust and safety divisions as recently as March 2024. Microsoft had already fired every member of its team responsible for directing ethical AI innovation. The corporations aren’t explaining the mass firings, but industry analysts say the goal is to channel money away from safety and toward developing products and profits. 

A recent poll conducted by the staffing firm Adecco Group indicates a wave of automation, with 41 percent of executives at big organizations globally anticipating a reduction in workers over the next five years as a result of artificial intelligence. Adecco CEO Denis Machuel told Reuters that AI is emerging “as a great disruptor in the world of work. Companies must do more to re-skill and redeploy teams to make the most of this technological leap and avoid unnecessary upheaval.”

Adecco surveyed executives from nine different countries and eighteen different industries—including both white-collar and blue-collar jobs. Recent layoffs in the tech sector support these worries. Employers such as Google and Microsoft are cutting staff in favor of AI chatbots like ChatGPT and Gemini. AI has been used as a justification for downsizing even by outside tech companies such as Dropbox and Duolingo. Goldman Sachs economists have already issued a warning that generative AI has the potential to remove or drastically affect up to three hundred million jobs worldwide, with white-collar professionals being the most susceptible. According to the Adecco poll, this forecast could come true in the next five years.

In those five years, managers will recapture many income streams from employees who do things generative AI can do a lot faster: write reports and newsletters, translate, perform legal research, provide customer service, program, and create other forms of content. Peter Park sagely noted, “The iteration of AI we’re experiencing was made for the managers of the world.”

If you want to know how big tech thinks about human workers, their opinions aren’t veiled. As noted, OpenAI’s charter says their goal is to create “highly autonomous systems that outperform humans at most economically valuable work.” That value was echoed recently when a TV presenter asked Mustafa Suleyman, co-creator of DeepMind at Google, if artificial intelligence was “going to replace humans in the workplace in massive amounts.”

Suleyman replied, “I think in the long term—over many decades—we have to think very hard about how we integrate these tools because, left completely to the market … these are fundamentally labor replacing tools.”

Try Big Think+ for your business

Engaging content on the skills that matter, taught by world-class experts.

Request a Demo

I think Suleyman threw in “over many decades” in order to avoid being struck by a hail of smartphones. Most experts’ prognostications are substantially shorter than that. Obviously he is stating the intelligence explosion’s long version. I don’t think we have many decades to preserve a role for ourselves because as a labor-saving tool, generative AI, with some work under the hood, is poised to be unequaled as it advances toward every big tech company’s goal: AGI. And there’s more than one route to AGI. The impact of automation and computers has been felt for decades, but the pace and scale of job losses have accelerated over time as technology has advanced.

Manufacturing and administrative occupations started to disappear in the 1950s and 1960s due to automation and computers, though this was a gradual trend that only affected repetitive and regular work. But starting in the 1980s, automation’s reach increased and human labor began to be replaced more frequently and extensively in all economic sectors. After a brief slowdown in the 2000s, concerns over automation-driven unemployment have risen again as capabilities in artificial intelligence and machine learning have increased.

The late I.J. Good and Stephen Hawking, as well as Geoffrey Hinton and Stuart Russell, are far from alone in thinking that machines might take over from their human inventors. The Victorian novelist and evolutionary theorist Samuel Butler (1835–1902) predicted that “the time will come when machines will hold the real supremacy over the world.” In his profound and pioneering study Darwin Among the Machines (1997), in which he discusses Butler, Good, and natural selection, the science historian George Dyson summarizes the theoretical process through which humankind would be surpassed: “In the game of life there are three players at the table: human beings, nature and machines. I am firmly on the side of nature. But nature, I suspect, is on the side of the machines.”

According to studies and predictions over the last five years, AI is expected to replace millions of jobs across various industries by 2030, and create many as well. But it’s not apparent—unlike with industrialization in the eighteenth century—that artificial intelligence will lead to the creation of new employment. However, estimates of job loss vary widely. For example, a study by the Brookings Institution predicts that by 2030, AI could displace between 1.3 million and 2.3 million jobs in the United States, but it could also create between 1.4 million and 2.4 million new jobs. On the other hand, a study by the McKinsey Global Institute predicts that by 2030, AI could displace between 400 million and 800 million workers globally, but it could also create between 130 million and 230 million new jobs.

But now consider the impact of multimodal AI, which is taking impressive baby steps today. If ChatGPT and the other generative AI will take jobs based on words alone, imagine AI that can ingest images, videos, speech, text, and sensor data, and create sensible ideas from this diverse data together. Multimodal AI will help make better translation apps, intelligent virtual assistants, autonomous vehicles, augmented reality systems, social robots, and more. Unifying different sensory streams allows these systems to learn complex concepts of the world, far more complex and complete than words alone can express. Research is ongoing into fusing modalities to achieve more generalized artificial intelligence.

As Professor Melanie Mitchell said, a fundamental part of our intelligence involves “… physical abilities, physical reasoning, physical interaction with the world.” This doesn’t just point to robots. It points to a multidimensional way for the cognitive creatures we are creating to ingest the richness of the planet, and help them take a big step toward a general intelligence. Intelligence isn’t only about processing information, but about physically experiencing the world around us. Multimodal AI seems aimed at helping AI “minds” take in the planet’s richness the way they currently gobble up words.

This connects to how the intelligentsia used to say AI art could never replace human art. They had two good reasons: humans need to make art to feel fulfilled, and hey, what kind of art could a computer possibly make anyway? Then DALL-E and Stable Diffusion came along and knocked us flat with images that look just as good as human-made paintings and photos. They get everything right from the basics to the smallest details. And we didn’t see it coming.