A recent Bloomberg article cited an MIT economist's claim that AI is only capable of doing 5% of jobs, even warning about a potential economic crash due to AI's limitations. This perspective might come across as cautious, but it misses the bigger picture of AI's transformative role across industries and its steady expansion into far more than what the numbers suggest.
One of the biggest misconceptions about AI is the notion that it’s either replacing human jobs entirely or doing nothing useful at all. In reality, AI's power lies in augmenting, enhancing, and reshaping work rather than just replacing it. Even if only 5% of jobs could be fully automated today, many more occupations are being fundamentally transformed by AI. Healthcare is a good example: AI can’t replace a doctor, but it can analyse medical images, flag anomalies, and suggest diagnoses with an accuracy that supports doctors. The role of radiologists is evolving, as AI allows them to work faster and with more confidence. This isn’t just a healthcare story; finance, law, and marketing are seeing similar shifts. So instead of focusing solely on jobs replaced, we need to look at how many jobs are changing, and that number far exceeds 5%.
The 5% claim also treats AI as if it’s stagnant and limited in scope. The truth is, AI is a general-purpose technology, like electricity or the internet. Both of these technologies started with limited uses, electricity powered lights, and the internet connected research labs, but eventually permeated nearly every aspect of life and work. AI is on the same trajectory. It might seem like it can only do a small range of tasks today, but its capabilities are expanding at a rapid pace. If AI automates 5% of jobs today, it could be 10% next year, and far more in five years. AI keeps improving as machine learning algorithms advance and new techniques, like self-supervised learning, emerge.
Another issue with focusing on jobs that can be completely replaced is that it misses AI’s real strength, automating portions of jobs, which allows humans to focus on tasks that require creativity, strategy, or interpersonal skills. McKinsey estimates that 60% of all jobs have at least some tasks that can be automated. These are often the repetitive or mundane tasks, and this is where AI adds immense value, even if it doesn’t take over entire roles. For instance, in customer service, AI-driven chatbots handle common inquiries quickly, while human agents are left to address complex issues. In manufacturing, robots perform high-precision tasks, freeing humans to focus on quality control and problem-solving. AI might not be doing the whole job, but it’s transforming how the work gets done, driving major efficiencies.
The economist’s fear of an economic crash due to AI's supposed limitations also warrants a closer look. Historically, economies adapt to new technology. AI contributes to productivity gains in ways that may not be immediately visible, and these gains offset concerns about job displacement. The argument that a lack of AI-driven transformation will lead to economic failure seems to rest on a flawed assumption: that if AI isn’t replacing the entire labour market instantly, it will fail catastrophically. Technological change doesn’t work this way. Instead, we’re likely to see a gradual redefinition of roles and skills. This will require investments in reskilling, but it’s not a situation that leads to sudden collapse. If anything, AI adoption will foster productivity growth, reduce costs, and create new opportunities, all of which suggest economic expansion rather than contraction.
AI shouldn’t be seen as a monolithic technology either. Different industries adopt AI at different paces, with varied applications ranging from basic automation to sophisticated decision-making. Limiting AI's impact to just 5% of jobs ignores its broader role in driving innovation. In retail, for example, AI-driven logistics and inventory management have massively increased efficiency, even if store staff aren’t being replaced by robots en masse. AI’s value is much broader than direct labour substitution, it’s about optimising supply chains, enhancing customer experience, and providing data-driven insights that weren't possible before.
The idea that AI can only perform 5% of jobs overlooks its real impact. AI is not just about outright replacement; it's enhancing roles, automating parts of jobs, and proving to be a general-purpose technology that continues to grow more powerful each day. From augmenting human work to automating mundane tasks and driving productivity gains, AI's economic influence extends far beyond replacing jobs. If we focus solely on what AI cannot do today, we risk ignoring the subtle yet significant changes it’s already bringing to the workforce and will continue to bring in the future. The success of AI isn’t about reaching an arbitrary target for jobs automated, it's about how well we adapt, evolve, and make the most of a technology that is still only in the early stages of revolutionising our world.