At a summit in Paris on February 10th and 11th, some of the world’s biggest AI leaders made bold predictions about the future. Sundar Pichai, the CEO of Google’s parent company Alphabet, called AI “the most profound shift of our lifetimes.” Dario Amodei, CEO of Anthropic, said it would create “the largest change to the global job market in history.” Meanwhile, Sam Altman, CEO of OpenAI, went even further, suggesting that in a decade, AI could make everyone on Earth more productive than today’s most successful people.
This idea was exciting at first. Many experts believed that AI would level the playing field—helping people with less experience work just as efficiently as experts. AI tools could assist with tasks like writing, research, and customer support, allowing everyone to improve their skills. Nvidia’s CEO Jensen Huang even imagined a future where every worker could act like the CEO of their own AI assistant.
But new research suggests that AI may do the opposite—instead of making everyone equal, it could make the best people even better, while leaving others behind.
Who Benefits from AI? The Best or the Rest?
Early studies showed that AI was most helpful to beginners. In 2023, researchers from Stanford and MIT found that AI boosted productivity by 34% for new customer service workers, helping them respond faster and better. Similar studies showed that AI helped weaker writers improve and even made law students better at drafting contracts. The idea was simple: AI would give less-experienced workers a shortcut to better results.
However, new studies now tell a different story. Instead of helping everyone, AI seems to work best for top performers—those who already have skills and knowledge. Why? Because using AI isn’t just about taking its answers—it’s about knowing how to judge what’s useful and what’s not.
For example, a study at MIT found that AI helped the best scientists double their productivity when discovering new materials. But for less experienced researchers, AI made no difference at all—they couldn’t tell which AI-generated ideas were good or bad.
Another study on Kenyan entrepreneurs found that the best business owners increased their profits by 15% using AI, while weaker ones actually lost money because they followed AI’s advice without knowing how to adapt it.
AI Will Change Jobs—But Not Equally for Everyone
AI is already automating simple, repetitive jobs. For example, self-checkout machines replaced cashiers. Now, AI-powered chatbots are replacing customer service workers. A tech executive at ServiceNow said that 85% of customer support cases no longer need a human. This means that while AI may first make workers faster, in the long run, it may replace them entirely.
Even in high-skilled jobs, AI is shifting responsibilities. At major law firms, AI now reviews contracts in seconds, a job once done by junior lawyers. As a result, young professionals have to learn new skills faster to stay valuable. A partner at a top law firm explained that the best lawyers now use AI to make strategic decisions, while weaker ones struggle to keep up.
The Future: A Bigger Gap Between Top and Average Workers
Throughout history, new technologies have created winners and losers. In the Industrial Revolution, engineers who understood machines earned more, while manual laborers lost jobs. When computers became mainstream, software engineers thrived, but typists became obsolete. AI seems to be following the same pattern—rewarding those who can think critically, analyze complex information, and adapt quickly.
As AI improves, it may eventually create semi-autonomous AI assistants that can act independently, just as Nvidia’s CEO predicted. If that happens, every worker might have access to an AI-powered “CEO assistant.” But not everyone will know how to use it effectively. The most skilled people will still find ways to stay ahead, making the gap between top performers and the rest even wider.
In short, AI isn’t a magic tool that makes everyone equal. The future belongs to those who can use AI wisely—not just follow its answers, but question, refine, and improve them.
Prepared by Navruzakhon Burieva
