“People who look at a hard problem and say ‘I know this is really, really important, and we can figure this out’” are the next change-makers, according to the semiconductor leader.
The internet transformed communication, mobile computing changed “how we live,” and cloud computing altered the way of work. But the MIT alum said that the AI era is different from all the previous technological shifts that have taken place because it will accelerate discovery in every sector—and finally get to the bottom of unsolvable problems.
And while AI may supercharge human capability across industries, Su stressed that the technology itself won’t decide the future—people will, making it critical for young workers to learn how to use the tools responsibly.
“Now, the way to think about [AI] is it makes each of us more capable, whether you’re talking about medicine, science, energy, [or] climate,” Su said. “But let me be clear about something: Technology itself does not decide what the future looks like, the best people do.”
“For everything that AI can do, AI can’t decide which problems are worth solving. It can’t make the hard judgments when the data is not there. It can’t take responsibility for the outcomes,” she continued. “These are actually our responsibilities, and they matter now more than ever.”
Available roles that mentioned large language modeling (LLM) grew from 5,000 to 20,000, and those that referred to prompt engineering also rose from 1,400 to nearly 6,300 in the same period.
The leader said it typically takes 10 years to develop those types of skillsets, and professionals are being thrown into the deep end. Essentially, staffers are wielding 2025 tools while held to 2015 job structures.
Getting comfortable with the tech is now a requirement, but the real unlock lies where budding professionals intertwine it with their own human judgement.
