Episode summary
Kai-Fu Lee on How AI Can Help Grow a Company's Earnings
At Barron's
· · 24 min · Hosts: Andy Serwer · Guests: Kai-Fu Lee
Summary
Kai-Fu Lee argues that most companies are using AI too narrowly, treating it as a set of tools at the edges of departments instead of making it central to how the business runs. He says the right test is whether AI changes earnings, because AI should improve revenue, profit, growth, and risk management when used by CEOs and top executives rather than just frontline staff.
He expects AI adoption to create both job losses and job gains, with routine and cognitive work hit first while physical jobs remain safer for now. He also says the biggest near-term danger is not rogue superintelligence, but bad actors using AI or humans under-specifying goals badly enough to create harmful outcomes. On governance, he favors serious safety and guardrail work, but thinks government will need help from companies because regulators are too slow and often do not understand the technology.
Lee is bullish on the U.S. in enterprise software, semiconductors, and model quality, while saying China has strengths in hard engineering, consumer AI, and robotics. For investors, he says applications are the most attractive part of the stack, and he believes AI could make the overall opportunity at least 10 times bigger than the internet revolution.
Key takeaways
- Kai-Fu Lee says AI should be embedded at the center of a company, not glued to peripheral departments like finance or customer service.
- He argues CEOs should judge AI by the earnings call test: whether it changes revenue, profit, growth, and risk.
- Lee expects job losses to come before job gains, with routine and many cognitive roles most exposed first.
- He sees the U.S. leading in enterprise software and semiconductors, while China leads in consumer AI and robotics.
- Lee says the most attractive investment area is applications, not chips or base models, because that is where long-term value accrues.
Notable quotes
“AI needs to be at the center of the company, making everything run around it so that AI can self-improve”
“you have to stop thinking about AI as just another app because another app doesn't change your earnings numbers”
“the job losses will come first. The job gains will come later”
“I don't take it seriously. I don't rule it out completely”
“At least at least 10 times bigger because this is intelligence. We're talking about internet commoditized information”
How the episode opens
Hello everyone and welcome to At Barrens. I'm Andy Serur and welcome to our guest Kyu Lee, an AI scientist, former executive at Apple, Microsoft, Google entrepreneur and author of a new book AI native, the mandate to transform your company. Welcome, sir. Great to see you.
Thank you. Thank Andy. So, um it really is great to be um talking to someone who's so wellversed in AI. My understanding is you've been working on it for four decades. And one thing I wanted to ask you right from the top sort of about your book is that you say that companies are pouring money into AI, but they're maybe doing it the wrong way. What does that mean generally?
Well, it's an okay start, but they got to change. And what they're doing wrong is kind of like um they're gluing AI at the periphery of the departments. So finance department, legal department, um you know, customer service.
They're saying why don't you use AI? And that's all well and good as a starting point. That's how we started with the internet, right?
Departments had browsers and but eventually companies stop operating without an internet. when the internet's pulled, the companies don't run anymore. Everything runs on it, applications and everything.
And and AI is going to be exactly like that. AI needs to be at the center of the company, making everything run around it so that AI can self-improve and it has to be done fast because AI adoption is at least five times faster than internet adoption. And you said something about the earnings call test,
Right? What does that mean? Explain that. Well, that means what I talk to a lot of CEOs. They're usually trying to tell me what a great job they're doing.
From the episode’s captions: roughly the first two minutes, lightly cleaned. Keep watching on YouTube →
Topics
- ai for earnings growth
- company productivity
- artificial intelligence applications
- investor perspective
Watch the episode
Watch “Kai-Fu Lee on How AI Can Help Grow a Company's Earnings” on YouTube
Summary, takeaways and quote selection by Octagon. Quotes are short excerpts; the full conversation is in the episode.
More from At Barron's
-
Moderna CEO on the Biotech’s Cancer Vaccine and Pipeline -
Anthony Scaramucci on Katie Zacharia, Trump, and the Midterms | At Barron's -
IMAX CEO on 'The Odyssey,' Chris Nolan, and Company's Sudden Ascent | At Barron's -
PG&E's CEO on Data Centers, Energy Rates, and the Stock | At Barron's -
'Stay Calm' Author Discusses the Science of Investing | At Barron's -
'Stay Calm' Author Discusses the Science of Investing | At Barron's
At Barron's, this episode and its artwork belong to At Barron's. Octagon is not affiliated with the show; this page is an independent summary.
