The recent buzz surrounding artificial intelligence and mass layoffs roiling the technology industry is familiar ground for Ginni Rometty, whose nearly 40-year career at IBM culminated in her being promoted to CEO in 2012
Rometty, 65, also had to occasionally jettison employees in an extension of cost-cutting layoffs that began in the 1990s as IBM adjusted to waves of technological upheaval that undercut its revenue.
After retiring from IBM in 2020, Rometty spent two years writing “Good Power, " a book she describes as a “memoir with purpose." She recently spoke with The Associated Press about her career and the state of the tech industry now.
Q: In your book, you mentioned you graduated from Northwestern in 1979 with just $4,000 in student debt. What do you think of the current debate about student debt relief?
A: Whether or not we have debt forgiveness, the bigger issue is around the educational institutions. I feel strongly universities should not be the only pathway in this country. Fifty percent of good jobs in this country are over credentialed. They require a degree when you don’t really need one. Somewhere at the end of World War II, the American dream got attached to this idea that it’s college or bust.
We have to have more accountability for community colleges and colleges so they teach what the market needs. And I don’t mean hard skills, I mean the soft skills the market needs. And they don’t do that today because even if you get a degree you often can’t get a job.
Q: What are your thoughts about the current state of AI, especially with so much attention centered on Microsoft's use of the ChatGPT language tool?
A: I am a bit worried about that, I want to be sure we bring AI safely into the world. One thing I learned in the early days of AI is that this is a people and trust issue. It is not a technology issue. Because of how fast ChatGPT has spread, people almost immediately noticed it wasn’t always right yet it acted authoritatively and it did some things that our values didn’t appreciate.
You have to manage the upside and downside of the technology in parallel. And that is not what has always happened with technology. We have celebrated all the positives and then all of a sudden said, “Oh, oh, there are some bad things here.” I think this is our chance to at least be signaling to the public, “Hey understand, this has got downsides and upsides.”
Q: Is it important for governments to impose regulations on AI?
A: In fairness to lawmakers, do you think they really understand this? What we need is something I call “precision regulation” because I am afraid that in an effort to control AI we will completely inhibit the positive side of it. We will lose the upside as we try to manage the downside.
If you go to the doctor and say, “My finger hurts,” you don’t want to cut your arm off, right? My example of precision regulation is to regulate its use, not the technology. Talk about the areas you think it’s OK to use it in and the area where you think it should not be used in. I think it is impossible to regulate the technology itself.
Q: Have you been surprised by the magnitude of layoffs sweeping the tech industry?
A: I think you are seeing everyone reacting to the environment. Those that over hired (during the pandemic) are adjusting. I also think you see a reaction in this economy to what is being valued as not growth at any price. It’s profitable growth. You have to be efficient.
And so now I think for the very first time efficiency is entering the picture for some companies. It may be because the environment changed. It may be because someone attacks your business model. So what you are seeing is a recalibration reacting to the external environment.
Q: How do you think Elizabeth Holmes' recent conviction for fraud while she was running Theranos has affected the perception of women leaders in tech?
A: To me, she doesn’t define the future of women in tech. I consider that situational. I think there are things to learn from it, but I think it speaks more to the great hope that people have for technology. You don’t want to set the expectation so high that you can’t make it.