Governments across the globe are examining AI and AI regulation as 2023 made it clear that businesses of all kinds are making heavy investments and advancements in AI technology. While the White House has put out a "Blueprint for an AI Bill of Rights," there are still questions around how effective the US government and governments around the world will be at regulating AI.
Tom Davenport, Babson College President’s Distinguished Professor of Information Technology and Management, joins the Live show for Yahoo Finance AI Revolution, to discuss adoption of AI and AI regulation across the globe.
When asked about the US and its potential for regulation, Davenport comments: "How do you transform the current operations of the US government? That's an issue we're all wrestling with every day. As I say, I don't see that happening at a federal level, and it will, I think, be somewhat chaotic if individual states and even cities start to pass AI legislation. So I kind of think regulation in the United States is a lost cause, but I do think there are some potentials for the AI act having an influence on US-based companies. "
For more expert insight and the latest market action, click here to watch this full episode of Yahoo Finance Live.
- Generative AI was center stage last week at the World Economic Forum in Davos. Now leaders must focus on turning the hype and investment from 2023 into actual value all while governments examine just how much regulation is needed for the new technology.
Here to discuss where AI goes from here as part of Yahoo Finance's theme week, the AI revolution. We welcome in Tom Davenport, president's distinguished professor of Information Technology and Management at Babson College.
Thanks so much for being here, Tom. So I would start with how this all becomes real. A lot of the talk that we heard last week was a lot of big talk about how much Gen AI is going to play into these businesses.
But I'm looking at actual deployment and monetization. Where are we in that cycle and how are we going to see that play out?
TOM DAVENPORT: Well, that's the right place to be looking. And you wouldn't see very much if you're looking in that direction. I did a survey at the end of last year sponsored by Amazon Web Services.
And it suggested 6% of companies have any production deployment of generative AI. So lots of talk, but not much action yet.
- And when you survey these chief data officers, Tom, these CDOs, and they talk about AI granted, obviously, as you said, it's early-- it's early days early beginnings here. But what are the benefits, the potential priorities, but also the roadblocks that they call out to you?
TOM DAVENPORT: Well, I think the primary activity now is in individual experimentation. And that's going to be hard to measure the benefits. And I think they're not ultimately going to be terribly large.
If you look at use cases like customer operations, call centers, contact centers, I think there is likely to be a substantial amount of value. Automated code generation, I think is going to yield a lot of value. Although that's going to take some time I think to really settle into development processes.
As you can see on the chart, R&D, none of the above are pretty low level. So a lot of potential opportunity. The biggest I think barrier now is in the data and how organizations get their data ready for generative AI.
- And how do they do that, Tom? How do they get good data sets? Because there's only-- I know there's an enormous quantity of data being generated every second. But is it the right kind of data on the right kind of information?
TOM DAVENPORT: Well, most organizations these days are interested in how do I train a generative AI model to deal with their own data and their own content. And, unfortunately, they haven't really curated or managed that terribly well over time. So the organizations that did this successfully early on companies like Morgan Stanley, they have a small army of people curating their investment reports and making sure they're accurate and up to date and not overlapping with other investment reports.
Because generative, AI as we know, if Fed poor quality information is going to give poor quality outcomes.
- And as you talk to these folks, Tom, across different companies, you survey these chief data officers. So they're experimenting-- at least experimenting with the technology. It's against the backdrop, Tom, of regulation.
And I'm curious how you see regulation of this technology kind of playing out at least here in the US. And do you do you believe Tom that American regulators and lawmakers have the background, the experience, frankly the tech chops to regulate AI responsibly and effectively?
TOM DAVENPORT: What do you think? I think we all know that they don't.
- I've seen the hearings, Tom. This is why I'm asking because you watch these hearings. The tech execs get dragged up before Congress to often becomes sort of embarrassing for frankly everybody that's involved.
TOM DAVENPORT: Yeah, I think that's mostly a show. You know, Congress can't pass a budget. So they're not going to be able to pass meaningful AI legislation.
I think the only good news for those people who want more regulation, and I am among them, is that I think it's likely that some aspects of the AI Act in the EU will be adopted by companies globally. Large companies have to do business in the EU no matter where they're based.
And so I think many of them will adopt some of the AI Act regulations. But there's also the issue of how do we enforce it. There was an interesting report in The Wall Street Journal this morning suggesting that one of the few pieces of meaningful legislation we have in the US comes out of New York.
It's called local law 144 that regulates the use of AI in recruiting and hiring. And this regulation is basically not being enforced at all. Something like 18 out of 400 companies even did any sort of audit at all on it.
So we have to put a regulatory infrastructure in place in addition to just passing the laws if we're going to be successful with it.
- Well, and, Tom, what do you think the urgency level is here on this? Because we can joke about Congress not really being up to the task. But, you know, I take the example of social media, which was not regulated and exploded.
And I think one could argue there have been pretty darn negative consequences from the lack of proper regulation there. So how do you get this done? How do you light a fire? How do you make sure it's done properly?
TOM DAVENPORT: You know, how do you transform the current operations of the US government? That's an issue that we're all wrestling with every day. As I say, I don't see that happening at a federal level.
And it'll I think be somewhat chaotic if individual states and even cities start to pass AI legislation. So I think regulation in the United States is a lost cause. But I do think there are some potentials for the AI Act having an influence on US-based companies.
- Tom, it's a big important topic. Thank you so much for joining the show today and helping us walk through it.