IBM (IBM) like many other tech giants, have put considerable investments into AI. IBM Chairman and CEO Arvind Krishna spoke with Yahoo Finance's Brian Sozzi and Akiko Fujita at the Milken Global Institute Conference to discuss the company's performance, its investments in AI, and how the rest of the world views AI.

Krishna speaks on AI regulation in the US: "I'm not concerned about the US falling behind. What I'm concerned about is that regulators should not regulate the technology. What regulators should regulate is the risk of the use cases. Can you imagine sitting back in 1995, let's regulate the internet, so lets regulate dial-up modems. How useful would that have been to today?"

He continues with: "Don't regulate the technology. All you will do is it will go offshore, these are digital technologies, but regulate the use case, so we are sure it's being used responsibly. One that I also talk about, is make people who develop models accountable for what the models can do. I think those are tools that are classic policy tools and can be wonderfully applied."

For more expert insight and the latest market action, click here to watch this full episode of Market Domination.

This post was written by Nicholas Jacobino

Video Transcript

You're talking a I you have to also be talking IBM.Let's welcome in IB MC O, Arvind Krishna here at the Milken conference Good to see you again.It's great to be here talking with Yahoo Finance and Water Energy and buzz here on a I here at the Milken conference.All right, well, let's stay on that buzz Arvin.How is IBM reinventing itself for this?A. I era Look a. I is a fundamental technological shift.We believe it will cost $4 trillion of total value to the industry, not all of that to the tech industry, but to the industry at large.All our clients want to tap into that.It's going to cause productivity.It's going to cause revenue gains.It's going to allow them to have more nibble enterprises.So what we are doing is building the A I.That's good for the enterprise efficient, open source in a way that our clients can control the IP and deploy it where they want.That's really what we are about.You talked a lot about that increase in productivity that's likely to come out of a I.Can you talk specifically in these early stages at least how your clients are deploying?A. I absolutely, I think there's three use cases that really, I think, bring life for us.So if you think of contact centres, you know typically, when people are doing customer experience, can you help that person become way more productive, so they might have to navigate amongst three or four screens?If you ever call the bank, you might notice the person on the other side is hesitating.They ask you, Can you hold on for 30 seconds?Well, that's because they're looking something up.Could a generative A I step in and simply get the information by querying the underlying myriad of systems?Can it make that person way more productive?By the way, productive means they get your query as an end user.Answer quicker.You're happier.You're more likely to get a correct answer.The A. I never gets tired or angry or snappy or anything like that.I think that's a great use case.I think that's going to be endemic.Everybody will use it second, quoting making quotas more productive.Everybody has tech debt.Can it help you get rid of the tech debt faster means that every enterprise is going to have more capabilities to offer their end users and their employees.I think these two without a doubt, the third one.We're seeing a lot.We call it digital Labour.In these days of tight job markets, demographics that point to less and less people of working age.Can I get a digital work to augment the human worker?I think these are three great use cases.All our clients are interested in them.We ourselves use them as the first client and it and it's giving us all of those benefits.Warren Buffett, at his annual meeting, talked about how the a I genie is out of the bottle, and he sounded concerned.I love all those things.Please tell me my big bank balance without getting snippy with me.I get it.Are you concerned that a I will have too much power?And what are the risks from a I I actually have no such fear, so I'm putting it black and white.I think there's a lot of people who hypothesise and as you have seen in many technologies, thinking that the rate of progress, which has been 100 times over the last three or four years is going to be another 100 to 1000 times.But just for the current A I we need this much infrastructure and power for it to be 100 to 1000 times.How much more infrastructure and power.I think it's a little.I use the term fantastical to keep predicting that it carries on on an exponential curve.I think it will get widely deployed.I think millions of people will use it across hundreds of processes.That will happen.But to imagine it's going to reach this point of super intelligence.I'm a bit of a sceptical on that topic.One of the big debates that's happening within the big players in a I is open source versus closed source.I know you're shaking your head because you're leading in Alliance, pushing for open source for those who are not necessarily following close source.We're talking about those like open A I as well as Microsoft Google.What is the risk of a closed source?A. I so I think, on any technology for it to get that escape velocity for it to really become ubiquitous.You got to balance between three things.Competition.You need competition because, as we all know, walled gardens tend to generally have very little innovation.So you need competition innovation.But you do need some safety.As you think about those three.I think open source is a great way to achieve a balance on all three points so open source is visible.Everybody can look at it.You can red team it if you want to.By yourself.You don't have to cross the bender to do all of the red teaming.By the way, as we all know, a I is built on data.So in effect, the IP you're capturing is the data you're bringing open source means that somebody can take a model that we might provide.If they add their own data to it, they can keep it.We don't ask for it back.If it's proprietary, in effect, you get it back.So I think it allows for a much faster proliferation while giving transparency back to those who want to get scrutiny.On that issue of transparency, there is a conversation about a I regulation.The US has executive orders in place, but not necessarily a framework for a federal regulation.You've got Europe leading the way.You just had Japan's leader coming in their own framework.Are you concerned about the US pop falling behind in that conversation?I'm not concerned about the US falling behind.What I'm concerned about is that regulators should not regulate the technology.What regulators should regulate is the risk or the use cases.Can you imagine sitting back in 1995?Let's regulate the Internet.So let's regulate dial up modems.How useful would that have been till today?I may not know how to use the computer, so I think that it's really to the point of Warren Buffett saying the genie is out of the bottle.Don't regulate a technology because all you'll do is it will go offshore.These are digital technologies but regulate the use case, so be assured it's being used responsibly.One that I also talk about is make people who develop models accountable for what those models can do.I think those are tools that are classic policy tools and that can be wonderfully applied.The black box.If you open up the box, let people innovate as fast as they can look at the last 18 months, how much innovation has come.There was open a I.Then there was mis.Now there's IBM.An 8 billion parameter model can do what a 200 billion parameter model can do.Let the innovation go.It will make the technology cheaper and ubiquitous.Marvin, it sounds like just listen to you here.IBM is working on some really cool and stuff on a I.Why do you think that the valuation on IBM may not reflect or be on par with some other A I place.So I'm gonna go cup helpful and then address your question directly.First, I think we've had a pretty good run over the last four years.Um, if I look at the last four years, I think we might have added 60 to 80 billion worth of, uh uh, total enterprise value.So I'd say that's good.Now that does not meet our one aspiration.So let me acknowledge that point.But I think that as we begin to make progress, we've got to be more consistent in our overall revenue and cash flow performance.I think our cash flow performance has been pretty consistent, but we've probably got to be a bit more consistent on revenue performance.And as that happens, I think we'll get both the multiple re rating, but also the growth that comes with the underlying cash flow and earnings growth.So I think those are the two pieces.I'm very confident our strategy is in the right place.We are building.We're staying in the places that we have credibility.We work with the Enterprise hybrid Cloud A I and Quantum down the road.And by the way, I don't think quantum or A I is really fully priced and into where we are.But I'll acknowledge Quantum is still 3 to 5 years out and a I is kind of in the early innings.I use a baseball analogy.A. I is kind of in the first innings of a game.Don't quite know how it will go is more like the fifth or sixth inning.So you have a pretty good idea.Not over yet, but a pretty good idea.We like sports analogies here, right?Let's finally talk about sort of bigger picture.You've got a good pulse on where the economy is given just the scale of your client base on your recent earnings call you talked about sort of a pull back and discretionary spending.How is that a pause?Is that a dry up?How would you characterise where that is?What does that tell you about the investments companies are willing to make, particularly in technology?So I'm not an economist.I'm a businessman and an engineer.Let me acknowledge that.But I see are spending and not spending.So I think that what is happening is really down to one reason, and I'm actually quite optimistic about both the US and the global economy, and I'll come back to that.Our interest rates are higher for longer than most people supposed in the middle of last year.So as you take that into account, if interest rates are 2 3% higher for the five and 10 year, then people are thought that means that all of the debt is going to carry an extra price.So I think CFOs are belt tightening against that now.That's not systemic.That's not really secular, but it is a trend that's going to be there for a year or two.If that makes any sense.I think that's the belt tightening.Look at anybody who carries tens of billions of debt.An extra 2 to 3 points is going to cost you an extra few 100 million per year.OK, let's look after what the Treasury spent.Let's not do anything catastrophic.Let's not make long term decisions.Would you go belt tighten against distress projects that's going on now?If I then look at demographic trends, I look at inflation trends, supply chains, cyber.All of that leads to Well, let's spend more on technology.So that's where you kind of see I want more software.I want more infrastructure.I want more a I. I want more cloud.But if I can, I'll kind of tighten on Labour based services.I think that's the two speeds we see and we see that in The software grew below 6% in the first quarter and consulting goes down two at C currency flat at acts.So we see that in our own business and I think it's a reflection of this macroeconomic frame.I wish we had some more time with you, but we don't.You're a busy guy.You're very in demand at this conference.IB MC, O Arvin Krisher Thank you for making time for your finance.We really appreciate it.Pleasure to be here with you guys.