Artificial Intelligence: Why Experts Disagree About its Future State
Tony Roth: Chief Investment Officer
Neil Sahota, CEO ACSILabs
NOTE: Data in this podcast are as of April 28, 2023.
Tony Roth: This is Tony Roth, Chief Investment Officer of Wilmington Trust. You're listening to Capital Considerations, and we're back with Neil Sahota in our second half of our conversation about artificial intelligence. You can go back, by the way to the first episode for Neil's credentials, which are profound. I just want to say a couple of them though.
He is the CEO of ACSILabs where he focuses on cognitive research and technology development and improvement of human thinking and a slew of other really impressive credentials. But he's deeply into the ecosystem of AI that is occurring both within government and within the private domain, which is, I think, why we picked Neil to be someone to help us with the understanding of the definition of AI, the possibilities and the limitations. So, I think we've done a good job sort of laying out what AI is and isn't. And we ended the last episode with your example of social media creating an echo chamber, and the idea that one of the things that AI could do is lead people to believe in fake news.
That's something that applies to both sides of the political spectrum. Remind everybody that Wilmington Trust is nonpartisan and our goal here is to try to suss out some of these risks, not to take a political stance on any individual risk. Let's get into that letter again. What it signifies to me more than anything is how fast things are changing.
And so how fast do you think things are going to change over the next five years, let's say? And do you think that in five years from now, there are, you know, 5% of our workforce will have been displaced because AI will be able to undertake what they used to do. In addition to that, our lives will be different, hopefully better because AI will be enhancing the quality of our lives in different ways.
What's the future going to look like as we move forward?
Neil Sahota: It’s going to be an even more rapid pace of change, Tony. I think, if I remember the latest World Economic Forum report, as of 2022, about 30% of tasks have been automated, like job tasks. By 2025, they’re projecting that 50% will have been automated. So you've got a near doubling in just a couple of years right there.
And if I remember correctly as well, the estimate was 120 million global workers will essentially need retraining because of AI systems picking up some of this work. We're not slowing down, we're moving faster because AI, generative AI particularly, has become the next big arms race, and not just for companies, but for governments.
So, everyone's trying to get out the head of the curve. Companies are trying to push things out quickly to establish market dominance, obviously get an ROI. But governments also realize that they have to jump out for thought leadership as well as somebody's going to weaponize some of this technology, we better figure out how that's going to happen so we have the proper defenses or we're, we're better at weaponizing it than they are.
Tony Roth: So, Neil, where do you see the clearest applications from a commercial standpoint, I suppose is probably the right way to start? I'm operating a university, I can fire all my professors and just put computers in the front of the classrooms.
Is it, I'm running an investment organization and I can hire robots to pick the stocks in our equity portfolios, or is it that I'm an accounting firm and I don't need accountants anymore? I could just load all the data into the machine and they can put together very complicated tax returns. Is it going to get to the point where I can roll a machine into the courtroom and I don't need the lawyers anymore?
Where does it stop or where are the greatest in the next decade applications?
Neil Sahota: I don't know if we'll go full automation across to that level. We'll still need some lawyers and accountants and some of those things, but we're already on track for some of these things. H&R Block has an AI system that does, I think, 70% of their tax returns now.
It knows every tax code, every tax law for every municipality, state, federal. Because of that, the average return prepared by the AI system has a refund of about $236 more because it can find some of those hidden or lesser-known deductions. You talk about the courtroom. Well, China is using virtual courtrooms because there's space constrained, but they've developed an AI arbitrator.
So they have a AI system that listens to 10,000 different arbitrations at the same time, and that's helped them clear out their court backlog. Actually, Estonia has AI robot judges doing traffic court. They can do about 90% of the cases because they're pretty standardized, speeding, or moving violation.
Tony Roth: So I have to be really careful not to drive on Estonia because when I get traffic violations I do try to find a way to weasel out of them and weasel is the right word, so I'm sure it would not work in Estonia.
Neil Sahota: Uh, you’re not going to get much sympathy from an AI robot judge.
Tony Roth: Okay. I talked about the inflation problem that we have. We have in the United States, 165 million workers, and I think these numbers are really fascinating, Neil, and we have a deficit relative to long-term trend of what's called two million workers.
It's that deficit of only two million workers across the economy. It's mainly not in the white-collar space, but in the blue-collar space that is driving wage pressures. We got the employee cost index, the ECI, we get it every quarter. We got it this morning, came in at 1.3% higher than expected. If you'd multiply that by four, that was for the first quarter, you end up with 5.2% for the year. Wages are growing at 5.2% a year because for every job opening, we have only half a person looking for a job, so it's driving wages up. All we would need to do is replace three or four million people, let's just say in the workforce, and all of a sudden we're on the other side of that equation and we're driving wages lower.
What do you think? What's the timeline?
Neil Sahota: It’s already happening, Tony it's happening not just because of wage inflation, but some of these jobs, people just don't want to stick around for the work. Look at McDonald's. They’re already converting their drive-through systems to use AI chat bots to take your order.
That's like one less person you have to staff, and the AI will capture the orders more accurately. It allows for more customization and the AI will flash on the screen there: Did I capture your order, right?
Tony Roth: As investors, we really need to keep our eye on this because it's going to fundamentally change the kind of problem that we're confronting from an investing and an inflation standpoint.
If it's going to happen as fast as you're prescribing, in a sense, or describing, and maybe our capital market forecast for next year needs to squarely take this challenge on of the impact of artificial intelligence on the economy.
Neil Sahota: The hottest skillset right now is prompt engineer, and it is a huge, massive deficit. Everyone is looking for prompt engineers. There's literally just a handful.
Tony Roth: What’s a prompt engineer?
Neil Sahota: Someone that works with AI systems like ChatGPT, and actually prompts the production of work. So, you actually have to be really good about what you ask for and making sure you're asking the right questions, otherwise, the quality of what gets produced by these AI systems is shoddy.
So you don't ask right, you don't instruct right, you don't guide right, garbage in, garbage out.
Tony Roth: So maybe one of the reasons the list of suggested questions to use today for our conversation was, so, I think the word I used was anodyne, was because we didn't really ask it in a very edgy way. We just asked it in a very generic way, and we got a generic set of answers.
Neil Sahota: Could be.
Tony Roth: You talked earlier about winning in this space. Who is winning in this space? And from a cybersecurity standpoint, should we be concerned personally that someone's going to be able to learn about me and figure out what my passwords are and take all my money? Or should we be concerned that I'm going to be on an airplane with the pilot, but they're going to shut down the FAA systems because they’re going to be able to figure that out through AI. Who's winning and what are the concerns for us as a sort of somewhat organized society?
Neil Sahota: Definitely Open AI right now is the frontrunner that is for sure.
Tony Roth: And they're a private company, just so everyone understands.
Neil Sahota: They started off as a nonprofit, but they actually switched to a for-profit a little over a year ago.
They're the guys that developed DALL-E 2 and ChatGPT, the stuff that everyone is using right now.
Tony Roth: So they're winning. And when I think about technology across the board, I think of the U.S. as being the leader hands down. I mean, there are areas like Nvidia, which is the Dutch company. They may be the leader in terms of developing, as I understand it, hardware microchips for AI. That's just one example.
Are there countries that you think are leading other countries? Is the U.S. at the forefront of this or are we behind on this relative to our dominance on—in other areas that we had established over the years?
Neil Sahota: The realistic answer is that the U.S. has not the forefront. China has actually done a really good job and actually feeling their AI kind of wave.
They've invested hundreds of billions of dollars, not just in companies and startups, but they've also made huge investments in education. They crank out 600,000 data scientists a year with PhDs.
Tony Roth: Wow.
Neil Sahota: China, when it comes to like AI and health care is really advanced because they basically collected all the data and created a repository to make it available for all these companies to use and train their systems on.
Nobody else really has anything like that. China's up there. Singapore has made huge investments as well. So has the UK, Canada. Has even turned Montreal and Toronto into municipal hubs. The training, the funding, the promotion of these companies, it's not like the U.S. is way, way behind, but we're definitely not the forerunner when it's coming to these things right now. Not across the board at least.
Tony Roth: Would you infer that our defense apparatus, if you will, is behind and at rise relative to whether it be China or Russia, et cetera?
Neil Sahota: It’s possible. They don't have a full look behind the curtain, so to speak. The U.S. government is doing a great job of trying to keep on top of this, the DOD and all that has made huge investments into this, but it's tough to say. I think the news just came out yesterday. They found one of the Russian drones shot down. They were looking at it. They found all this proprietary technology, which then they realized they think they got it from Iran who had captured a drone like five years ago.
So, I think there was some surprise by this level of sophistication of this Russian drone as a result. As we know, it's an never ending arms race, especially when it comes with government versus government, but I think the bigger challenge is that we definitely have a talent deficit when it comes to exploiting this technology.
Tony Roth: Let’s just pivot back to the areas where we can see the most dramatic and beneficial applications for a moment. So we've talked about accounting and we've had a conversation in the first episode about autonomous driving and such, but systematically, where do you think the best applications are for the greatest benefits, if you will, for humankind around AI in the next 5 or 10 years?
Neil Sahota: If we're talking about like social benefits, then health care, food production/food safety, and education are really the top three areas. I've worked with farmers in like Bangladesh, Malawi, using some of these tools and using less water top soil, that kind of stuff. They're able to produce about 30% more crops.
So the ability to grow an abundance of food, create more cash crops to bring more money in the community is totally feasible. We're already putting out more educational tools like AI tutors. People that don't have access to, universities or, even a good K through 12 school. There's some things we can do to supplement that so students have an opportunity to learn and, and learn things that they would like to learn without having to rely on having teachers that are knowledgeable and experienced in some of those subjects.
Then, of course health care, it’s a global challenge with the aging population. But we've been able to take even very rural hospitals and upgrade the quality of care to that of like a big city by supplementing some of these AI tools for doctors and nurses. Not replacing them, but giving those those tools.
So there's a wealth of benefits that can be created in some cases are being created. They're often though not the sexy stories we tend to hear in the news.
Tony Roth: It's really important to recognize that there are so many places where AI is impacting our lives already that we don't recognize. Health care, I think also, is one of the most interesting and richest areas from what I've read and understood.
And diagnostics specifically, is one of the applications within health care, whether it's collecting a set of information and identifying early on an illness or disease, et cetera, that a patient may have. Or another application is radiology. A friend of mine, he’s a radiologist said, you know, I'm glad I've only got another 10 years of my career because I can't compete with some of these computers that can come in and can read a film and see things better than he can, already.
Neil Sahota: And again, we're not trying to replace everybody. But we've already seen AI systems that can detect lung cancer from x-rays at a 90% success rate. Whereas you benchmark against the human, it's 50%. That's marginally better. But I think that's one of the things we're looking at. Look at pharmaceuticals. You've got a lot of companies out there now investing heavily in AI drug discovery tools. You give it some information about the disease and other things, and it can formulate different molecules and bindings and then simulate which of these might be the most effective, or even at a precision level, might be dosage, might be the most effective for certain genders, ethnicities. But it's radically cutting down time and cost. It’s cutting down cost to a degree that diseases that are a little bit more rare, suddenly it's becoming cost effective to develop pharmaceuticals for them. So, our ability to actually treat more diseases will—should expand out.
Tony Roth: What’s so fascinating is that AI could do so much good in terms of performing tasks that humans would have to labor at and performing tasks that humans really aren't that good at or aren't perfect at, or couldn't do on their own.
Where do you think it goes in a longer period of time than five years? Are we at risk that we have a population on this planet of X billions of people and we don't have enough jobs for them to fulfill because so many things will be able to be accomplished by AI? Or is it like robotics that as the productive output of the supply chain increases and, and evolves, we'll find new things for people to do, to support it and to tackle new problems and challenges that come along?
Neil Sahota: It’s another inflection point. If I remember right, the World Economic Forum said that by the end of the decade, 97 million new jobs, jobs that never exist before, will enter the economy fueled by this technological growth we're talking about; global economy.
Tony Roth: Global economy, right.
Neil Sahota: Some jobs will get lost, some new jobs will get created. That's what always happens with, you know, these kinds of, you know, industrial revolutions if you will. The challenge that we have, I think, is twofold. One is obviously getting the future generations ready, but also retraining the current workforce.
And the problem we have is everyone thinks like, well we’ve got 15, 20 years to do this. The truth is we probably have four or five years to really start getting people ready for those new jobs. We know some people, unfortunately, will get left behind. The second thing though is historically throughout human history, every time we've had kind of this inflection point, people have been really good about rising to the challenge.
We're going to take on that next set of more complicated value add-type of work. This time around though, I'm hopeful we'll do that. We've also never had so much entertainment in our lives before. Like other things to do didn't work, and I call it 50/50 that will either rise and take ourselves to that next level, or we're all going to wind up spending our time binge watching Yellowstone.
Tony Roth: It’s a really fascinating observation and I was going to make a similar kind of observation, but I wasn't sure that where I take it, and it's simply that you can imagine a world where people don't need to work anymore, in a sense for a couple reasons. One is that AI is fulfilling so many human tasks and humans can anesthetize themselves and entertain themselves with consumption off of a small little screen.
What they need in order to be some acceptable level of fulfillment is potentially delivered in a much less extensive way. Which is scary in and of itself obviously, but maybe that's where we're going.
Neil Sahota: It’s almost like that movie, Wall-E, if you've ever seen it, how humanity kind of devolves because everything is taken care of by AI systems.
Everyone just kind of floating around in these chairs, eating and be entertained. Unfortunately, that's a future I could see. I hope that doesn't happen, but I can see it.
Tony Roth: A lot of concerns. A lot of worries for the children, right? The grandchildren. We're going to have to end here, but for me this conversation has really driven home the profound opportunity set for investors in AI and that we need to really be challenging ourselves as an investment organization to make sure that we are really fleshing out those opportunities and we're thinking about where the best entry points are for capital at each stage of that supply chain, if you will, of AI so that we're maximizing our opportunity, and that's what we're going to be doing at Wilmington Trust.
So, Neil, thank you for a really terrific set of conversations here. It's one that I'd like to repeat at some point periodically if you'd be willing, because I think that taking stock of the development and the pace going forward is going to be really important.
Neil Sahota: Thanks for having me on. Tony, I would love to be on with future discussions with you.
It was a great conversation and hopefully informative for a lot of people.
Tony Roth: It's enough for me at least to get really better oriented and I hope that our listeners will agree. So thanks everybody for listening today. Wilmington Trust.com for a complete roundup of our latest take on the economy and the markets and how we're responding in portfolios.
Thanks everybody, and we'll talk to you next time.
This podcast is for educational purposes only and is not intended as an offer or solicitation for the sale of any financial product or service or recommendation or determination that any investment strategy is suitable for a specific investor.
Investors should seek financial advice regarding the suitability of any investment strategy based on the investor’s objectives, financial situation, and particular needs. The information on Wilmington Trust’s Capital Considerations with Tony Roth has been obtained from sources believed to be reliable, but its accuracy and completeness are not guaranteed. The opinions, estimates, and projections constitute the judgment of Wilmington Trust as of the date of this podcast and are subject to change without notice.
The opinions of any guest on the Capital Considerations podcast who are not employed by Wilmington Trust or M&T Bank are their own and do not necessarily represent those of M&T Bank Corporate or any of its affiliates.
Wilmington Trust is not authorized to and does not provide legal or tax advice. Our advice and recommendations provided to you is illustrative only and subject to the opinions and advice of your own attorney, tax advisor, or other professional advisor.
Diversification does not ensure a profit or guarantee against a loss. There is no assurance that any investment strategy will be successful. Past performance cannot guarantee future results. Investing involves a risk and you may incur a profit or a loss.
Any reference to company names mentioned in the podcast should not be constructed as investment advice or investment recommendations of those companies. Third-party trademarks and brands are the property of their respective owners. Third parties referenced herein are independent companies and are not affiliated with M&T Bank or Wilmington Trust. Listing them does not suggest a recommendation or endorsement by Wilmington Trust.
Private market investments are only available to investors that meet the U.S. Securities and Exchange Commission’s definition of qualified purchaser and accredited investor.
Facts and views presented in this report have not been reviewed by and may not reflect information known to professionals in other business areas of Wilmington Trust or M&T Bank and may provide or seek to provide financial services to entities referred to in this report.
M&T Bank and Wilmington Trust have established information barriers between their various business groups. As a result, M&T Bank and Wilmington Trust do not disclose certain client relationships or compensation received from such entities in their reports.
Investment products are not insured by the FDIC or any other governmental agency, are not deposits of or other obligations of or guaranteed by Wilmington Trust, M&T Bank, or any other bank or entity, and are subject to risks including a possible loss of the principal amount invested.
Wilmington Trust is a registered service mark used in connection with various fiduciary and non-fiduciary services offered by certain subsidiaries of M&T Bank Corporation including, but not limited to, Manufacturers & Traders Trust Company (M&T Bank), Wilmington Trust Company (WTC) operating in Delaware only, Wilmington Trust, N.A. (WTNA), Wilmington Trust Investment Advisors, Inc. (WTIA), Wilmington Funds Management Corporation (WFMC), and Wilmington Trust Investment Management, LLC (WTIM). Such services include trustee, custodial, agency, investment management, and other services. International corporate and institutional services are offered through M&T Bank Corporation’s international subsidiaries. Loans, credit cards, retail and business deposits, and other business and personal banking services and products are offered by M&T Bank, member FDIC.
© 2023 M&T Bank and its affiliates and subsidiaries. All rights reserved.