© 2024 M&T Bank and its affiliates and subsidiaries. All rights reserved.
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), Wilmington Trust Asset Management, LLC (WTAM), 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. 
M&T Bank Corporation’s European subsidiaries (Wilmington Trust (UK) Limited, Wilmington Trust (London) Limited, Wilmington Trust SP Services (London) Limited, Wilmington Trust SP Services (Dublin) Limited, Wilmington Trust SP Services (Frankfurt) GmbH and Wilmington Trust SAS) provide international corporate and institutional services.
WTIA, WFMC, WTAM, and WTIM are investment advisors registered with the U.S. Securities and Exchange Commission (SEC). Registration with the SEC does not imply any level of skill or training. Additional Information about WTIA, WFMC, WTAM, and WTIM is also available on the SEC's website at adviserinfo.sec.gov. 
Private Banking is the marketing name for an offering of M&T Bank deposit and loan products and services.
M&T Bank  Equal Housing Lender. Bank NMLS #381076. Member FDIC. 
Investment and Insurance Products   • Are NOT Deposits  • Are NOT FDIC Insured  • Are NOT Insured By Any Federal Government Agency  • Have NO Bank Guarantee  • May Go Down In Value  
Investing involves risks and you may incur a profit or a loss. Past performance cannot guarantee future results. This material is provided for informational purposes only and is not intended as an offer or solicitation for the sale of any security or service. It is not designed or intended to provide financial, tax, legal, accounting, or other professional advice since such advice always requires consideration of individual circumstances. There is no assurance that any investment, financial or estate planning strategy will be successful.

From smartphones to supercomputers, and automobiles to advanced medical devices, semiconductors are at the heart of it all. To introduce us to the world of semiconductors, the power behind modern technology including artificial intelligence, Tony welcomes Dr. Mark Lundstrum, distinguished professor and chief semiconductor officer at Purdue University. Tune in to gain insights into the bedrock of our digital society and the boundless possibilities that semiconductors continue to unlock.

Unraveling the Power of Semiconductors: Enablers of Modern Technology

Tony Roth, Chief Investment Officer

Dr.  Mark Lundstrum, Chief Semiconductor Officer, Purdue University

Tony Roth: This is Tony Roth, chief investment officer of Wilmington Trust and you are listening to Capital Considerations. Thank you for joining today. We have a very interesting episode that is extremely topical relative to behavior we've seen in the markets this year, as well as the broader geopolitical narrative in the world, particularly as relates to the U.S. and China. And the topic is semiconductors. What we'd like to explore today and try to provide some really interesting and hopefully helpful clarity on is not just what are semiconductors, but why are we at a moment where it seems as though, even though they've been on our landscape, in our economy, in our lives for so many years now, they've reemerged as a cutting edge technology, as a technology with intellectual property associated with it, which has raised all kinds of difficulties among countries, particularly the U. S. and China. There has been a lot of focus on our own domestic independence, national independence from a semiconductor standpoint.

Here to help us today sort through the topic of semiconductors is Professor Mark Lundstrom, distinguished professor of electrical and computer engineering at Purdue University. He currently serves as senior advisor to the university's president and as Purdue’s chief semiconductor officer, which is in fact quite important because Purdue happens to be a major center for semiconductor research.

He previously was the dean of engineering at Purdue, and he is a member of the executive committee of the American semiconductor Academy. Dr. Lundstrom has asked me to call him Mark graciously today. Mark has also been working in the semiconductor industry, his entire career, and has been on the faculty at Purdue since 1980. So, welcome Mark and perhaps you could help me with the correct pronunciation of your seat at Purdue.

Mark Lundstrom: Thank you, Tony. It’s a great pleasure to be here. So, I hold the Don and Carol Scifres electrical and computer engineering and Don Scifres was a pioneer in semiconductor lasers that enable Internet communications.

Tony Roth: Thank you for joining us today. I'd like to remind everyone that any companies Mark and I mentioned in the course of our conversation are not an endorsement to buy that company.

We all know that semiconductors are these little pieces of silicon that add and subtract zeros, essentially and do them very quickly.

And we know Moore's law is essentially an idea that allows the speed or capacity of a semiconductor to double every 18 months, which is to say that, given that semiconductors have been around for many decades, they must be really fast now because they keep on doubling every 18 months, yet they’re still not fast enough.

So, we’ll talk about that today. What would you want us to know, Mark, about a semiconductor beyond what I've said in order to lay the foundation for our conversation today.

Mark Lundstrom: Maybe we should just define what exactly a semiconductor is to begin with. We all know what a metal is. Metals like copper, aluminum conduct electricity well. And we know what insulators are. Materials like glass don't conduct electricity. A semiconductor is somewhere in between. It's neither a good conductor nor a good insulator. So, what makes it so important? What makes it so important is that by introducing small numbers of other atoms into the semiconductor, we call these dopants, you can dramatically change the electrical properties, make the semiconductor more metallic or more insulating.

And that's the way we make electronic devices, by selectively putting these dopants in different regions of the semiconductor. That's the way we make transistors. Every silicon chip contains billions of transistors that are wired up to perform electronic functions to do computing ones and zeros to store data. Communicating with wireless communication, taking pictures, all kinds of things.

Tony Roth: Semiconductors allow us to essentially reduce things that may not seem like mathematical functions or equations. Essentially down to mathematical functions or equations. So if I want to drive to the store and pick up some milk and come home, there's probably hundreds of semiconductors transistors in the automobile that are somehow processing zeros and ones and converting my trip to the store to some type of process that requires zeros and ones to be computed, added and subtracted in order for me to make it there and back safely.

Mark Lundstrom: There's something like 1500 chips in your car. That's a typical car and they do everything. They do the sound system, the audio, they do the radar and collision avoidance. They do the cruise control, they do the electronic ignition. But the smartphone that you're carrying with you while you're in your car is just chock full of transistors, all of the computing and communication and the taking of the pictures and the text messaging and everything you do.

Tony Roth: Without getting too complicated, because I'm sure that I would get overwhelmed pretty quickly. Why is it that there's such a range of different kinds of semiconductors out there, if they're all essentially doing the same thing, which is adding, subtracting these zeros and ones? We have the advent of AI now, and we have these specific semiconductors that NVIDIA is very well known for.

Everybody wants to get access to these. They can provide AI development and then ultimately AI services and make money to the world. How is it that there's such a different range of semiconductors? And what is it about the particular types of semiconductors that NVIDIA is creating and offering to the world that's so different and special. And why can't everyone do that?

Mark Lundstrom: First of all, I should point out the most common semiconductor is silicon, which is highly purified sand, incredibly purified. The vast majority of all semiconductor chips are made out of silicon and the vast majority of chips do digital signal information processing, ones, and zeros, because all computations can be reduced to performing mathematical manipulations of ones and zeros.

So that's the vast majority of what's done, what AI chips do. But there are needs to do other things. Your sound system uses audio electronics, analog electronics, your solar cells, when you see solar cell arrays, when you're driving outside, that's a semiconductor device.

If you're driving an electric vehicle, you have large power transistors that are controlling the electrical current through the motor. So, there's a variety of different specialized applications, but by far the largest volume of chips do computing and digital information processing, and that's what NVIDIA has done. More and more of the high-end chips are doing very specialized calculations for artificial intelligence applications.

And these are highly sophisticated designs. A single chip might have hundred billion transistors. The design team to produce that chip might have hundreds or close to 1000 engineers working on it. The cost of just designing one of those chips is 1 billion dollars. If we're talking about the most sophisticated high-end chips for artificial intelligence applications.

Tony Roth: So, I read an article in the New York Times. They describe these Nvidia chips as GPUs, graphics processing units, essentially, and they were sort of the fastest chips or the most capable chips for handling mathematical calculations, there's got to be more to it than that.

In other words, what is going on within these chips other than just being really fast and be able to handle a large volume of zeros and ones that make them particularly suited for artificial intelligence?

Mark Lundstrom: For many years, the semiconductor industry produced generic microprocessors that could be programmed to do anything you want them to do. They could do artificial intelligence, but they could do the computing you need on your desktop or whatever else you need them to do.

But you mentioned that Moore's law and the doubling of the number of transistors, which is the measure of the power of a chip every 18 months. Well, actually, that's getting more and more difficult. The things are slowing down and what people are finding is that if you design a chip or a specialized application, not to do everything, but to do one particular type of calculation very efficiently, you can get dramatic improvements in the performance of the chip.

So, it turned out many years ago, the NVIDIA entered the market making these GPU chips, graphical processing units that were optimized to do the graphical processing. You knew that you need to present 3D images and all of this. Now, it turns out that the mathematical calculations that you used to do those kinds of calculations are well suited to the calculations needed for AI applications.

And because of that, these highly specialized chip to do these particular mathematical algorithms. These are the chips that companies like Nvidia are producing and more and more of this specialized design has become a key feature of the chip industry. Design the chip to perform exceedingly well for the specific application that you're most targeting.

Tony Roth: So, NVIDIA must be, by extension of what you're describing, one of the leaders, if not the leader in the graphical processing space. And it's really because of that they're so well positioned in the space.

Mark Lundstrom That that's exactly right. In fact, the demand for Nvidia chips is so high that people are finding it difficult to get them, which is creating opportunities for some other companies as well.

Tony Roth: Again, in this times article that was talking about, particularly in the, what I'm going to describe as venture space, many startup companies in the AI arena need access to these chips in order to be able to develop and then ultimately provide the solutions and services within the generative AI space and they're not able to get these chips access to these chips. And they often don't actually take the chip and stick it in their laptop. The chips are actually hosted by some 3rd party and they're going to get time on the chips allocated to them. Something like that. Is that right?

Mark Lundstrom: They’ll be off in a server farm somewhere and they're, they're accessing them through the cloud.

Tony Roth:  Can we expect that NVIDIA is going to continue to be the dominant player here, or will other companies catch up really quickly, in terms of creating these graphic process chips? And I assume that there must be many patents associated with them, such that they're probably pretty well protected, at least within the community of countries that follows patent law and respects those types of things.

Mark Lundstrom: They certainly have the lead, especially because they've been in this GPU market for some time and now suddenly the AI chip market is exploding and they're in the right place at the right time with the right technology. But I think you're going to begin to see other companies entering this field and one in particular AMD is already entering this field.

Tony Roth: They're Taiwanese, right?

Mark Lundstrom: Right, NVIDIA will see more and more competition.

Tony Roth: And are there any other cutting-edge applications of semiconductors that we should be aware of other than AI that are dramatically changing the landscape here for chip making, and for the importance of chips in our society, that's different than just what's been going on for decades where you get continually new, gradual evolution.

I sort of view this AI thing is really a major disruption, sort of a break. Are there any other things? Or is it really everything else is sort of continuing along as expected, but then you have this one area of AI?

Mark Lundstrom: AI is certainly what's driving things now, especially at the leading edge. And the impact of AI on our economy and everything is likely to be enormous over the next decade or two. But there's much more to this industry than the leading-edge AI technology. Chips are pervasive. They’re in all of our appliances. They’re thousands of them in our automobiles. We call them mature technologies. They're not the smallest transistors. They're not the most sophisticated designs, but they're so pervasive.

You have them in your smart appliances, your thermostat, your toaster, your oven, your refrigerator, everything is networked. So, there's more and more use of chips in a wider variety of applications. Companies that you don't think of as semiconductor companies are designing their own chips because it's, they're becoming the differentiating factor between one product and another.

So, that's why a number of years ago, Apple began designing its own chips for its laptops. Why every automobile manufacturer has announced that they're going to be designing their own chips, too. So, we're going to see more and more pervasive use of chips, but not necessarily the leading-edge AI chips.

Tony Roth: Are there major shortages right now in areas other than AI chips? Of course, we had this one disruption of the pandemic where the supply chains were all messed up. That's been largely ironed out. In fact, we have an excess of certain kinds of chips like memory semiconductors and such where they sort of store the information in ones and zeros.

They're not necessarily computing it. But are we are we seeing continued shortages there and what can you tell us sort of around the state of play in terms of the United States’ participation in that industry. I think of Intel as being the 1st maker of chips that showed up in Microsoft computers or IBM computers, driven by Microsoft software back in the late 70s and early 80s. And now most of the chips come from Taiwan. So, it seems like even away from those very specialized AI chips, perhaps we have some work to do as a country to try to create less dependencies there. Maybe just talk to us a little bit about what you're seeing in that sense.

Mark Lundstrom: The chip shortage that we saw during the pandemic has largely been alleviated. What we're seeing now is a shortage of these high end ultrasophisticated AI chips. But in other chips, as you mentioned, memory chips, for example, right now there's a downturn or a pause in the economy and some of those manufacturers are experiencing difficulties right now.

So that doesn't seem to be in general and acute shortage of chips right now, but there's an expectation. This industry has always been cyclical. It's about a half a trillion dollars right now, and by 2030, it's projected to be 1 trillion. So, there's an expectation by the companies that are building factories that this demand is going to continue to rise over the next decade and beyond.

We've been talking about the leading-edge and this is one thing I've seen over the course of my career. When I began my career, many companies were building their own chips, but as they became more and more sophisticated and more expensive to manufacture and produce, fewer and fewer companies could afford to be in that leading edge game.

Intel was always the one that was leading the leading edge. But now we're left with 3 companies in the world that can produce leading-edge chips. It's Intel, DSMC in Taiwan and Samsung in Korea and actually, Intel has lost its lead in the leading edge and is working very hard to regain it. Over 90% of our leading-edge chips now come from Taiwan and that's a serious supply chain vulnerability for us.

Tony Roth: Even away from the idea of having a formative role in the new chips that are generating AI, just in terms of the more mundane chips, if you will, that are the advanced chips, we're not really as a nation, at the cutting edge of either design or actually production. How much progress do you think we'll make in the next few years, next decade, given the CHIPS [Creating Helpful Incentives to Produce Semiconductors] Act and the support that the CHIPS Act has provided for foundries to be developed and undertaken here in the U. S.?

Mark Lundstrom: The semiconductor supply chain is. Is very broad, so design is part of that. And the U.S. really leads the world in chip design. That's a key strength of ours that we don't want to lose. What we have done over the past decades is outsource the manufacturing of chips to other countries. The CHIPS Act is a big part of it is about reshoring some of that manufacturing.

Now, we have, I think it's 12% now and declining. We're never going to be able to re-shore a majority of our chip manufacturing, but we need to re-shore a significant chunk of it just to harden our supply lines. So, we may end up in a decade or more with 20% or so of our chip manufacturing done domestically, but we're going to need these partnerships with like-minded countries, like Taiwan, Korea, Japan, Europe, and, you know, to form these robust supply chains that we will need. We can't do it all ourselves.

Tony Roth: In addition to building, it domestically, we also want to diversify it away from Taiwan, obviously, given the relationship between Taiwan and China. So, I think that's just an obvious observation.

Mark Lundstrom: It’s also in an earthquake prone zone. So, there’s some vulnerabilities there.

Tony Roth: And we talk about the U. S. being a leader in chip design still, that’s hardening to hear, as a patriot. NVIDIA is a Dutch company. It sounds like NVIDIA is the leader in terms of all these AI chips and where's the locus of activity, what companies are the leaders in design that are U.S. companies?

Mark Lundstrom: Well, AMD would be one.  Qualcomm is another one. I guess that's another leader in this space. Apple is doing an awful lot of its own high end chip design. The tools that are used to design chips to computer aided design tools are really ultra-sophisticated and only a few companies in the world can produce these computer aided design tools that are needed to design 100 billion transistor chips. The U. S. is very strong there, synopsis, cadence, mentor graphics, most of its operation in the U. S. So, we have strength in the design space. The other place that we have strength in is the inside these factories that produce chips are ultra-sophisticated manufacturing tools.

And we're very strong in that space as well, in terms of these tools. Some of our partners are very strong. For example, there's a company in the Netherlands, ASML, that is the only company in the world that can make this extreme ultraviolet lithography tool that produces these incredibly fine patterns that are used to make these very small transistors and nobody else is even close to being able to do that.

Tony Roth: If I recall correctly that the Biden administration was successful in working with the Dutch to prevent access by the Chinese to some of those tools that ASML is in the market with.

Mark Lundstrom: And it will be very difficult for China to compete in the leading edge without that tool. It will take some time to develop a corresponding tool of their own.

Tony Roth: And that's just one tool all the various tools that are U.S companies synopsis, et cetera.

Mark Lundstrom: Right.

Tony Roth: Those are also not available.

Mark Lundstrom: So you hear people talking, you know, if Taiwan were to be invaded. If we simply stop supplying spare parts to these sophisticated tools, TSMC production would stop quickly.

Tony Roth: Which would ultimately hurt everybody.

Mark Lundstrom: Right.

Tony Roth: So, we've crossed over into these issues around production and security. Is there anything different about the AI space specifically? So, we talked about NVIDIA being a leader, but we also have discussed it as almost an unexpected token or usage of this type of GPU chip. Is there something that is unique around NVIDIA's ability to design these chips that the U.S. is behind on? How do you think about that? The AI space specifically in terms of design and chip development.

Mark Lundstrom: Well, I think most of that NVIDIA design takes place in the U. S. The sophistication of these designs with these design teams that approach 1000 engineers, this is something it takes a while to learn and to be able to hone and do and NVIDIA has long practice in doing that. Other companies are entering this space and have done some of that as well. AMD is maybe going to be a key competitor.

Tony Roth: Tell us more about what you're doing at Purdue and where your activities are focused and what's happening there and what role you guys are playing within the ecosystem.

Mark Lundstrom: When you ask semiconductor CEOs, what is your number one worry about addressing the nation's semiconductor challenge? They will all give you the same answer, and its talent. It's where are we going to get to people for the factories we're building and for the design operations that we're expanding. There’s a critical shortage of talent in this country for engineers and other workers in these industries.

That's the challenge that we're addressing here. We're one of the nation's largest engineering programs. And we felt that we just had a responsibility to help the nation address this challenge. A few months before the CHIPS Act was passed before we even knew that it was going to pass for certain, we announced that we were making this a priority and we were launching something we called the semiconductor degrees program. So, this is a college wide program. You know, that engages electrical engineers, computer engineers, materials engineers, mechanical engineers, industrial engineers. We’re expanding it beyond our own engineering college into statistics and data science and computer science and other fields of the university. So, we, as a university, we've really made a commitment to be a major supplier of talent to this country and that's the key challenge that the industry faces.

Tony Roth: When you look, Mark, at the shortage that you're describing, obviously, there's a shortage coming out of universities that are producing the talent, is that primarily because as you've described it, there aren't enough cutting-edge programs, or is it because there aren't enough kids that are attracted to the space?

Mark Lundstrom: So that's interesting. Our program has been growing every year for the last 15 years. We set an enrollment record. And these students have no shortage of jobs. Something like 95% of them have a job when they graduate or shortly after.

Tony Roth: These are undergrads or graduate students?

Mark Lundstrom: I'm referring mostly to undergrads, but we have a large graduate program as well.

You know, I think our challenge has been as we've outsourced more and more of this industry and chips have become kind of invisible to people. They're just not on the radar of young people. They're not thinking about careers in this industry. So that's been the key question that people have had, you know, can we get students interested in considering, exploring careers in semiconductors and microelectronics and chip design?

We've had some positive indications over the last year. We're seeing students are paying attention. They're hearing about chip shortages. They're hearing about the CHIPS Act. They want to know more about it. And we're seeing a lot of interest in the new courses and programs that we're setting up.

So, for example, we launched a summer program to give students some hands-on experience over the summer and semiconductor chip design or chip fabrication and we were hoping we might have 50 students the 1st summer. So, we had 350 sign up. We couldn't do that many, but we've had a number of indications like that that that students are really paying attention and they're interested in learning more about what career options there are in this field. We're just at the beginning of what will probably be a 10, 20, 30-year effort to grow and expand and reenergize semiconductor manufacturing in the U. S. So, this is a great time for young people to get into this field.

Tony Roth: How many universities are there that you would describe as being, sort of at the top of the field in the space? In other words, are there 50 Purdues out there at 5 Purdues? How many students are being minted each year? What's the best way to think about the volume today versus where it needs to be?

Mark Lundstrom: When people go through these analysis of what is the shortfall going to be? I forget the specific numbers now, but the bottom line is we need to increase by between 2 and 3 times the number of semiconductor engineers that the nation's colleges are producing. That's the challenge that we're facing. It’s a particular challenge in this field. If you ask the number of large programs that can really address this challenge in a significant way, there are really only a handful, probably less than 10.

You need a critical mass of faculty that can teach it. It's such a broad range of skills and knowledge that are needed. You need expensive laboratory facilities for the students to be trained in. Only a few universities can afford those kinds of laboratory facilities.

We're looking at how we can partner with other universities that don't have the ability to train their students for these careers and maybe bring them to campus in the summer offer them various kinds of programs that we can work in partnership with them to provide an education that they couldn't provide by themselves.

It's a big challenge and it's probably not going to be possible to address this all with domestic talent. An awful lot of really talented students from all over the world want to come here and study, especially in our graduate schools. This is a competitive advantage for the nation.

You know, we need to make it easier for these students to stay here, have careers.

Tony Roth: We spend a lot of time in our investment group at Wilmington Trust thinking about what I call economic exceptionalism of the U. S. There are a lot of reasons why the U. S. has been the dominant economy since World War II. And they range from demographics, whether it be the pluralism of the country, immigration that we've had for many, many decades. It could be the strength of the dollar, the reserve currency of the dollar, the openness of our capital markets. There’s lots of reasons, but one of the key reasons really is the strength of our educational system. And I think that one of the pillars of that is, in fact, the pluralism that we see in the diversity of perspectives and the talent we attract.

So, Mark, can you talk a little bit about that and the pros and cons of attracting some of the smartest people and the most capable students that are out there, versus being able to then, if you will, ring-fence or circumscribe that knowledge that's created so that it doesn't just go back to whether it be China or some other place.

I mean, I would imagine you still have plenty of Chinese students. We’re open for business for Chinese students to come in and study in the U. S. How do you think about that? Is that a problem? Is it not a problem? What are the dynamics around that?

Mark Lundstrom: For years, we've brought some of the brightest people in the world here.

They become citizens, they've worked in our industry for years and years. They started new businesses and then sometimes they return back to their home countries later on. I had the opportunity to visit Taiwan about one year ago with the governor of our state on a state visit.

We were able to get 30 minutes with Morris Chang, the founder of TSMC, 92 years old.

Tony Roth: Wow.

Mark Lundstrom: Sharp as a tack. He spent many years in the U. S. I think the, the last 20 or so years was at Texas Instruments, and then he returned to Taiwan. He still thinks of himself as an American. He talks about us, our challenge.

He contributed an awful lot to our electronics industry. He's now a strong partner of ours in Taiwan. So, I think it's to our benefit to bring these people here. To keep them here if we can. If they return to their home countries to continue to partner with them and build those bridges.

I mean, there are some real challenges with China in particular that we're going to have to be careful about.

Tony Roth: And I think that there's probably an inherent limitation on their, on the ability of a student to go back to their home country, perhaps, and deploy their knowledge, because there's a lot more than that knowledge base required in terms of, or even capital. You need to have access to all the things we've talked about earlier.

And perhaps that also provides an inherent incentive for folks to stay local after their education, if we, as a country and politically have the right policies to allow them to do that.

Mark Lundstrom: And, you know, also, I think that a lot of our graduate programs have a high percentage of international students.

You bring these students in, they contribute to the research done here in America, you publish it largely in the open literature. They contribute to the strength of the research group and the added expertise and experience that the group has, that stays with us. If that student returns to their home country, our own research group right here in the U.S. has become stronger as a result of their contributions. And that continues to be here. We don't want to lose that.

Tony Roth: Absolutely, I mean, again, that's the pluralism that I see at least as being one of the major pillars of the dynamism of the U. S. economy and has been for many years. When you think about, just given your experience, seeing really the birth of a computer industry and the advent of digital across our entire economy, but now we're on this threshold, if you will, of generative AI, and lots of other various neural networks and other applications of AI. How significant to human society do you see this movement into AI as being and do you see it as overwhelmingly a positive development?

There's possibility for job loss and such as well, but tremendous advancement, whether it's medicine, diagnostics, so many other areas. How exciting is it to you to see this threshold we're at in terms of AI?

Mark Lundstrom: People have been working on AI since what the 1960s and it always seemed something in the future that was slow and incremental progress And then suddenly, when, you know, when I asked people in this field, what happened, why suddenly do we seem to have crossed the threshold and there's been a discontinuous step change and suddenly everything seems possible. What they point to is the fact that just Moore's law. The incredible steady increase in computing power and the increase in the available data to train these machines with. There weren't any really fundamental advances in new algorithms and things that change things.

It was this increase in computing and data storage capability enabled by semiconductor technology that led us to where we are today. Now when I think back, you know, when I started as a grad student, I had one of the first email addresses on the internet that I thought it was really cool to be able to send an electronic mail.

And we thought, well, the internet was something that might be of interest to a few, geeks and people like me. I had no appreciation of where it might lead. And, in some sense, we're sort of like that with AI. You can speculate and you know that there can be enormous enhancements to the quality of life that will come, applications to medicine and trying to address the main climate challenges, energy challenges, everything that we have. But just as we learned with the Internet, there are downsides to, you know. How do we balance this all?

It's going to be very, very difficult.

We've always been concerned whenever there's a new technology that loss of jobs. So far, I guess that hasn't happened. I have a real concern now that there's going to be a group of people that will be left behind and there'll be a group of people that will become super-efficient because they will be enabled by the power of AI to become 10 times more productive.

That's a challenge I think we're going to have to deal with. That might help us address the semiconductor workforce challenge. If we're not able to get enough people to design chips and manufacture chips, we may be able to make each of those people much more efficient. So that might be part of the challenge that we face in semiconductor manufacturing.

So, I don't know where it's going to lead. It's an exciting time. We have to be, we have to do this carefully. We have to think about what lessons we've learned with previous technologies and see if we can avoid some of the downsides and focus on the positive aspects that are sure to be enormous.

Tony Roth: One of the things to me that really resonates, that’s so interesting and it's something that I think about in a lot of different contexts of commercial activity and the evolution of our economy, is that when we think about different advancements and technology, we often think about making money. We often think about efficiency.

But we also often neglect to quantify or qualify the increase in quality of life that is not quantifiable in dollars and cents, et cetera. And AI, I think, could be one of those areas that while there'll be lots of money to be made and lots of efficiencies, there could also be lots of increases in the quality of life of human beings on the planet that I think is very important and exciting as well. I think that's a really important point that you alluded to there.

There’s a takeaway from the conversation for me. We are at a really interesting evolution point in the role that semiconductors play in our lives with the advent of AI and that while we have a tremendous production dependency right now on Taiwan, and we will have a tremendous production dependency for many decades on non U. S. sources. The fact that we continue to be a leader in the development and design of the semiconductors means that we have an important role to play and as long as we continue to have policy, like the CHIPS Act, continue to move it right in the right direction there. There's a lot of leverage that we have and hopefully, if the broader relationship with China can be maintained in a constructive fashion, these new usages of technology can be available and productive for everybody. And we just have to be smart around how we continue to put the emphasis on things like what you're doing at Purdue, developing more people, more engineers in a very pluralistic context. So, thank you for all you're doing.

Mark Lundstrom: Thank you. I enjoyed talking with you, Tony. And, you know, I just thinking back when I was a kid, Sputnik was launched and we had to look what happened. Look how we responded to that challenge. And Covid was sort of another Sputnik moment. And I'm optimistic. I think we're going to respond to this challenge just as we've responded to challenges in the past.

Tony Roth: Great to hear. Thank you very much, Mark. And for all of our listeners again, as always, WilmingtonTrust.com is where you can go for a full roundup of all of our latest thinking on the economy and the markets.

We will be with you again with another episode of Capital Considerations in coming weeks. So, thank you again, everyone.



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), Wilmington Trust Asset Management, LLC (WTAM), 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.


Follow Capital Considerations on your favorite podcast channel

Featured Guest

Dr. Mark Lundstrum  
Professor and Chief Seminconductor Officer
Purdue University 


Stay Informed


Sign up here to receive insights designed to help you succeed.

Sign Up Now

WTU Newsletter Card
WTU Newsletter Handler