Name: Brainstorm AI Singapore 2024: Growth, Wages, Jobs, And Wealth—Debating AI’s Economic Impact
Uploaded: 2024-07-31
Duration: 21 min 5 s
Description: Panos MADAMOPOULUS-MORARIS, Founding Managing Director of Industry Programs & Partnerships, Human-Centered AI (HAI) institute, Stanford University, Anant MAHESHWARI, President and CEO, Global High Growth Regions, Honeywell Moderator: Clay CHANDLER, FORTUNE

Panos MADAMOPOULUS-MORARIS, Founding Managing Director of Industry Programs & Partnerships, Human-Centered AI (HAI) institute, Stanford University, Anant MAHESHWARI, President and CEO, Global High Growth Regions, Honeywell Moderator: Clay CHANDLER, FORTUNE

Brainstorm AI Singapore 2024: Growth, Wages, Jobs, And Wealth—Debating AI’s Economic Impact

Conferences

Panos MADAMOPOULUS-MORARIS, Human-Centered AI (HAI) institute, Stanford University, Anant MAHESHWARI, Global High Growth Regions, Honeywell, Clay CHANDLER, FORTUNE

So there has been um tremendous rhetoric about the advent of A I and what it means to firms, what it means to the global uh economy. And we have heard a huge range of uh you know, from techno optimism to techno doom is uh about, about a is impact. Uh you know, people comparing it to the industrial revolution, electricity. I've even heard people compare it to the discovery of fire, which feels a little over the top. Um And we've seen a number of really uh fascinating studies come out. I mean, uh the, uh you know, there was one that mit economist Darren uh as molo uh came out that was quite pessimistic. Uh mckenzie's had one that was a little more in the middle. Goldman Sachs has had one that uh has estimated that it could add as much as 15% to the global economy uh over the next 10 years, which seems extraordinary. Um So I'd love to, you know, think about those a little bit with you. And I'm, I'm just reminded of uh the seventies and the eighties when we had all these massive investments in, in uh computer technologies and robotics. And there was this great conundrum among economists because as, as, uh you know, Nobel Prize winner, Robert Solo observed in 1987 that you could see evidence of the computer age everywhere, except in the productivity statistics. It wasn't really showing up. Well, eventually it did show up. But I, I wonder whether, you know, they called that moment the productivity paradox. We're seeing a similar productivity paradox uh with A I and both of you are kind of perfect people to address this topic. Uh You know, an not you're right on the front lines of this uh at Honeywell, tell us a little bit about what you think about these kind of wildly disparate macro productivity uh estimates. Yeah. And maybe just to start optimism and pessimism can be seen both ways depending on which side you're looking at it from. But just a quick context of where I come from, we operate in three mega trends on the planet. That is the future of aviation, automation and the energy transition. So my comments will be centered around that. But these are real world industries. These are, these are really people who operate aircraft, people who are operating buildings and plants, manufacturing plants and people who are in the in the energy business. Now, we recently did a study across 1600 people, executives across different companies in 12 countries, with each of them having 1000 people at least. And we realize that nearly 17% of companies have actually implemented scale sort of projects which are related to productivity primarily because that's the number one play that people are seeing around productivity and efficiency. There is of course another play which is around cybersecurity where people are deploying a lot of the productivity I would say is really around saving cost of energy saving materials and getting more with less, say a little bit about your, your role uh at, at the Stanford Center and, and love to hear your take on this kind of, you know, paradox. Yes, thanks for having me Clay and excited to be here. Always excited to be in uh uh Southeast Asia. So yeah, at, at Stanford, the uh Institute of Human A I, we are short of the umbrella organization for all things A I encompass and we are very committed in uh uh advancing um A I policy research education and practice underscoring practice to uh accelerate human condition as part of this. Uh We work with a number of uh Fortune 50 companies uh help them accelerate the I journey and this means different things to different people. So there is a been indeed um a lot of uh skepticism around productivity or Roy, a number of reports came out uh both from uh uh analyst in Wall Street, but also in, in our world in Silicon Valley. Um But also economists are following this very closely. You mentioned productivity paradox uh uh Eric Bloo and my colleague at DRE has been doing work on this topic for the last 30 years or so. Um Let me unpack. Uh let's have a couple of things here. We should be unpacking one in terms of the concerns. Yes, there are some really credible concerns. Um People are looking at the Capex. Um This Goldman report mentioned, you know, a trillion dollars in Capex uh built out, built out infra hyper scale is alone, they are committing 60 to 80 billion this year uh in excess of the already committed Capex for uh for research. And then um in start up land, you have unlimited uh amount of capital uh getting into A I start ups and we don't have at the same time that many start ups in the 100 million territory. So people are skeptical of what extent to those start ups can live to the valuations. They uh they, they raise money and then for capital allocators, there's a lot of pressure because companies now uh stay in private for a long year. And with scaling laws advancing uh the hyper scalar and the magnificence seven MN AM N A will become harder and then it will be more challenging for them to build sustainable modes. So two things on our end, having worked with a number of those fortune 50 companies over the course of the last three years. One, I think people on more of the camps that we are um underestimating uh the long term impact of A I at scale and overestimating the short term. So that has created realistic expectations when it comes to Roy, both in terms of revenue or productivity or whatever you want, whatever way you want to measure it. Uh It's one thing to see public interest spiking and just open A I went to 100 million users in two months, but I was reading the other day when 20% of us retailers, it took them 20 years to get online. And there are multiple use cases, you know, apple revenue curve, sales force revenue curve. Many large organizations still uh they are not in cloud properly. So expectations versus reality is one like those things take time and second is what lies beyond tech. So most of the work we do with companies, it's not really on, you know, let's buy this tech or let's build this model or let's dump some A I to just replace the existing workflow. We feel that most of the value will come by thinking A I native, which means how you really re invent your business model, how you revisit your processes, uh how you upskill and rescue your people. This is where most of the value will come from, that will take more time. And I think even Goldman in this report, recognize the railroad analogy, you know, the first wave and the second wave. So we have more of this confidence Well, you said something this morning that I thought was quite insightful and a couple of other panelists on the main stage this morning have mentioned it, um which is the difference between using A I uh to achieve incremental savings uh on your, you know, the the production process that you've always used versus saying, OK, now we have to completely re invent the production process around the new technology. And we definitely saw that in the, in the computer investment waves where originally what everyone was doing was fiddling at the margins but not rethinking how they do everything. And, and, and that this long term curve that you're talking about on us where the real gains kicked in was when people finally said, oh, you know what, we got to rip up the whole legacy system and rethink it from the beginning. Are you seeing that with, with sort of the firms that you work with? And so what, what does that imply about how long the kind of payoff horizon might be for this technology? So maybe again, putting in context of very three different kinds of industries and use cases as you say, and I'd really like to do a poll in this room if you would fly an aircraft, if I said I'm not going to have the pilot in the aircraft. And, and that's the, that's the aircraft that you're on for 16 hours or eight hours and that will take you to your destination. There would be a few adventure seekers here. But uh but I would imagine a lot would say, you know, I'll wait on it a few more people do it and let the regulators approve it before I get to do this. Now, this is an industry that is short of aircraft and short of pilots in the air. Right now. Today it's regulation two pilots in the, in the aircraft to fly. You will it get down to zero? Maybe not, but can it get down to one with the power of A I and therefore have a lot more aviation going? That's possible. And that's where I think productivity and the workforce really really comes in because you can get a lot of input and truly A I becoming a copilot to, to that pilot. Now, the second industry, if I go and this is again a thing that when people are not talking about A I, they're talking about sustainability and uh and the energy transition. Now, if you look prepa to today, actually, the number of jobs have gone up in the energy industry because A I and the need for energy is accelerating the, the energy transition, there's a lot more renewable, there's a lot more uh different kind of green electron coming along and it's short of people. So when you're short of people, whether you're in the United States or Germany or Japan, then how do you, how do you augment that workforce and you again, bring A I to help you augment that, that workforce. And if you go down to the developing part of the world, they need to train up a lot of people for new industry. And automation therefore, is coming in. And that's again where you need A I to help that labor come up the learning curve very quickly. So a lot of training a lot of simulation, a lot of testing where the system might fail and therefore training somebody to do that. So, so overall, I would say there is definite use cases of productivity of workforce augmentation of training and enhancement of savings and energy that, that are, that are satisfying the CS in every company to say, you know, there is something in it, you know, it's part of the, the, the question about productivity gains is in a way two questions, I mean, it's a question about uh you know, how do we, how much do we think A I is going to be able to reduce labor costs or the cost of total production? Um But then I guess the other part of the question is what percentage of tasks or jobs do we think that A I can really be able to automate? So if you think about your airplane example, um I'm not sure I would feel comfortable getting into, I, I would get into a robo taxi but would I get into an airplane that didn't have two pilots. I might not. Right. Just because the stakes are so much higher. But if you think about what already happens in the cockpit at most airplanes already, maybe nine percent. I think people don't realize that when you get on a plane, pilots aren't touching very much until the actual, you know, the take off in the landing. Right. They're focusing on those really critical tasks that, that still haven't been automated. So I think there's probably a lot of equivalent to that in the production processes of different uh companies. Uh you know, Darren argues that mit that in fact, the number of tasks that can be automated is probably not as great as people think. And, and so that's where he gets, you know, he says probably if we get a productivity gain, it will only be 1% at 1% point a year rather than these crazy, you know, uh two and 3% kind of estimates. Um What can we talk a little bit about the, the job I impact uh of this sort of this transformation, you know, um the um you know, the famous uh William Gibson quote is that the future is already here. It's just unevenly distributed. And I think that, that there is a little bit of that associated with, with A I is that um uh the the fruits, the benefits productivity gains of this are going to be different in different parts of the economy and for different types of jobs. Uh How do you, how do you see that panel? I know you're thinking about this with your industry, uh you know, contact, we are thinking about it a lot indeed. And in multiple dimensions, um you know, there are a lot of parlors in, in A I so organizations are trying to figure out how, how do they manage in a world where alpha is so asymmetrically distributed. So there's a big delta between how quickly uh research is moving and then the time that organizations have to effectively respond and how, how they respond to this in our view is gonna define pretty much wins and losers for the next, for the next cycle. Um So there are a few things. Um I, I think this year we've seen a lot of things functionality uh from Executive suite and boards. Uh Most of the transformation journeys are led by either the CEO or the CIO um boards are very, very actively involved. As a matter of fact, I was in New York before the, I ran a poll in a room like this with uh with CIO S and I asked them, how many of you are you expecting to present in every board? And the vast majority of them told me every board we expect to go in and uh report progress on um on our journey uh On the A I side. Uh Second is, uh I think alliances and ecosystems as a as a mode uh because things are moving so quickly, there's so much internal and external pressure to leader to, to executive leaders, costs are still very, very high. Uh And human capital is here, I think that has pushed the envelope a little bit and uh expanding the surface area of creativity in terms of business models. So even organizations that were traditionally very conservative, they wanted to do everything themselves. Now you see those like global uh alliances, ecosystems and both of us spend a lot of time around the world, actually managing those ecosystems. Uh which I think bottom line is great for the organizations is great for uh the discipline like the A I discipline as a whole. You see the rise of cooperations because of this many companies at the same time, they might be competing or they might be partnering. And one last thing I would say is that uh yeah, alpha will be as symmetrically distributed at least for the foreseeable future. But leaders in their own industries can identify and curate their own pillars of alpha because Alpha has many different colors and something that we've seen again and again, in particular, the last six months is uh the human Centricity of how uh increasing awareness from executive leadership in different organizations or how they build design, deploy at scale in a human surgery way can deliver, you know, Alpha uh both internally and externally on the safety side on uh on the design side, on the implementation side. Um And I think as a matter of fact, we're gonna, um we're gonna see find ourselves uh in a few months of those transformation journeys, they will be measured. Their success will be measured by humans how and that connects to your argument about jobs uh and displacement. We got a few minutes. Uh Love to take questions from the floor if uh anyone has them. Um Yeah. Yeah, we get a, do we have four mics? Where are there? He comes? Thanks. Uh Seth Dobrin. Uh You know, I think there's three things I really didn't hear, address that are important to this conversation. One is a lack of diversity and so much. So, you know, for instance, Panos like, you know, Bee who founded, you know, the Human Centered A I is, is really the grandmother of A I and no one ever talks about that. They just talk about the fathers of A I. Um And, you know, lack of global representation um which is really important because people talk about re you know, fine tuning models, but fine tuning actually breaks the model. So you can't do that when you look at other cultures, which can cause actual type of colonialism, a new type of colonialism because you're inflicting Western and Chinese cultures on the rest of the world. Um And then the short term impacts on jobs. Yes, we're gonna create new jobs but you're gonna have a transition period where you're gonna have a lack of job creation, you're gonna have people leaving the workforce. What are things that can be done and what is how, you know, how I found in the, in the near term, you know, what's going on through the research that, that you guys have done? Thank you. Is the question addressed to me exclusively or 1.5 minutes. So let me talk quickly on, on two of them and IV line about everything else. Yeah, I did is extremely uh committed in advising uh diversity. As a matter of fact, we had our five years anniversary uh a few weeks ago and we in our dinner in our dinner session, we have the godmothers of, of, of A I and I, as everyone go on youtube uh and what the conversation was illuminating. Um We are very actively on the research side trying to empower and enable this on campus for our grant programs. Um But there are a big issues that they are affecting diversity in A I which is the availability of resources overall. It has become extremely uh expensive to run major A I experiments on campus. So, um you know, academia is losing talent to the industry. Uh And we are committed both in terms of our public policy efforts, but also our work with the industry uh to reverse that as much as possible. Uh really quick on, on on the impact of jobs um short term versus short term. Yes. Right now, we see certain use cases um that they prove that this is that the human and the machine are better where human and machines collaborate are better for organizations and better for society. Um We see that in customer support, we see that in diagnosis, there are a number of use cases and there are few papers coming out of Stanford mit other places that they prove that this, this uh there's this economic concept of complementarities where there are certain tasks that humans are better. Um emotional, emotional intelligence, creative problem, problem solving, contextual understanding. And there are other tasks like the ones that are predominantly uh repetitive in the nature and heavy on data that A is are more effective. So bringing those two together um we believe will amplify human creativity and um augment like we're very committed and physical augmentation versus replacement. Se second, talking about pilots is uh uh another economic principle we did the devis paradox. So when the Z engines were invented and technology dropped, uh increased productivity dropped the cost, the demand for pilots skyrocketed, it didn't went down. There was a lot of talk about um um you know, in the health care community uh about radiologists and how radiologists are gonna be the space. Well, guess what uh us cannot have enough. There are two X number of postings on the people who are actually available to work as radiologists. And that goes down to this physic where for some tasks, we're gonna still need humans in the loop. And we have agency, business leaders and policy makers, they have agency in making those decisions. Those are not decisions that will come embedded with products, but humans have agents in directing the strategic direction of the organization. We are unfortunately out of time. Uh But a lot of challenges, a lot to think about. I would encourage everyone to uh find both uh an A and uh and panels, uh you know, uh in the, in the hallway, in the lobby to follow up and, and uh explore more about these topics. Gentlemen, thank you very much. Thank you so much.