In this episode of Executive Connect, Melissa Aarskaug sits down with Irving Wladawsky-Berger, former IBM innovation leader, MIT research affiliate, and longtime advisor to global enterprises, to talk about what actually makes technology matter. Drawing on decades of experience through the rise of computers, the internet, and now AI, Irving explains why breakthrough technology only creates lasting value when it is deployed at scale across business and society.

This conversation brings a rare long-view perspective on AI, open systems, enterprise transformation, and the economic realities behind innovation. From ERP and e-commerce to open-source AI and self-driving cars, Irving makes a strong case for focusing less on hype and more on usefulness, productivity, and business results.

Chapters:
(0:01) Why this technology moment matters
(1:32) What leaders misunderstand about innovation
(4:36) Lessons from IBM’s biggest bets
(8:00) Why AI adoption still stalls
(13:42) What computers and the internet taught us
(23:22) Why AI hype misses the point
(29:08) Why open-source AI matters
(35:08) Where AI may create near-term value
(41:55) The leadership this moment requires
(47:02) Why real breakthroughs take time

Irving

(0:00) I think the number one leadership, in my opinion, based on my experience with computers first and then with the internet, is business and market leadership. (0:13) I spend a lot of time trying to figure out how to make companies more profitable, how to make them more productive, and don’t try to wow me with something that one person did that was great.

Melissa

(0:32) Every few years, technology is declared transformational, but the real question isn’t what’s new, it’s what actually lasts the test of time. (0:44) Today’s guest shaped the internet era, open systems, enterprise computing, and digital strategy, often before the rest of the world has language for it. (0:56) Irving Walowski-Berger is a research affiliate, MIT Salon teacher, a former IBM innovation leader of nearly four decades, and an advisor to organizations ranging from Citigroup to HBO to Mastercard.

(1:16) This conversation isn’t about trends, it’s about how technology and business truly evolve over time. (1:25) Welcome to the Executive Connect podcast, Irving.

Irving

(1:29) Melissa, it’s a pleasure to be talking to you.

Melissa

(1:32) Now, you’ve been involved in multiple major technology shifts. (1:39) What do you think leaders misunderstand most about how innovation actually happens?

Irving

(1:49) Okay, I would say if I look at my life, I’ve lived through now three major, what I would call historical transformative technologies. (2:05) The first one is the advent of computers, which happened before you were born, and in fact, I started being involved with computers this summer before entering college in 1962. (2:23) Computers were really just starting.

(2:27) The second one is the internet in the 1990s, which has been transformational like crazy, and now AI is the third. (2:38) Each one of these has been historically transformative. (2:45) I’m using that word going back to the industrial revolution, maybe the invention of the printing press in the 15th century, they changed every aspect of technology and society.

(3:04) My biggest lesson, if I had to be very (3:08) crisp about it, is the technology is really important, and clearly computers are really (3:17) important, the internet has been really important, AI now is really important, but what people (3:25) sometimes don’t realize or don’t take into account is they only have an impact when they deploy (3:38) across economies and societies at scale, and this deployment is very difficult, (3:48) and that’s the part that takes the most effort.

(3:52) Sometimes people say, oh no, we have AI, it’s transforming everything, we’ll have AGI, artificial general intelligence, maybe in three years, and my view is give me a break. (4:09) This AI is a young technology, it’s only been with us in its present incarnation for, I don’t know, 10-15 years. (4:22) That’s a child, so it will take, it’s hugely transformative, but it will take time.

(4:31) That’s my sort of overall feeling.

Melissa

(4:36) Yeah, so let’s talk a little bit about the lessons from IBM Big Bets. (4:42) Now, you’ve led company-wide initiatives around the internet, supercomputing, and Linux. (4:49) What separates the bets that worked from the ones that truly struggled?

(4:54) Ready to lead smarter and invest wiser? (4:57) On the Executive Connect podcast, we unpack executive strategies for wealth and influence. (5:04) Hit the subscribe button now, don’t just watch, act.

Irving

(5:09) Well, it’s how they are embraced by society, by the economy and society. (5:14) That’s what separates the ones that work from the other ones, and let’s take supercomputing. (5:24) We had supercomputers around in the 60s.

(5:32) I don’t know if you remember a company called Cray Research. (5:36) You’re probably too young to have heard of Cray Research. (5:42) What happened is that as microprocessors got cheaper and cheaper and (5:52) cheaper in the 1980s because of their use in personal computers, all of a sudden the research (5:58) communities realized that rather than having a very expensive mainframe with vectors or a Cray (6:10) computer, they can do very high-performance supercomputing by stringing together lots and (6:21) lots of these microprocessors, lots and lots and lots of them. (6:28) They would be much less expensive than these other big machines that had to be water-cooled and all that. (6:38) They were scalable.

(6:40) You can keep adding more and more microprocessors. (6:45) Parallel supercomputing became embraced in the research world in the early 1990s. (6:56) I was involved in IBM’s initiative to do that.

(7:00) Then they absolutely took off in the internet age because we needed highly scalable web servers. (7:11) When people were accessing information over the internet, remember all that information had to be stored someplace. (7:20) It had to handle the increasing number of people accessing that information.

(7:27) The way that happened is to use these parallel computers that were initially developed for scientific computing for the internet. (7:40) That’s one major lesson that we got. (7:47) Costs are very, very important.

(7:52) Of course, we’re facing that issue with cost of computing with AI now.

Melissa

(7:59) So true. (8:00) Talking a little bit about technology versus transformations, I find as we’re in this era of AI right now, AI everything, companies are focusing on the new technology, but they’re not focusing on the transformation that goes with technology. (8:23) The question is, why do organizations adopt these new technologies, but they don’t truly transform?

Irving

(8:32) I would say, and let me focus on AI rather than technology in general, there are two main aspects to AI to be very simplistic. (8:46) One is personal productivity. (8:50) How do you as an individual use it to do things that if you had to do it on your own would take you a lot longer?

(9:02) So personal productivity, and let me give you a very personal example. (9:08) I write a weekly blog. (9:11) I’ve been writing a weekly blog for the last 20 years since 2005.

(9:19) For a while, it was reposted in the Wall Street Journal, CIO Journal. (9:29) One of the things I liked about that is the professional journalists at the Wall Street Journal would take my blog and they would edit it in a way that make it better because they’re editors, they know how to do this well. (9:47) But at some point, they start reposting my blogs and I missed having an editor.

(9:55) So not only do I have to go through the pain of writing the damn blog, writing is painful. (10:02) I don’t know if people realize that. (10:05) You have to figure out what do you want to say, how to say it, etc.

(10:10) But then you have to go through revision after revision after revision. (10:14) So a few months ago, I said, you know, maybe I can ask ChatGPT to help me edit my blogs, not to write them. (10:26) I have to do that myself.

(10:29) But once I write them, can you help me edit it, like improve the flow and make sure it’s understandable and so on. (10:42) So I went to ChatGPT and I always talk to the chatbot in a very nice way, as opposed to saying, Goddamn you chatbot, help do this. (10:56) I like to say, and we can discuss why I do that, would you be able to help me edit my blog?

(11:05) I would really appreciate if you can help me. (11:08) But the way I’d like you to do that is to keep my voice. (11:13) That is, don’t change it, do it for clarity and so on.

(11:17) And the chatbot would reply, of course I can help you edit your blog and I’ll follow your instructions. (11:25) And just paste the blog here. (11:30) And I did.

(11:32) And I got it edited in five seconds. (11:37) I mean, it’s impressive how good they can do things. (11:41) And I have to say, in the same way that my blogs were better after the Wall Street Journal professionals edited it, I was amazed what a good job the AI did.

(11:59) That, I would say, it’s personal productivity. (12:02) It’s me as an individual asking a chatbot to help me do something and so on. (12:11) And that people are using it now for programming.

(12:16) People are using it for, you know, sort of give me a way of celebrating something in the style of Shakespeare. (12:30) And the AI is very happy to do that and so on. (12:33) That one is advancing nicely.

(12:36) But the most important one is the way businesses, governments, universities will use it for business value, for marketing value. (12:53) And that part has gone much, much, much slower. (12:59) And that’s why you read that so many companies, you know, they have, they’re using AI, but they’re not making much money from it, because you only make money if it helps improve your business.

(13:19) That’s very different from the personal productivity. (13:23) And that’s a very different thing. (13:27) And that’s the part that I believe in taking a technology to market takes much, much longer.

(13:37) Can I give you a couple of examples?

Melissa

(13:40) Yes, please.

Irving

(13:42) So computers, I would say, became commercially available at scale in the 1960s. (13:52) That’s when IBM came out with the 360 mainframes. (13:57) And they’re used in transaction processing in banks, in inventory management, and so on.

(14:05) But there is something called the Solow Paradox. (14:09) Have you heard of the Solow Paradox? (14:12) S-O-L-O-W.

(14:15) And its name after a famous Nobel Prize winning economist at MIT, who in the 1980s, early 90s, said computers are now everywhere, except in the productivity measures. (14:37) Because despite having 40 years of computers everywhere, the productivity of the economy and of businesses hadn’t changed. (14:49) So he said, what the hell is going on?

(14:53) In the 90s, what people discovered is that companies were using computers, but they were applying them in their companies in each isolated part of the business. (15:10) In other words, they were applying the computers to individual processes in the business, but not to the business as a whole. (15:23) And if you really wanted to get a transformation with the computers, you had to transform the whole way you manage the business.

(15:38) That’s when ERP came along, Enterprise Resource Processing. (15:46) You’ve heard of ERP, correct? (15:50) Which SAP is one of the top companies.

(15:53) It came along. (15:55) It was very painful for companies to do that because all of a sudden, they had to create a shared database for all the different parts of their companies. (16:09) And they had to make sure that, let’s say, the manufacturing operations were putting their data in this shared database.

(16:20) So the finance people could analyze that data and see how they were doing. (16:27) And it took often years to embrace ERP. (16:35) Once they did that, the productivity of computers went way up because they really transformed the operations of the business.

(16:48) So that’s one example. (16:50) A similar thing happened with the internet, that when the internet came along, and I was very involved in that in IBM because I was the general manager of our internet division. (17:06) And the way we brought the technology to clients is to work with them and say, okay, let’s identify an application that could improve if you use the internet.

(17:24) And this is very important. (17:27) The simpler the application to begin with, the better it is. (17:33) So here is one example.

(17:35) If you were FedEx or UPS and customers use you to ship their packages or people to receive them. (17:50) We’re talking now about the early 1990s before the internet spread out. (17:57) If you wanted to know where your package was, you called an operator on an 800 number and you said, hi, I’m Irving.

(18:11) Here is the track ID of my package. (18:15) Could you tell me when it’s going to get here? (18:18) And the operator then entered that in their computers, got an answer and told you that.

(18:28) So that worked. (18:30) But then we discovered something that you may say, God, Irving, that’s so trivial. (18:35) That was good that it was trivial.

(18:37) All of a sudden, we developed an application, and I don’t just mean everybody, where you could now link your personal computer to the identical mainframe database that had the data where your package was. (18:58) And instead of having to call an operator, you went online, you enter the package ID in the PC, the PC went to the mainframe, looked up the data, came back, and showed you where your package was. (19:19) And that took no time.

(19:20) That really revolutionized package tracking. (19:27) Now, you say, well, what’s the big deal of that? (19:31) The big deal is that people felt much better about it.

(19:35) You could wake up at 3 a.m. and say, where is my damn package? (19:42) And just from your PC, you can access it. (19:45) So customer satisfaction went way up.

(19:52) The same thing happened with, if you wanted to check, you were buying a car, you wanted to see multiple cars, you call the car company or the dealer and you say, send me a manual about your car. (20:10) And all of a sudden came edmunds.com. (20:14) I don’t know if you’re first of edmunds.com was one of the first websites that had lots of information about every kind of car you can imagine. (20:25) They’re still around. (20:27) And so you could go to edmunds.com, you could tell this is what I’m looking for. (20:33) Can you help me compare this and this and that?

(20:36) And all of a sudden, you could decide what kind of car you want without having to get lots of different manuals. (20:46) Electronic commerce was the same thing. (20:50) Our first e-commerce customer in IBM was LL Bean.

(20:58) Now, why LL Bean? (21:01) Because LL Bean, have you heard of LL Bean?

Melissa

(21:04) The clothing store. (21:05) Yeah.

Irving

(21:06) Yeah. (21:07) So they had a catalog business before the internet where they would mail you a catalog. (21:16) And then you had the catalog.

(21:20) You then, when you decided what you wanted to buy, you called an 800 number. (21:27) You spoke to an operator. (21:29) You told him or her what you wanted.

(21:33) The operator did that. (21:36) And then they told you, we’ll ship you this stuff. (21:40) Guess what we did?

(21:41) We put the catalog online so that now you could see the catalog and get what you want. (21:53) It went automatically into the LL Bean website. (21:59) Everything got done.

(22:01) So productivity and customer satisfaction went way up. (22:05) I’m sorry to be- No, no, no.

Melissa

(22:07) It’s good points. (22:08) I think when we think of the digital economy today, and from your work at MIT, what are leaders and companies missing about the digital economy? (22:21) And I think you really set that up with talking about the magazines.

(22:25) I still get magazines from clothing stores and I’m not mailing what I want in back to the clothing stores. (22:33) I’m either A, going to the clothing stores, which now I know have less choices at the store, or I’m going online because I know what my sizes are. (22:44) And so talk to me a little bit about what the world is missing.

(22:50) Money Ripples is on a mission to help professionals just like you get their money working harder for you. (22:57) Their clients free up an average of $35,000 their very first year without having to work extra hours to show you how they’ve put together a powerful training called Cashflow Secrets, and the listeners of the Executive Connect podcast can get it completely free. (23:14) Just visit moneyripples.com forward slash secrets and enter the promo code EXEC.

Irving

(23:22) What the world is missing with AI, first of all, because people are so excited about all the things AI can do, which are incredible, and we can talk about some of them. (23:37) Instead of looking for the simplest business applications to implement, they look for the most complicated ones to show the power of AI. (23:50) Now remember, when I went through a few examples of the internet in the 90s, picking simple applications was critical, because that’s how the companies learn how to use the technology.

(24:06) We haven’t done that. (24:08) Now we’re talking about agentic AI, which is going to be a fabulous technology where you can get things done with some people and lots of AI agents that are going to help them autonomously. (24:27) That is a very complicated application, because all of a sudden you have to have all these AI agents interacting with each other and stuff like that.

(24:42) That feels to me like give me a break. (24:47) Maybe in 5 to 10 years we’ll be able to do it well, but you’re picking something that is really difficult to do. (25:00) Then you wonder why it’s not bringing you business value.

(25:08) It’s because you haven’t implemented a very good application, (25:13) or even worse, when companies, and this is especially happening in Silicon Valley, (25:21) and there is a reason for this, when a lot of the discussion is when are we going to have (25:30) artificial general intelligence, meaning AI can do everything a human can do, (25:39) or even better artificially, the next step is now AI can do ASI, AI can do everything humans can do (25:53) much better.

(25:55) They are talking, are we going to have it in two years, or three years, or five years? (26:03) Give me a break. (26:06) First of all, why do we care that AI can do something better than a human or not?

(26:17) What we really want is for AI to be helpful to me, like I used it for editing, or use it in a business application that has real value. (26:30) If you said to me, Erwin, but wouldn’t you like AI to be better than you at everything it does? (26:41) I don’t give a hoot about AI.

(26:44) I want business value. (26:47) I want something that really works well. (26:51) Now, the reason I think in Silicon Valley, people have written quite a bit, they spend so much time in AGI and ASI, is I think they’re chasing valuations.

(27:03) They’re spending so much money, and they’re not getting revenues for it. (27:09) They’re trying to say, Melissa, just wait three years. (27:12) This kid is going to be like Lionel Messi.

(27:16) The kid is only 11. (27:19) In three years, he’ll be 14. (27:22) Trust me, Lionel Messi is toast.

Melissa

(27:26) To your point, I think how we all use AI and technology differs from person to person based on our age, our knowledge, our experience, our comfort with it. (27:43) I love your example about writing.

Irving

(27:47) Now, I am- Or how a business uses it.

Melissa

(27:50) Yeah, of course.

Irving

(27:51) That’s even more important, how a business or government uses it.

Melissa

(27:55) Go ahead. (27:56) 100%. (27:56) But I think it really comes down to finding ways to use it to augment, to get more time.

(28:05) I use it like you do for writing, because I’m terrible with commas and apostrophes. (28:10) The spacing, sometimes I mix up. (28:12) But I think about how many times my perfectionist tendencies went back and re-read and read and changed and changed again.

(28:19) If I could just do it once or twice and have a tool save me an hour, it’s an hour back in my life I could use in other areas.

Irving

(28:28) Yeah. (28:29) So I want to try- And it improves. (28:31) The writing is yours.

(28:33) It improves the overall- 100%. (28:37) The overall readability of the article you wrote. (28:41) Exactly.

Melissa

(28:42) Easy to read, quick to read. (28:43) We know that we’re all impatient. (28:46) We want our stuff delivered today and everything done tomorrow.

(28:50) And so I want to talk a little bit about open systems in the future.

Irving

(28:54) Yes.

Melissa

(28:54) So open technologies like Linux, reshaped computing. (28:59) So what do you see as far as openness mattering next in this new world we’re in and moving into?

Irving

(29:08) Well, the reality, and there has been some good work. (29:15) In fact, I posted a blog on this recently by Frank Mace, who is the top economist in the Linux Foundation, but he’s at MIT now in the initiative on the digital economy, which I am a fellow there also. (29:36) And he analyzed that open source AI models are 20% of the cost of proprietary models, like the ones open AI has been building, or Google has been building, or Anthropic has been building.

(30:00) Those proprietary models, as I’m sure you’ve been reading, they need gigantic amounts of data to train the model. (30:12) They need huge amounts of electricity, and so on, and on, and on, and on. (30:18) And they need water to cool the stuff.

(30:22) And people say, yes, but to make them better, we have to have even more. (30:30) And I suspect that the reason is you want one model that can do everything, like we’ve accomplished with large language models. (30:41) With open source AI, yes, you can use large language models and so on, but you may have more focused models to help you solve a problem you’re trying to solve, as opposed to a model that’s as good as the human brain.

(31:04) And well, but why do you need it to be as good as the human brain? (31:08) This is a goddamn machine. (31:10) So use it to help you do something that you want to do better.

(31:16) Don’t worry about the human brain. (31:20) First of all, the human brain has evolved over millions of years, and here is this poor AI, and you’re trying to make it evolve over five or 10 years. (31:32) Give it a break.

(31:33) Give it time to grow up. (31:36) And so I think that we’re so fascinated by what AI can do, we forget that, you know, in the end, it is probably the most transformative technology of the 21st century, and it’s going to do incredible things. (32:04) And my point of comparison is the industrial revolution, which started with steam power in the 1780s, whatever.

(32:18) And was it transformative throughout the 19th and 20th century? (32:25) Of course. (32:27) You know, in the mid-19th century, in the 1840s, 50s, all of a sudden you had railroads, which wouldn’t have happened without the steam power.

(32:38) Remember, it took about 40 years from the development of steam engines to railroads. (32:44) And then another 40 years later, you had electricity in the, you know, with Edison and so on. (32:54) That came along in the 1880s.

(32:58) And you know what happened in the early 20th century, Melissa? (33:03) We had cars. (33:05) And a few years later, we had airplanes.

(33:10) I mean, you would say to me, Irving, but that’s magical. (33:16) Nobody, when the industrial revolution started in, I don’t know, 1780, would have predicted that in 1920, we would have airplanes. (33:29) Humans could fly.

(33:31) Well, that’s what happens with technology. (33:36) So what I go through in my mind is, don’t tell me about AGI and ASI. (33:44) I’m much more interested in what is the economy going to be like in 2040, in 2050?

(33:56) What is it going to be like in another 30 years later when your children have grandchildren? (34:07) That’s going to be a fascinating question. (34:10) I don’t know the answers, Melissa, anymore that I could have predicted railroads and airplanes when the industrial revolution started.

(34:23) Now, you could say, God, Irving, you’re being so conservative. (34:28) You are saying, give AI a chance to grow up. (34:35) But that’s what I’ve learned from history, that these technologies take time.

(34:43) And what takes time is not the technology itself. (34:48) It’s the deployment across the economy and society. (34:53) And it will only get deployed when you find applications that turn out to be incredibly useful to individuals, to businesses, and to governments.

Melissa

(35:08) Yeah. (35:08) And I’m so glad you brought that up because I want to piggyback that a bit. (35:12) And you advise a lot of governments as well as enterprises, like we discussed.

(35:17) But where do businesses and policy most need to align today as it relates to what you were just talking about?

Irving

(35:27) I think they need to work closer to individual companies to together figure out and advise the companies what will really bring you good value. (35:41) And I don’t think they’re doing that. (35:44) I don’t mean all of them.

(35:47) I think they’re spending too much time just on the technology and not enough on its deployment across the economy. (35:57) With one exception, the part where AI, in my opinion, will have the biggest near-term success is in the research world, such as finding new healthcare methods for curing cancer or trying to understand what causes senility, what causes our brains to age, what causes autism. (36:36) And the reason I’m much more comfortable with that is the use of AI in those kinds of research applications are building on the huge history of using supercomputers to do research.

(36:56) Remember, scientists have been using supercomputers to do all kinds of research on different kinds of problems from how to find extragalactic planets by analyzing astronomical data to finding all kinds of new methods of doing research. (37:28) And so using AI for finding cures for cancer or senility goes back to, well, you’re using it as a supercomputer, Irving. (37:43) You’re cheating.

(37:44) No, that’s not cheating. (37:46) A supercomputer is really good at analyzing lots of information. (37:52) That’s what it’s good at.

(37:54) Here we are. (37:56) We have more information at our disposal than ever.

Melissa

(38:02) Right. (38:03) So true. (38:04) Yeah.

Irving

(38:04) More than ever. (38:06) It would be impossible for people to do it any other way. (38:12) You know, Melissa, AI is really good at analyzing information.

(38:21) That’s its claim to fame. (38:24) So you’re not using it in a more supercomputing application. (38:30) And the people doing that are people who have been very good at using supercomputers for research for the last three, four decades.

(38:42) So I’m very comfortable with that. (38:45) I am less comfortable. (38:47) For example, you saw how many companies, Salesforce said that AI all of a sudden is going, a lot of users of computers will be dead.

(39:00) Well, that hasn’t happened. (39:02) They, a number of companies have said that many online applications are going to die because of AI. (39:10) And the answer is that hasn’t happened because they really haven’t found the economic value to the business.

(39:20) And that’s why so many promises haven’t come through. (39:24) Melissa, just to give you an example, I think I first read this that in 2015, Elon Musk was a brilliant man said that, I don’t know, in five years, we would have self-driving cars all over Manhattan. (39:45) He said that, I don’t know, in 2015.

(39:49) Melissa, we don’t have self-driving cars all over Manhattan yet. (39:54) So, you know, a lot of those predictions are, I don’t know, macho kind of predictions of saying something that you don’t know how to do and willing it into existence. (40:12) Self-driving cars will come along.

(40:15) They will be much, much better. (40:18) They’ve gotten much better. (40:20) Do you have them in Austin also?

Melissa

(40:22) Yes, we have them in Austin and we’ve seen them, you know, in other big cities. (40:27) And so I love that you brought that up because kind of to tie it up here and kind of put a bow at the end here. (40:34) So when we look about the future and we look at leadership, we’re going through a, you know, a time where there’s a lot of change, there’s a lot of fear, there’s technology, whether you like AI or you don’t like AI, whether you are moving forward with technology and transformation or not.

(40:54) So what kind of leadership does this moment require for people in the world today, personally and professionally? (41:03) This episode is brought to you by Summit Ventures. (41:09) If you’re an accredited investor, Summit gives you access to one of the greatest tax advantage opportunities, direct ownership in oil and gas.

(41:20) Their projects deliver what they call the triple play, cashflow, equity growth, and powerful tax benefits. (41:28) And here’s the best part. (41:30) These investments qualify for 1031 exchanges.

(41:34) That means you can roll gains from real estate into energy while deferring capital gains. (41:41) To learn more and get a free white paper, oil and gas demystified, just visit www.summitven.com forward slash executive connect.

Irving

(41:55) I think the number one leadership, I, in my opinion, based on my experience with computers first, and then with the internet, is business and market leadership. (42:13) Spend a lot of time trying to figure out how to make companies more profitable, how to make them more productive. (42:24) And, you know, don’t try to wow me with something that one person did that was great, but didn’t improve productivity, didn’t create a new product.

(42:38) You know, I want to see hard business value being created. (42:47) And that’s the part that I think has been going so much slower because what people tell me is AI is great. (42:58) I don’t want to hear how great AI is.

(43:00) I agree AI is great. (43:02) I want to see what new products it’s created, what new business value it has brought, what new overall productivity. (43:13) And I don’t think we’ve done that enough because a lot of people may say, Irving, that’s boring.

(43:20) It’s much better to write articles that are 90% hype about what we expect. (43:31) And, you know, business is tough. (43:35) I mean, it’s a lot of blood, sweat and tears.

(43:39) And, you know, companies sometimes lose their head and they go under and so on. (43:46) So I think I would like a lot more attention to the market forces that would help us bring AI. (43:57) I mean, let me just give you a tiny example with self-driving cars that, you know, they’re really all over the place in San Francisco, which is great.

(44:09) But did you know that Waymo, you know, Waymo is owned by Google, whatever. (44:18) But for self-driving cars in Waymo, they’re self-driving. (44:24) But did you know that in every city where they are, there is a group of people monitoring the cars?

(44:34) And if the car gets into trouble, they get involved and they start changing things. (44:40) Now, you said to me, Irving, is that cheating? (44:44) No, that’s not cheating.

(44:46) But what that says is that the overall cost of having self-driving cars is much higher than you can imagine, because not only do you need the car with all the technology, but you need now a group of experts that are hanging around to make sure that the car works. (45:11) So if I try to convince you, Melissa, to buy a self-driving car for your own personal thing, I would say, OK, the car will cost you, I don’t know, $100,000. (45:24) And if you say, Irving, that’s too high.

(45:26) I say, well, radars and all the technology. (45:31) But you have to pay me, I don’t know, $300 a month for the access to the experts that you need when you get into trouble with the car. (45:45) And you say, hey, wait a second, I never signed up for this monthly fee.

(45:50) Well, then you won’t get the self- We are sometimes ignoring the economic realities of what it takes to deploy a technology at scale. (46:06) And the reason we’re doing that is because the technology is in its early stages. (46:11) We still need a lot more experience.

(46:14) Am I being too pessimistic, you think?

Melissa

(46:17) No, I think anything, you know, everything good takes time, right? (46:23) Even us as humans, we’re learning, experiencing, making mistakes. (46:29) We take time.

(46:30) We’re never going to be a finished product. (46:33) And I don’t believe that technology is ever going to be, as soon as we have figured out a technology, something else comes in or something else happens. (46:42) So I think the more that we, you know, lean into that and lean into nothing’s going to be black and white, and we’re going to forever be living in this gray time.

(46:55) And with each year and each technology advancement, we’re going to have to learn a new thing.

Irving

(47:02) And Melissa, let me finish by saying something that may be obvious, but you have four children. (47:15) And I think we can agree that watching babies, how quickly they learn is magical. (47:27) Can we agree on that?

(47:30) And the reason is that our brains have evolved over millions of years. (47:37) So your babies weren’t born with nothing in their brain. (47:44) Evolution made sure that in order for humans to survive, their brains had all this ability to learn and to smile.

(47:56) And I mean, you have to agree, it’s amazing how quickly they pick up language and so on. (48:05) But it’s because of all the evolution that came before. (48:11) AI is just a machine.

(48:13) It hasn’t been born with all that evolution. (48:17) And if you said, but over time, it will get better. (48:21) Yes, it will get better over time, but not in three years or five years.

(48:27) So I am making an analogy with the fact that to be really good at something takes a lot of time. (48:36) Sometimes it is evolution. (48:39) Sometimes it is like Lionel Messi.

(48:44) Do you know who he is? (48:47) Yeah.

Melissa

(48:48) And I want to kind of put a bow on this. (48:50) I agree with you. (48:53) I think the more that we give children, businesses, technology, time, the more we’re going to evolve.

(49:01) And as technology comes, we need to evolve with it and choose how we’re going to use it in our life and in our business and educate ourselves in the process. (49:13) I want to thank you so much for being here today. (49:15) I love the conversation.

(49:17) I loved the insights that you left with our listeners. (49:22) Please connect with Irving on LinkedIn. (49:25) And thank you for joining us today on the Executive Connect podcast.