AI's Impact on Productivity and Innovation

Knowledge at Wharton

Ask about Knowledge at Wharton

Episode Description

EPISODE OVERVIEW Wharton professor Daniel Rock discusses the impact of artificial intelligence on productivity, economic growth, and workforce transformation. The conversation covers key challenges such as mismeasurement, restructuring lags, and the role of AI in business innovation. Insights include how companies invest in AI, its influence on labor markets, and the long-term effects on economic development.

PODCAST OVERVIEW Every day, business scholars answer pressing questions in their research - but what do their insights mean for you? In this podcast, Wharton faculty dive into what inspired their studies and how their findings resonate with the world today. Learn how research insights translate into knowledge you can use, with host Dan Loney.

Learn more at http://whr.tn/rippleeffect.

#AIProductivity #ArtificialIntelligence #FutureOfWork #BusinessInnovation #AI #artificialintelligence ----- The Wharton School of the University of Pennsylvania is committed to sharing its intellectual capital through Knowledge at Wharton, the school’s online business analysis journal.

Launched in 1999, Knowledge at Wharton offers free access to:

- Articles, podcasts, and videos highlighting Wharton faculty research and analysis of current business trends - Interviews with book authors from Wharton School Press - In-depth series of curated content like Wharton Executive Education’s Nano Tools for Leaders - A searchable database of more than 10,000 articles covering all aspects of business - A weekly newsletter that delivers Knowledge at Wharton insights directly to your inbox

Visit the Knowledge at Wharton homepage: https://whr.tn/3qUpSBE Subscribe to the Knowledge at Wharton YouTube channel: https://whr.tn/3DEAgAU

Read the full transcriptHide the full transcript
  1. Daniel Rock: That's a really fascinating point. I mean, you think about the value of an OpenAI or Anthropic, even a Microsoft,
  2. Google— you know, whoever's building it, Llama— like, a lot of the value of those companies is in the complementary
  3. investments that their customers are making. Or, like, the ecosystem at large is making. That's really interesting,
  4. right? Like the— they get more valuable as their customers and as their consumers learn how to integrate that toolkit. So I
  5. think that's something that'll take a little while, but they're well aware of it too.
  6. You know, they're trying to make it easier to use. - Welcome to <i>The Ripple Effect</i>,
  7. the podcast that takes you on a journey through the minds of Wharton faculty. I'm your host, Dan
  8. Loney, and in each episode, we'll be diving deep into the inspiration behind the groundbreaking research that
  9. Wharton professors have conducted and exploring how their findings resonate with the world today.
  10. Dan Loney: Well, when we think about innovation these days, there's a good chance that
  11. artificial intelligence is going to come into the conversation. Even though it's been around for some time, it only now feels
  12. like the majority of the public at large are seeing the impact of artificial intelligence in our lives. Daniel Rock is an
  13. Assistant Professor of Operations, Information and Decisions here at the Wharton School. He and colleagues have
  14. looked at how AI can impact something like productivity. Dan, great to have you here today. Thanks for your time.
  15. Great to be here. Thanks for having me. So when I bring up AI, doesn't it seem like productivity kind of is a— a
  16. natural— a natural first thing for people to think about? Oh, absolutely. And I think when we talk about productivity, it's important to
  17. define terms here. For economists, productivity can be a few different things. It can be how much output per worker
  18. you have, how much revenue per unit of input, but generally, all of it points to one big idea, which is, what are the
  19. number of outputs we get per unit of input? So it's not, you know, how do we cut jobs, necessarily, or reduce the resource use. It's
  20. also, how do we create more? And I think with these tools empowering people to do, you know, greater and more
  21. interesting things, productivity in the long run has to kind of be positively impacted by what we can do with them.
  22. - Would you say that, though— that there's a paradox when you think of this? Yeah, sure. I think here's— here's the core thing with a
  23. sufficiently transformative technology, what economists would call a general purpose technology. That is, it's
  24. pervasive, it improves over time, and then it kind of necessitates and spawns complementary innovation. That
  25. is the other stuff you need to build to get this stuff to go. So, yeah, there's a lag. It takes a long time to build up
  26. those additional assets, to reconfigure your organization, to train people to use stuff. Over time, that's going to pay
  27. off in a big way, and we're seeing people make huge investments in that. But it's not going to be, right off the
  28. bat, super powerful. Actually, it's funny with AI, there are some applications that are right off the bat super powerful, but
  29. the long run implications are going to take a while to play out, I think. - You have four areas
  30. of potential impact that you've come up with in the work that you've done, the first being false hopes. Explain that a
  31. little bit. Oh yeah. So this is the explanation for the paradox. So, why it takes a while. So I already preempted what I think is going
  32. on. But yes, there is the— there is the chance, right— this is sort of a Bob Gordon view. I don't want to put too many words
  33. in his mouth, but, you know, AI just isn't that big a deal. Or you could broaden this to say any technology just isn't that big a
  34. deal. And we see lots of promise and hype, but it's just never going to materialize. That's a consistent way to view the world
  35. in the early stages, if you don't know what's going to happen. But then you do have to, you know, change tack if you see
  36. the benefits start to show up. I think with AI, we're starting to see that a bit. So that's number one.
  37. The second one is mismeasurement. So this is kind of— there's some folks in Silicon
  38. Valley who say this, and there's some evidence that this might be going on too. The idea here is that the gains are real. They're
  39. happening, but we're not capturing them properly in the economic statistics. And I think this— the folks at the BLS and
  40. the BEA do a really great job of trying to measure the economy. You know, where it might be tougher to measure things is
  41. like, you know, Google's free. I asked my MBAs, would you rather have, you know, search or indoor plumbing? And you know, after
  42. trying to wriggle out of that conundrum, many of them will still pick search over indoor plumbing. I'm kind of with them
  43. on that one. - That's probably a good idea. Yeah. It gets cold in Philly, but not too cold, - Exactly.
  44. So that's the second one. We could be mismeasuring things. And yes, there are some
  45. cases where that may be the case. But in general, you have to make an argument for why it's different now. What changed to
  46. make us worse at measurement, given what the economy is producing? And I think that's a tougher case. My co-author, Chad
  47. Syverson at University of Chicago, has kind of disposed that argument. At least, you know, up until 2017 or so.
  48. So that's the second one. The third one is sort of a rent dissipation argument. What does that mean? It's, the gains are
  49. real, but they're accruing to a really small proportion of people in the economy. They're taking all of the gains, and
  50. nobody else is seeing anything there. I think you could make an argument that a lot of that is still happening, but it would
  51. have to be really enormous to take away the gains from the technology, given the expectations. And then the last
  52. one, which we just discussed, restructuring and implementation lags. The stuff can take a while. We see a lot of promise,
  53. but let's not confuse a clear view for a short walk. It's going to take a long time to implement this stuff.
  54. So is the expectation, then, with where we are kind of currently, that we're still going to see innovation coming from other
  55. areas to kind of complement what is, I think a lot of people believe, the core of— of what AI is right now?
  56. Yeah, absolutely. I think that's that's already happening in a lot of areas. One of the really cool things about AI is— I don't
  57. know if you you code, or if you if you use coding assistance, but the fact that these tools— the software you need to augment
  58. them, to augment the AI tools, can be partially written with AI help. So we can— we really get in this nice, like, flywheel,
  59. where you use AI to improve the sorts of tooling you need to make AI more effective.
  60. That's— when I talk with companies, I'm like, "Are you guys doing this? Because you should be."
  61. It's very helpful. There's an element that you talk about in this paper, about the added
  62. costs being a type of capital that can be just part of the build out, I guess, correct?
  63. Yeah, absolutely. When you have these adjustment costs or fixed costs of investment, these are things like the training or
  64. intangible capital, or even the culture around how you build an organization that— that works with machine learning or AI
  65. software. Like, this is a different type of software. It creates output that's non- deterministic, or maybe a little
  66. bit fuzzier. It's not perfect every time, it's not cookie cutter. That's a mindset shift, too. You can't expect the same
  67. results as you could with— with ordinary kind of rules-based software. So you have to pay that upfront cost, or maybe even
  68. ongoing cost, to keep people in the loop, to structure your organization processes properly. And then what you— what you get
  69. out of that is, you know, competitive advantage, basically. You can do things other companies can't, if you
  70. crack that. - Sure. - Yeah. And obviously, part of that also probably fills into,
  71. with the companies that are in the development of AI, the value that those companies have— you know, that that is a significant
  72. beneficial capital component that they have to their companies, which, you know,
  73. obviously is larger than other companies. Sure. That's a really fascinating point. I mean, you think about
  74. the value of an OpenAI or Anthropic, even a Microsoft, Google— you know, whoever's building it, Llama— like, a lot
  75. of the value of those companies is in the complementary investments that their customers are making. Or, like, the
  76. ecosystem at large is making. That's really interesting, right? Like the— they get more valuable as their customers and
  77. as their consumers learn how to integrate that toolkit. So I think that's— that's something that will take a little while,
  78. but they're well aware of it too. You know, they're trying to make it easier to use. In some sense, ChatGPT is more of a UX
  79. innovation. The playground existed. You could use stuff before, but ChatGPT just, like showed people, "Hey, you can—you
  80. can really engage with these models and do something cool." I massively updated how important I thought UX was after I saw the
  81. success of that app. So where we are right now then, you mentioned about some of the
  82. metrics that will come into play here. It feels like we're still at a point where the development of some of those metrics is kind
  83. of either— it either hasn't happened or it's ongoing right now, and so it may be hard to truly gauge the value or the
  84. component of productivity, especially when we don't have the dynamics fully tweaked to what we need, right?
  85. Yes, I agree with that. Though what makes AI, this wave of software, a little bit easier to do a good job there is, we're
  86. using the last wave of IT to instrument it, right? So we have software to track the software. Before, I mean, the best you
  87. could do, best you could hope for, was surveys of some kind, like, say, in the early '90s. Now we can scale up those efforts.
  88. You can track sort of what some of my colleagues called "digital exhaust", right? You connect to APIs for companies, you can see
  89. how they're changing what they're doing. You know, there's little pockets. This is like this big iceberg, and we're
  90. seeing just the tip of it. But you can use those sort of points where we can see the tip of the iceberg and how it's changing as
  91. a way to gauge what's going on. And a concrete example is something— you know, something I've done in my own work. I've
  92. tracked how many people with AI skills are being hired company by company, assuming that if you're hiring people to do this,
  93. you're probably building out all the other complements to make them effective. And, you know, that's— we're trying to measure
  94. the— kind of the size of the— the investments on that front, which Which for many companies, is probably the way you need to do it, is
  95. bring on the talent before you actually do the level of implementation to get to that point, so you have people who understand it
  96. going— going— going into the process. 100%. You can't get away from— from labor markets and those
  97. complementary investments. If you do AI well, then you probably did data science well before. If you did data science
  98. well, you probably did the cloud well before. There's a whole stacking of these technologies. It actually makes AI super
  99. concentrated in— in only a few firms right now. And it's a little bit, you know— a little bit of an explanation where that
  100. might be coming from. But yeah, you're bottlenecked in three potential areas. It's either talent or data or compute right
  101. now. But you mix those three things together in the right proportions, and you start to get, you know,
  102. AI internal capabilities. What do you think, then, doing this research helped you to better
  103. understand about where we are going, and that connection between AI and productivity? - Yeah.
  104. So I started to think, what are the ingredients inside of a company that would generate productivity? Like, how— where
  105. does it come from? And there's a few models that colleagues in other places have put together that I can use as sort of
  106. workhorse models. There's sort of the task-based approach. There are researchers like David Auter and Daron Acemoglu at MIT
  107. who have used this heavily. The idea is like, let's break down a job into bundle of tasks and track all those tasks. Those
  108. tasks or those bundles are changing. I think the more change you see at that level, the more of an indication you
  109. have that something is different now. So that's one early kind of check you can do to see what's going on. So I'm working on some
  110. of that, looking at job postings with a few colleagues. And then there's kind of a perspective that Tim Bresnahan at Stanford
  111. as well as Joshua Gans, Avi Goldfarb and Ajay Agrawal at University of Toronto have kind of put forth. It's the sense that
  112. nobody's ever lost their job to task-based automation. It's not happening task by task, necessarily, or change isn't
  113. happening task by task. It's happening at the system level. So when we change the direction in what's possible with a model,
  114. like, I can discover new drugs with these tools, or I can— I mean, I can't. But I can make pretty lousy images or paintings
  115. that I couldn't do before. I'll stop there with my new capabilities. But like, as— as you give people these new
  116. capabilities, you redesign the system, and that new system's got different demands for people. Different demands for capital.
  117. - Right. - So let's see if, like, there are companies making big changes saying "We're going to reconfigure this module." It's
  118. kind of hard to do that if things are moving really quickly. You don't have certainty. You don't feel like
  119. you're standing on solid ground when you do that. Does it feel, like, then— I'll finish up on this. Does it feel like
  120. that where we are with AI and the workforce right now, that obviously AI is going to play a significant role, but the human
  121. component will be there. And to a degree, maybe even the human component becomes kind of even more of a learning experience as
  122. we move forward, because of how AI is kind of guiding the ship a little bit here.
  123. Yeah, I get accused of being an optimist on this point, so I strongly agree with that. But I do think there are, of course,
  124. going to be pockets where things go better or things go worse. One thing I've grown fond of saying recently is that we can't
  125. get away from labor markets. So the— I will say, you know, perhaps it's a little bit of a stretch. I don't want to get too
  126. far out of my skis here. But the augmentation versus automation debate matters at some unit of analysis. But at the individual,
  127. like, you know, worker or manager making a decision about where to deploy the technology— you can augment someone that can
  128. do the job of 20 people, and if the company doesn't want 20 people to do it, then that's not great news. For the workers, that
  129. is. On the other hand, you can automate things that people hate doing and refocus their work on to other stuff where demand
  130. expands. So these are choices that companies and managers and workers can make. They're not foregone conclusions. And I
  131. think I've— I have confidence in the— you know, the talent of people out there in the world to make good choices there, and,
  132. you know, ultimately end up in a more fulfilling sort of work configuration scenario.
  133. Dan, great to have you in here today. Thanks very much. - Thanks so much for having me. Great to be here.
  134. Thank you. Daniel Rock, who's Assistant Professor of Operations, Information and
  135. Decisions here at the Wharton School. - Thank you for listening to <i>The Ripple Effect</i>. We hope you found this episode
  136. informative and engaging. Don't forget to subscribe and leave us a review so that we can continue to bring you the best insight
  137. from the Wharton School.

Brief Summary

Daniel Rock discusses AI's impact on productivity and the challenges of measuring its effects in this episode of The Ripple Effect.

Key points

  • The Ripple Effect Podcast Introduction. Join host Dan Loney as he explores groundbreaking research from Wharton faculty.
  • AI's Impact on Productivity. Daniel Rock discusses how AI can enhance productivity and its long-term implications.
  • The Paradox of AI. Exploring the paradox of AI's potential versus its actual impact in the early stages.
  • Complementary Innovations. The value of AI companies lies in the complementary investments made by their customers.

Tension

45/ 100Some tension

Related episodes