IBM‑MIT Lab Builds AI Foundations
Key Points
- Malcolm Gladwell introduces “Smart Talks with IBM,” focusing on how AI acts as a game‑changing multiplier for businesses, with guest Dr. David Cox, IBM’s VP of AI models and director of the MIT‑IBM Watson AI Lab.
- Cox explains his dual role: leading the MIT‑IBM Watson AI Lab—an academic‑industry partnership that dates back to the 1950s origin of AI—and overseeing IBM’s development of large “foundation” generative models.
- The MIT‑IBM collaboration highlights a long‑standing history, tracing back to IBM engineer Nathaniel Rochester’s 1956 Dartmouth workshop that coined “artificial intelligence,” showcasing how the partnership has evolved to push the frontiers of AI research.
- Cox breaks down foundation models in accessible terms and describes how these massive generative AI systems enable new capabilities across industries, from product design to data analysis.
- He also discusses the practical impacts of AI on business operations, the future of work, and how enterprises can leverage these models to drive innovation and competitive advantage.
Sections
- AI Foundations and Business Impact - Malcolm Gladwell introduces Dr. David Cox, IBM Research VP and director of the MIT‑IBM Watson AI Lab, to discuss the evolution of AI, foundation models, and how artificial intelligence is reshaping business, work, and design.
- IBM‑MIT AI Strategy Partnership - David Cox explains how the IBM‑MIT collaboration places them at the forefront of AI research—especially foundation models—while pursuing broader societal impacts such as climate change mitigation and advanced materials development.
- From Data to Self‑Supervised AI - The speaker outlines how the surge in digitized data and faster computers paved the way for breakthroughs, culminating in self‑supervised learning that removes the need for labeled data and fuels the creation of increasingly powerful models such as ChatGPT.
- Foundation Models: The AI Oven Analogy - Jacob Goldstein likens foundation models to a versatile oven that replaces single‑purpose AI tools, emphasizing their broad utility and how they drastically cut the labor needed to build and deploy AI solutions.
- Beyond Dams: Foundation Model Power - The speaker likens traditional automation to building a dam only on large rivers—ignoring countless smaller data “puddles”—and explains that foundation models, trained on massive unlabeled corpora, serve as a universal base that lets users efficiently address any downstream task without starting from scratch.
- Specialized Small Models Outperform Giants - Jacob Goldstein discusses how a modest 2.7‑billion‑parameter model trained on biomedical literature surpasses far larger models in that domain, highlighting gains in efficiency, cost, and domain expertise while illustrating the industry's trade‑off between generic big models and task‑specific tools.
- AI Hallucinations and False Confidence - The speakers illustrate how language models can confidently fabricate information—such as a completely invented biography—showing that the persuasive tone of AI outputs can easily mislead users into accepting inaccurate content as true.
- Beyond Sentiment: AI Summarization Use Cases - David Cox explains how modern LLMs extend traditional AI tasks like sentiment analysis to more complex applications such as automatically condensing customer call transcripts and meeting notes for quicker insight.
- Conversational AI Transforming Code Generation - The speaker envisions AI dramatically reshaping software development by allowing users to describe requirements in natural language and have the system generate, reason about, and execute code autonomously.
- AI's Ubiquitous Business Revolution - The speaker argues that AI will permeate every product, cutting costs and creating new opportunities for firms while freeing workers from tedious tasks, drawing parallels to past technological shifts like mechanized agriculture and email.
- Watsonx Enables AI-Driven Creative Enterprise - The speaker outlines how watsonx unifies foundation models, tooling, and data management to let businesses safely harness AI’s power, while emphasizing the creative research role that fuels innovative enterprise solutions.
- AI as a Natural Resource - David Cox envisions AI becoming an abundant, resource‑like tool that seamlessly augments human work, creating both job displacement challenges and unprecedented opportunities for intellectual productivity.
Full Transcript
# IBM‑MIT Lab Builds AI Foundations **Source:** [https://www.youtube.com/watch?v=kLBdpNsXh1A](https://www.youtube.com/watch?v=kLBdpNsXh1A) **Duration:** 00:38:52 ## Summary - Malcolm Gladwell introduces “Smart Talks with IBM,” focusing on how AI acts as a game‑changing multiplier for businesses, with guest Dr. David Cox, IBM’s VP of AI models and director of the MIT‑IBM Watson AI Lab. - Cox explains his dual role: leading the MIT‑IBM Watson AI Lab—an academic‑industry partnership that dates back to the 1950s origin of AI—and overseeing IBM’s development of large “foundation” generative models. - The MIT‑IBM collaboration highlights a long‑standing history, tracing back to IBM engineer Nathaniel Rochester’s 1956 Dartmouth workshop that coined “artificial intelligence,” showcasing how the partnership has evolved to push the frontiers of AI research. - Cox breaks down foundation models in accessible terms and describes how these massive generative AI systems enable new capabilities across industries, from product design to data analysis. - He also discusses the practical impacts of AI on business operations, the future of work, and how enterprises can leverage these models to drive innovation and competitive advantage. ## Sections - [00:00:00](https://www.youtube.com/watch?v=kLBdpNsXh1A&t=0s) **AI Foundations and Business Impact** - Malcolm Gladwell introduces Dr. David Cox, IBM Research VP and director of the MIT‑IBM Watson AI Lab, to discuss the evolution of AI, foundation models, and how artificial intelligence is reshaping business, work, and design. - [00:03:50](https://www.youtube.com/watch?v=kLBdpNsXh1A&t=230s) **IBM‑MIT AI Strategy Partnership** - David Cox explains how the IBM‑MIT collaboration places them at the forefront of AI research—especially foundation models—while pursuing broader societal impacts such as climate change mitigation and advanced materials development. - [00:07:04](https://www.youtube.com/watch?v=kLBdpNsXh1A&t=424s) **From Data to Self‑Supervised AI** - The speaker outlines how the surge in digitized data and faster computers paved the way for breakthroughs, culminating in self‑supervised learning that removes the need for labeled data and fuels the creation of increasingly powerful models such as ChatGPT. - [00:10:06](https://www.youtube.com/watch?v=kLBdpNsXh1A&t=606s) **Foundation Models: The AI Oven Analogy** - Jacob Goldstein likens foundation models to a versatile oven that replaces single‑purpose AI tools, emphasizing their broad utility and how they drastically cut the labor needed to build and deploy AI solutions. - [00:13:13](https://www.youtube.com/watch?v=kLBdpNsXh1A&t=793s) **Beyond Dams: Foundation Model Power** - The speaker likens traditional automation to building a dam only on large rivers—ignoring countless smaller data “puddles”—and explains that foundation models, trained on massive unlabeled corpora, serve as a universal base that lets users efficiently address any downstream task without starting from scratch. - [00:16:32](https://www.youtube.com/watch?v=kLBdpNsXh1A&t=992s) **Specialized Small Models Outperform Giants** - Jacob Goldstein discusses how a modest 2.7‑billion‑parameter model trained on biomedical literature surpasses far larger models in that domain, highlighting gains in efficiency, cost, and domain expertise while illustrating the industry's trade‑off between generic big models and task‑specific tools. - [00:20:01](https://www.youtube.com/watch?v=kLBdpNsXh1A&t=1201s) **AI Hallucinations and False Confidence** - The speakers illustrate how language models can confidently fabricate information—such as a completely invented biography—showing that the persuasive tone of AI outputs can easily mislead users into accepting inaccurate content as true. - [00:23:03](https://www.youtube.com/watch?v=kLBdpNsXh1A&t=1383s) **Beyond Sentiment: AI Summarization Use Cases** - David Cox explains how modern LLMs extend traditional AI tasks like sentiment analysis to more complex applications such as automatically condensing customer call transcripts and meeting notes for quicker insight. - [00:26:09](https://www.youtube.com/watch?v=kLBdpNsXh1A&t=1569s) **Conversational AI Transforming Code Generation** - The speaker envisions AI dramatically reshaping software development by allowing users to describe requirements in natural language and have the system generate, reason about, and execute code autonomously. - [00:29:17](https://www.youtube.com/watch?v=kLBdpNsXh1A&t=1757s) **AI's Ubiquitous Business Revolution** - The speaker argues that AI will permeate every product, cutting costs and creating new opportunities for firms while freeing workers from tedious tasks, drawing parallels to past technological shifts like mechanized agriculture and email. - [00:32:24](https://www.youtube.com/watch?v=kLBdpNsXh1A&t=1944s) **Watsonx Enables AI-Driven Creative Enterprise** - The speaker outlines how watsonx unifies foundation models, tooling, and data management to let businesses safely harness AI’s power, while emphasizing the creative research role that fuels innovative enterprise solutions. - [00:35:27](https://www.youtube.com/watch?v=kLBdpNsXh1A&t=2127s) **AI as a Natural Resource** - David Cox envisions AI becoming an abundant, resource‑like tool that seamlessly augments human work, creating both job displacement challenges and unprecedented opportunities for intellectual productivity. ## Full Transcript
Malcom Gladwell: Hello, hello. Welcome to Smart Talks with IBM,
a podcast from Pushkin Industries, iHeartRadio and IBM.
I’m Malcolm Gladwell.
This season, we’re continuing our conversation with New Creators,
visionaries who are creatively applying technology in business to drive change
— but with a focus on the transformative power of artificial intelligence
and what it means to leverage AI as a game-changing multiplier for your business.
Our guest today is Dr. David Cox, VP of AI models at IBM Research,
and director of the MIT-IBM Watson AI Lab,
a first-of-its-kind collaboration , industry-academic collaboration
between IBM and MIT,
focused on the fundamental researchof artificial intelligence.
Over the course of decades, David Cox has watched as the AI revolution steadily grew
from the simmering ideas of a few academics and technologists into the industrial boom
we are experiencing today. Having dedicated his life to pushing the field of AI towards
new horizons, David has both contributed to and presided over many of the major breakthroughs
in artificial intelligence. In today’s episode, you’ll hear David
explain some of the conceptual underpinnings of the current AI landscape—things like
foundation models (in surprisingly comprehensible terms, I might add). We’ll also get into
some of the amazing practical applications for AI in business, as well as what implications
AI will have for the future of work and design. David spoke with Jacob Goldstein, host of
the Pushkin podcast What’s Your Problem? A veteran business journalist, Jacob has reported
for the Wall Street Journal, the Miami Herald, and was a longtime host of the NPR program
“Planet Money.” Okay! Let’s get to the interview.
Jacob Goldstein: Great. So tell me about your role at IBM.
David Cox: So I wear two hats at IBM. So, one, I'm the IBM director of the MIT-IBM Watson
AI lab. So that's a joint lab between IBM and MIT where we try and invent what's next
in AI. It's been running for about five years. And then more recently,
I started as the vice president for AI models. And I'm in charge of building, IBM's foundation models.
You know, building these big models, generative models, that allow us to have all
kinds of new, exciting capabilities in AI.
Jacob Goldstein: So I want to talk to you a lot about foundation models, about generative AI, but before we get to that,
let's just spend a minute on the, on the, IBM-MIT collaboration. Where did that partnership start? How did it originate?
David Cox: Yeah, so actually, it turns out
that MIT and IBM have been collaborating for a very long time in the area of AI. In fact,
the term “artificial intelligence” was coined in a 1956 workshop that was held at
Dartmouth, and it was actually organized by an IBMer, Nathaniel Rochester, who led the
development of the IBM 701. So we've really been together in AI since
the beginning. And as you know, AI kept accelerating more and more and more. I think there was
a really interesting decision to say, “Let's make this a formal partnership.” So IBM
in 2017 announced that it'd be committing close to a quarter billion dollars over 10
years. To have this joint lab with MIT—and we located
ourselves right on the campus, and we've been developing very, very deep relationships where
we can really get to know each other, work shoulder to shoulder, you know, conceiving
what we should work on next and then executing the projects. And it's really, you know—very
few entities like this exist between academia and industry.
It's been a really fun last five years, to be a part of it.
Jacob Goldstein: And what do you think are some of the most important outcomes of this
collaboration between IBM and MIT?
David Cox: Yeah, so we're really kind of the tip of the spear for, for IBM's AI strategy.
So we're really looking at what, you know—what's coming ahead. And, you know, in areas like foundation models, you know,
as the field changes. MIT people are interested in working on, you know—faculty, students and staff
are interested in working on, “What's the latest thing?” “What's the next thing?”
We at IBM Research are very much interested in the same. So we can kind of put out feelers:
you know, interesting things that we, that we are, that we're seeing in our research,
interesting things we're hearing in the field—we can go and chase those opportunities. So when
something big comes, like the big change that's been happening lately with foundation models,
we're ready to jump on it. That's really the purpose. That's the lab
functioning the way it should. We're also really interested in how do we advance the
AI that can help with climate change or build better materials and all these kinds of things
that are, you know, a broader aperture, sometimes, than what we might consider just looking at
the product portfolio of IBM. And that gives us a breadth where we can see
connections that we might not have seen otherwise. We can think things that help out society
and also help out our customers.
Jacob Goldstein: So the last, whatever, six months, say, there has been this wild rise in the public's interest in AI, right?
Clearly coming out of these generative AI models that are really accessible—you know, certainly
ChatGPT, language models like that, as well as models that generate images, like Midjourney.
I mean, can you just sort of briefly talk about the, the breakthroughs in AI that have
made this moment feel so exciting, so revolutionary, for artificial intelligence?
David Cox: Yeah. You know, I've been studying AI basically my entire adult life.
Before I came to IBM. I was a professor at Harvard. I've been doing this a long time and I've
gotten used to being surprised. It sounds like a joke, but it's—it's serious. Like
I'm getting used to being surprised at the acceleration of the pace.
Again, it tracks, actually, a long way back. You know, there's lots of things where there
was an idea that just simmered—for a really long time. Some of the key math behind this—the
stuff that we have today, which is amazing—there's an algorithm called “back propagation,”
which is sort of key to training neural networks. That's been around, since the ’80s, in wide
use. And really, what happened was it simmered for a long time, and then enough data and
enough compute came. So we had enough data because, you know, we
all started carrying multiple cameras around with us. Our mobile phones have all, you know,
all these cameras,
and this—we put everything on the internet and there's all this data out there. We caught
a lucky break that there was something called a “graphics processing unit,” which, you
know, turns out to be really useful for doing these kinds of algorithms— maybe even more
useful than it is for doing graphics. (They're great at graphics too.)
And things just kept kind of adding to the snowball. So we had deep learning, which is
sort of a rebrand of neural networks that I mentioned, from the ’80s.
And that was enabled, again, by data, because we digitalized the world, and compute, because we kept
building faster and faster and more powerful computers.
And then that allowed us to make this, this big breakthrough. And then, you know, more
recently, using the same building blocks, that inexorable rise of more and more and
more data met a technology called “self-supervised learning,” where—the key difference there—in
traditional deep learning, you know, for classifying images, you know, like, “Is this a cat or
is this a dog?” in a picture. Those technologies require supervision. So you have to take what
you have and then you have to label it. So you have to take a picture of a cat and
then you label it as a cat. And it, it turns out that, you know, that's very powerful,
but it takes a lot of time to label cats and to label dogs. And, you know, there's only
so many labels that exist in the world. So what really changed more recently is that
we have self-supervised learning, where you don't have to have the labels. We can just
take unannotated data. And what that does is it lets you use even more data. And that's
really what drove this latest, sort of, rage. And then all of a sudden we start getting
these, these really powerful models, and then—really this has been—you know, again, simmering
technologies, right? This has been happening for a while. But one of the—and progressively
getting more and more powerful. One of the things that really happened with ChatGPT and
technologies like Stable Diffusion and Midjourney was that they made it visible to the public.
You know, you put it out there; the public can touch and feel and they're, like, “Wow,
not only is there a palpable change,” and “Wow, this, you know, I can talk to this
thing, wow, this thing can generate an image.” Not only that, but everyone can touch and
feel and try. My kids can use some of these AI art generation technologies.
And that's really just launched—you know, it's like a propelled slingshot at us into a different
regime in terms of the public awareness of these technologies.
Jacob Goldstein: You mentioned earlier in the conversation “foundation models,”
and I want to talk a little bit about that. I mean, can you just tell me, you know, what
are foundation models for AI and why are they a big deal?
David Cox: Yeah. So this term, “foundation model,” was coined by a group at Stanford,
and I think it's actually a really apt term. Because remember, I said, you know, one of
the big things that unlocked this latest excitement was the fact that we could use large amounts
of unannotated data, we could train a model. We don't have to go through the painful effort
of labeling each and every example. You know, you still need to have your model do something
you want it to do. You still need to tell it what you want to do. You can't just have
a model that doesn't, you know, have any purpose. But what a foundation model is—it provides
a foundation, like a literal foundation. You can sort of stand on the shoulders of giants.
You can have one of these massively trained models and then do a little bit on top. You
know, you could use just a few examples of what you're looking for and you can get what,
what you want from the model. In some cases, you don't need examples at all.
So just a little bit on top now gets you the results that a huge amount of effort—you
used to have to put in, you know, to get from the ground up to that, that, that level.
Jacob Goldstein: I was trying to think of, of an analogy for sort of foundation models
versus what came before. And I don't know that I came up with a good one, but the best
I could do was this; I want you to tell me if it's plausible. It's like before foundation
models, it was like you had these sort of single-use kitchen appliances.
You could make a waffle iron if you wanted waffles, or you could make a toaster if you
wanted to make toast. But a foundation model is like, like an oven with a range on top.
So it's like this machine and you could just cook anything with this machine.
David Cox: Yeah, that's, that's a great analogy. They're very versatile. The other piece of
it too is that they dramatically lower the effort that it takes to do something that
you want to do. And sometimes—I used to say, about the old world of AI, I would say,
you know, the problem with automation is that it's too labor intensive.
Which sounds like I'm making a joke.
Jacob Goldstein: Indeed. Famously, if automation does one thing, it substitutes machines or computing power for labor. Right? So what
does that mean to say AI is, or automation is, too labor intensive?
David Cox: I'm actually—it sounds like I'm making a joke, but I'm actually serious. And
what I mean is that the effort it took in the old regime to automate something was very, very high.
So if I need to go and curate all this data, collect all this data, and then
carefully label all these examples, that labeling itself might be incredibly expensive and time
consuming. And we estimate anywhere between 80 to 90% of the effort it takes to field
an AI solution actually is just spent on data. So that, that has some consequences, which
is, the threshold for bothering, you know, if you're going to only get a little bit of
value back from something, are you going to go through this huge effort to curate all
this data? And then when it comes time to train the model, you need highly skilled people
that might be expensive or hard to find in the labor market. Are you really going to
do something that's just a tiny little incremental thing? No, you're going to do the—only the
highest-value things that warrant that level of investment.
Jacob Goldstein: Because you have to essentially build the whole machine from scratch.
And there aren't many things where it's worth that much work to build a machine that's only
going to do one narrow thing.
David Cox: That's right. And then you, you tackle the next problem and you basically have to start over and— there are some nuances
here. Like for images, you can pretrain a model on some other task and change it around.
So there are some examples of this like nonrecurring cost that we had in the old world too, but
by and large, it's just a lot of effort. It's hard. It takes, a large level of skill to
implement. One analogy that I like: think about it as,
you know, you have a river of data, you know, running through your company or your institution;
traditional AI solutions are kind of like building a dam on that river. You know, dams
are very expensive things to build. They require highly specialized skills and lots of planning.
And, you know, you're only going to put a dam on a river that's big enough—that you're
going to get enough energy out of it that it was worth your trouble. You're gonna get
a lot of value out of that dam if you have a river like that, you know, a river of data.
But it's—actually, the vast majority of the water, you know, in your kingdom actually
isn't in that river; it's in puddles and creeks and babbling brooks. And, and, you know, there's
a lot of value left on the table, because it's like, well, I can't— there's nothing
I can do about it. It's just that that's too, you know, low value.
So it takes too much effort. So I'm just not going to do it. The return on investment just
isn't there. So you just end up not automating things because it's too much of a pain. Now,
what foundation models do is they say, “Well, actually, no; we can train a base model, a
foundation that you can work on.” We don't—we don't care. We don't have to
specify what the task is ahead of time. We just need to, like, learn about the domain
of data. So if we want to build something that can understand English language, there's
a ton of English-language text available out in the world. We can now train models on huge
quantities of it. And then it learned the structure. It learned
how language—you know, a good part of how language works, on all that unlabeled data.
And then when you roll up with your task—you know, “I want to, you know, solve this particular
problem,” you don't have to start from scratch. You're starting from a very, very, very high
level. So that just gives you the ability to just,
you know—now all of a sudden everything is accessible. All the puddles and creeks
and babbling brooks and kettle ponds. You know, those are all accessible now. Um, and
that's, that's very exciting. It just changes the equation on what kinds of problems you could use AI to solve.
Jacob Goldstein: And so foundation models basically mean that automating some new task is much less labor intensive. The sort of
marginal effort to do some new automation thing is much lower because you're building
on top of the foundation model rather than starting from scratch.
David Cox: Absolutely.
Jacob Goldstein: So that is, like, the exciting good news. I do feel like there's, there's a little bit of a countervailing idea that's
worth talking about here. And that is the idea that, even though there are these foundation
models that are really powerful, that are relatively easy to build on top of, it's still
the case, right, that there is not some one-size-fits-all foundation model. So, you know, what does
that mean and why is that important to think about in this context?
David Cox: Yeah. So we believe very strongly that there isn't just “one model to rule
them all.” There's a number of reasons why that could be true. One, which I think is
important and very relevant today, is how much energy these models can consume. So these
models, you know, can get very, very large.
So one thing that we're starting to see, or starting
to believe, is that you probably shouldn't use one giant sledgehammer model to solve
every single problem. You know, like we should pick the right-size model to solve the problem.
We shouldn't necessarily assume that we need the biggest, baddest model for, for every
little use case. We're also seeing that, you know, small models that are trained, like,
to specialize on particular domains can actually outperform much bigger models. So bigger isn't
always even better.
Jacob Goldstein: So they're more efficient, and they do the thing you want them to do better as well.
David Cox: That's right. So, Stanford, for instance—a group at Stanford trained a model.
It was a 2.7 billion parameter model, which isn't terribly big by today's standards. They
trained it just on the biomedical literature. You know, this is the kind of thing that universities
will do. And what they showed was that this model was better at answering questions about
the biomedical literature than some models that were, you know, 100 billion parameters.
You know, many times larger. So it's a little bit like, asking an expert
for help on something versus asking the smartest person you know. The smartest person you know
may be very smart. But they're not gonna beat expertise. And then, as an added bonus, this
is now a much smaller model. It's much more efficient to run. We are—you know, it's cheaper.
So there's lots of different advantages there. So I think we're gonna see attention in the industry between vendors that say,
“Hey, this is the one, big model,” and then others that say,
“Well, actually,there's, there's, you know, lots of different tools we can use that all have this nice quality that we outlined
at the beginning.”
And then we should really pick the one that makes the most sense for the task at hand.
Jacob Goldstein: So there's sustainability—basically, efficiency. Another, another kind of set of issues that come up a lot with AI are
bias and hallucination. Can you talk a little bit about bias and hallucination—what they
are and how you're working to, to mitigate those problems?
David Cox: Yeah. So there, there are lots of issues still. As amazing as these technologies
are—and they, they are amazing; Let's be very clear, lots of great things we're gonna
enable with these kinds of technologies—bias isn't a new
problem. So, basically we've seen this since the beginning of AI: if you train a model
on data that has a bias in it, the model is going to recapitulate that bias when it provides
its answers. So if every time, you know—if all the text
you have says, is more likely to refer to female nurses and male scientists, then you're
going to get models that—for instance, there was an example where a machine-learning-based
translation system translated from Hungarian to English.
Hungarian doesn't have gendered pronouns, English does. And when you, when you asked
it to translate, it would translate “they are a nurse” to “she is a nurse.” And
it would translate “they are a scientist” to “he is a scientist.” And that's not
because the, the people who wrote the algorithm were building in bias, and coding in, like,
“Oh, it’s gotta be this way.” It's because the data was like that.
We have biases in our society, and they're reflected in our data and our text and our
images everywhere, and then the models—they're just mapping from what they, what they've
seen in their training data to, to the result that you're trying to get them to do and to
give, and then these biases come out. So, there's a very active program of research.
And, we do quite a bit of IBM research at MIT, but also all over the community and industry
and academia, trying to figure out, “How do we explicitly remove these biases? How
do we identify them? How do you know? How do we build tools that allow people to audit
their systems to make sure they aren't biased?” So this is a really important thing. And,
you know, again, this was here since the beginning, you know, of machine learning and AI. But
foundation models and large language models and generative AI, um, just bring it into
sharper, even sharper focus, because there's just so much data and it's, it's sort of building
in, baking in, all of these different biases we have.
So that, that's, that's absolutely, um, a problem that these models have. Another one
that you mentioned was hallucinations. Um, so even the most impressive of our models
will often just make stuff up. You know, the technical term that the field has chosen is
“hallucination.” Um, to give you an example, I asked ChatGPT to create, you know, create
a biography of David Cox at IBM. And, you know, it started off really well.
You know, it identified that I was the director of the MIT-IBM Watson AI Lab and said a few
words about that. And then it proceeded to create an authoritative but completely fake
biography of me, where I was British, I was born in the UK, I went to British universities in the UK.
Jacob Goldstein: The authority, right? It's the certainty that it, that is, is weird about
it, right? It's, it's dead certain that you're from the UK, et cetera.
David Cox: Absolutely, yeah. And it has all kinds of flourishes, like “I won awards
in the UK.” So yeah, it's problematic, because it kind of pokes at a lot of weak spots in
our human psychology, where if something sounds coherent, we're likely to assume it's true.
We're not used to interacting with people who eloquently and authoritatively, you know,
emit complete nonsense. Like—we could debate about that.
Jacob Goldstein: Yeah. We can debate about that, but yes, its sort of blithe confidence,
throws you off when you realize it's completely wrong, right?
David Cox: That's right. And, we do have a little bit of like a Great and Powerful Oz
sort of vibe going sometimes, where we're like, “Well, you know, the AI is all knowing
and, and therefore whatever it says must be true.” But, but these things will make up
stuff, very, aggressively. Um, and—you know, everyone can try asking it for their bio.
You'll get something that—you'll always get something that's of the right form, that
has the right tone, but you know, the facts just aren't necessarily there. So that, that's
obviously a problem. We need to figure out how to close those gaps, fix those problems.
There's lots of ways we could use them much more easily.
Malcom Gladwell: I’d just like to say, faced with the awesome potential of what these technologies
might do, it’s a bit encouraging to hear that even ChatGPT has a weakness for flamboyant,
if fictional, versions of people’s lives. And while entertaining ourselves with ChatGPT
and Midjourney is important, the way ordinary people use consumer-facing chatbots and generative
AI is fundamentally different from the way enterprise businesses use AI. How can we harness
the abilities of artificial intelligence to help us solve the problems we face in business
and technology? Let’s listen on as David and Jacob continue their conversation.
Jacob Goldstein: We've been talking in a somewhat abstract way about AI and the ways it can
be used. Let's talk in a little bit more of a specific way. Can you just talk about some
examples of business challenges that can be solved with, with automation, with this kind
of automation we're talking about?
David Cox: Yeah, so the sky's the limit. There's a whole set of different applications that, these models are really good at.
And basically, it's a super set of everything we used to use AI for in business. So, you
know, the simple kinds of things are like, “Hey, if I have text and I have, like, product
reviews and I want to be able to tell if these are positive or negative, let's look at all
the negative reviews so we can have a human look through them and see what was up.”
Very common business use case. You can do it with traditional deep-learning-based AI.
So there's things like that, that are, you know—it's very prosaic, sort of. We're already
doing it. We've been doing it for a long time. Then you get situations that are, that were
harder for the old AI. Like if I want to compress something. Like I want to—say I have a chat
transcript, like a customer called in and they had a complaint. They call back. Okay,
now a new person on the line needs to go read the old transcript to catch up.
Wouldn't it be better if we could just summarize that? Just condense it all down? Quick little
paragraph. You know, “Customer call, they were upset about this,” rather than having
to read the blow-by-blow. There's just lots of settings like that, where summarization
is really helpful. “Hey, you have a meeting.” I'd like to just automatically have that meeting—or
that email, or whatever—I'd like to just have it condensed down so I can really quickly
get to the heart of the matter. These models are really good at doing that. They're also
really good at question answering. So if I want to find out what's—how many vacation
days do I have, I can now interact in natural language with the system that can go and—that
has access to our HR policies, and I can actually have a, you know, multiturn conversation where
I can, you know—like I would have with, somebody, you know, an actual, HR professional
or customer service representative. So a big part, of what this is doing is it's
putting an interface—you know, when we think of computer interfaces, we're usually thinking
about UI (user interface) elements, where I click on menus and there's buttons and all
this stuff. Increasingly now we can just talk. Just in words, you can describe what you want.
You want to answer, ask a question, you want to sort of command the system to do something,
rather than having to learn how to do that, clicking buttons, which might be inefficient.
Now we can just, sort of, spell it out.
Jacob Goldstein: Interesting, right? The graphical user interface that we all sort of default to—that's not, like, the state of nature,
right? That's a thing that was invented and just came to be the standard way that we interact
with computers. And so you could imagine, as you're saying, like, chat, essentially,
chatting with the machine could, could become a sort of standard user interface, just like
the graphical user interface did, you know, over the past several decades.
David Cox: Absolutely. And I think those kinds of conversational interfaces are going to
be hugely important for increasing our productivity. It's just a lot easier if I don't have to
learn how to use a tool, or I don't have to kind of have awkward, you know, interactions
with my computer; I can just tell it what I want and it can understand. It could potentially
even ask questions back to clarify, and—have those kinds of conversations.
That can be extremely powerful. And in fact, one area where that's gonna, I think, be absolutely
game-changing is in code, when we write code. Programming languages are a way for us to
sort of— between our very sloppy way of talking and the very exact way that you need
to command a computer to do what you want it to do. They're cumbersome to learn. They
can—you know, you create very complex systems that are very hard to reason about. And we're
already starting to see the ability to just write down what you want and the AI will generate
the code for you. And I think we're just going to see a huge
revolution of—like, we just converse. You know, we can have a conversation to say what
we want and then the computer can actually not only do fixed actions and, and, and do
things for us, but it can actually even write code to do new things, you know, and, and
generate the software itself. Given how much software we have, how much,
um, just like craving we have for software—like we'll never have enough software in our world—the
ability to have AI systems is—as a helper in that, I think we're going to see a lot
of, a lot of value there.
Jacob Goldstein: So if you, if you think about the different ways AI might be applied to business —
I mean, you've talked about a number of the sort of classic use cases. What are some of the more “out there” use cases?
What are some, you know, unique ways you could imagine AI being applied to business?
David Cox: Yeah, there's—um, really the sky's the limit. I mean, we have one project
that I'm kind of a fan of, where, we actually were working with a mechanical engineering
professor at MIT, working on a classic problem of how do you build linkage systems.
Which are like, you know—imagine bars and joints and motors, you know, the things that—
Jacob Goldstein: Building a thing, building a physical machine of kind.
David Cox: Yeah, like, real, like, metal and, you know, and—
Jacob Goldstein: Nineteenth century. Just old-school industrial revolution.
David Cox: Yeah. Yeah. But, but, you know, the little arm that's holding up my microphone
in front of me, the cranes that build your buildings, you know, parts of your engines—this
is like classical stuff. It turns out that, you know, humans, if you want to build an
advanced system, you decide what, like, curve you want to create. And then a human, together
with a computer program, can build a five- or six-, bar linkage. And then that's kind
of where you top out. It just gets too complicated to—
Jacob Goldstein: Huh. David Cox: We built a generative AI system
that can build 20-bar linkages. Like arbitrarily complex. So these are machines that are beyond
the capability of a human to design themselves. Another example: we have, AI systems
that can generate electronic circuits. You know, we, we had a project where we were working,
where we were building better power converters, which allow our computers and our devices
to be more efficient. Save energy. You know, less carbon output.
I think the world around us has always been shaped by technology. If you look around,
you know, just think about how many steps and how many people and how many designs went
into the table and the chair and the lamp. It's really just astonishing. And that's already
the fruit of, automation and computers and those kinds of tools. But we're going to see
that increasingly be, you know—and it's just going to be everywhere around us.
Everything we touch is going to have been helped in some way to get, get to you, by AI.
Jacob Goldstein: You know, that is a pretty profound transformation that you're talking about in business. How do you think about
the implications of that, both for the, business itself and also for, for employees?
David Cox: Yeah. So I think for businesses, this is gonna cut costs, make new opportunities,
delight customers. You know, like, there's just, you know—it's sort of all upside, right?
For the workers, I think the story is mostly good too. You know, like—how many
things do you do in your day that you'd really rather not, right? And we're used to having
things we don't like automated away. We didn't—if you didn't
like walking many miles to work, like, you can have a car and you can drive there. Or
we used to have a huge fraction, over 90%,of the US population engage in agriculture,and then we mechanized it.
now very few people work in agriculture. A small number of people can do the work of a large number of people.
And then, things like email, they've led to
huge productivity enhancements, because I don't need to be writing letters and sending
them in the mail. I can just instantly communicate with people.
We just become more effective. Like, our jobs have transformed. whether it's a physical
job like agriculture or it's—whether it's a knowledge worker job, where you're sending
emails and, and communicating with people and coordinating teams, we've just gotten better.
And, you know, the technology has just made
us more productive. And this is just another example of that. Now, you know, there are
people who worry that, you know, “We'll be so good at that that maybe jobs will be
displaced,” and that's, that's a legitimate concern, but just like now, in agriculture,
you know, it's not like suddenly we had 90% of the population unemployed. You know? People
transitioned to, to other jobs. And the other thing that we found too is that our appetite
for, for doing more things is—as humans, is sort of insatiable. So even if we can dramatically
increase how much, you know, one human can do, um, that doesn't necessarily mean we're
going to do a fixed amount of stuff. There's an appetite to have even more. So we're going
to, you know, continue to grow—grow the pie. So I, I think at least, uh, certainly
in the near term, you know, we're going to see a lot of drudgery go away from work.
We're going to see people be able to be more effective at their jobs. You know,
we will see some transformation in, in jobs and what they look like, but we've seen that
before. And, the technology at least has the potential to make our lives a lot easier.
Jacob Goldstein: So IBM recently launched watsonx, which includes watsonx.ai.
Tell me about that. Tell me about, you know, what it is and the new possibilities that it opens up.
David Cox: Yeah. So, watsonx, is obviously
a bit of a new branding on the Watson brand. T. J. Watson—that was the founder of IBM.
And our AI technologies have had the Watson brand; watsonx is a recognition that
there's something new. There's, there's something that actually has changed the game.
We've gone from this old world of, “Automation is too labor intensive” to this new world
of possibilities, where it's much easier to use AI. And what watsonx does is it brings
together tools for businesses to harness that power. So, watsonx.AI It includes, foundation
models that our customers can use. It includes tools that make it easy
to run, easy to deploy, easy to experiment. There's a watsonx.data component, which allows
you to sort of organize and access your data. So what we're really trying to do is give
our customers a cohesive set of tools to harness the value of these technologies and at the
same time be able to manage— the risks and the things that you have to keep an eye on
in an enterprise context.
Jacob Goldstein: So we talk about the guests on this show as, as “New Creators,”
by which we mean people who are creatively applying
technology in business to drive change. And I'm curious how, how creativity plays a role
in the research that you do.
David Cox: I honestly, I think the creative aspects of this job—this is what makes this work exciting. I should say, the folks who
work in my organization are doing the creating and I get to sort of—
Jacob Goldstein: You're, you're doing the managing, so They could do the creating. Yeah.
David Cox: I'm helping them be their best. I do—I still get to get involved —in the
weeds of the research, as much as I can. But, there's something really exciting about inventing.
One of the nice things about doing invention and doing research on AI and industry is it's
usually grounded in a real problem that somebody is having. You know, a customer wants to solve
this problem, it's losing money, or there could be a new opportunity. You identify that
problem and then you, you, you build something that's never been built before, to do that.
And I think that's honestly the adrenaline rush that keeps all of us in this field.
How do you do something that nobody else on earth has, has done before or tried before? So that,
that kind of creativity—and, and there's also creativity as well in identifying what
those problems are, being able to understand the places where, you know, the technology
is close enough to solving a problem, and doing that matchmaking between problems that
are now solvable. And in AI, where the field is moving so fast, there's this constantly
growing horizon of things that we might be able to solve.
So that matchmaking, I think, is also a really interesting, creative problem. So, I think
that's why it's so much fun. And it's a fun environment we have here too—it's people
drawing on whiteboards and, you know, writing on pages of math. And, uh—
Jacob Goldstein: Like in a movie, like in a movie.
David Cox: Yeah, straight from central casting.
Jacob Goldstein: You drawing on—the drawing on the window, writing on the window with a Sharpie.
David Cox: Absolutely.
Jacob Goldstein: So, let's close with the really long view. How do you imagine AI and people working together 20 years from now?
David Cox: Yeah, it's really hard to make
predictions. The vision that I like, actually,—this came from an MIT economist named David Autor.
Which was, “Imagine AI almost as a natural resource.” You know, it's like, we have—we
know how natural resources work, right? Like there's an ore we can dig up out of the
earth that comes from, you know—kind of springs from the earth.
We usually think of that in terms of physical stuff. With AI, you can almost think of it as, like,
“There's a new kind of abundance, potentially 20 years from now, where not only can we have
things we can build or eat or use or burn or whatever. Now we have, you know, this ability
to do things and understand things and do intellectual work.” And I think we can get
to a world where automating things is just seamless; we're surrounded by the capability
to augment ourselves—to, to get things done. And you could think of that in terms of, like,
“Well, that's going to displace our jobs, because eventually the AI system is going
to do everything we can do.” But you could also think of it in terms of, like, “Wow,
that's just so much abundance that we now have. And really, how we use that abundance
is, is sort of up to us.” You know, like, when you can—writing software is super easy
and fast, and anybody can do it. Just think about all the things you can do now.
Like, think about all the new activities. Think about all the ways we could use that
to enrich our lives. That's where I like to see us in 20 years: you know, we can do just
so much more than we were able to do before.
Jacob Goldstein: Abundance. Great. Thank you so much for your time.
David Cox: Yeah. It's been a pleasure. Thanks for inviting me.
Malcolm Gladwell: What a far ranging deep conversation. I'm mesmerized by the vision David just described.
A world where natural conversation between mankind and machines can generate creative solutions to our most complex problems.
A world where we view AI not as our replacements, but as a powerful resource we can tap into and exponentially boost our innovation and productivity
Thanks so much to Dr. David Cox for joining us on Smart Talks. We deeply appdicaite him sharing his huge breadth of AI Knowledge with us,
and for explaining the transformative potential of foundations models in a way even I can understand.
We eagerly await his next great breakthrough.
Smart Talks with IBM is produced by Matt Romano,
David Zha, Nisha Venkat, and Royston Beserve, with Jacob Goldstein. We’re edited
by Lidia Jean Kott. Our engineers are Jason Gambrell, Sarah Bruguiere, and Ben Tolliday. Theme song by Gramoscope.
Special thanks to Carly Migliori, Andy Kelly, Kathy Callaghan, and the EightBar and IBM teams, as well as the Pushkin marketing
team. Smart Talks with IBM is a production of Pushkin
Industries and Ruby Studio at iHeartMedia. To find more Pushkin podcasts, listen on the
iHeartRadio app, Apple Podcasts, or wherever you listen to podcasts.