BACK TO Business ResilienceZINNOV PODCAST | Business Resilience
Artificial Intelligence (AI) has the transformative power to change how businesses across verticals are run. And its newest avatar – Generative AI – stands at the forefront of the technological revolution, offering unprecedented opportunities for innovation and creativity. By harnessing the potential of advanced Machine Learning (ML) algorithms, generative AI has emerged as a game-changing force, pushing the boundaries of what was once considered possible. Today, we find ourselves in an era where machines can not only understand and analyze data but also generate new, original content, sparking a paradigm shift in diverse sectors.
In this episode of the Zinnov Podcast, Pari Natarajan, CEO, Zinnov sits down with James E Heppelmann, CEO, PTC, to explore the effects of AI on physical design. James goes on to explains how AI and ML are already being used in products such as Predictive Analytics, 3D Generative technology, and Computer Vision.
Whether Generative AI is ethical or not, emerges a key theme in the episode. After its recent success, ChatGPT emerged as not only a great tool to reduce human effort, but also is a great enhancer of product design, when used as a learning mechanism.
The two leaders also discuss the impact of Generative AI on various industries and its potential to transform the role of Marketing, Sales, and Software Development . They also explore how AI can enhance the design of physical products, particularly from a sustainability lens.
Jim emphasizes the importance of considering the entire product lifecycle and how AI can contribute to dematerialization, regeneration of designs using greener materials, and improving energy consumption.
Tune into this episode to understand the potential of AI in revolutionizing product design and its implications for sustainability and productivity.
1:52 – How Generative AI is going to change the design of physical products
3:58 – How the integration of ServiceMax is changing PTC’s product design strategy
6:38 – How Generative AI helps with discovering problems, over rule based hypotheses
8:49 – What questions customers have for innovation in sustainability
11:43 – Designing for sustainability
13:13 – How Generative AI is helping solve for productivity and talent in organizations
15:49 – Big concerns about AI
19:25 – The role of AI in light of a talent shortage and job losses
21:17 – AI proving to be a deflationary trend
22:53 – Upcoming technologies James is most excited about
Pari: Hello and welcome to the brand new episode of Zinnov Podcast.
I’m Pari Natarajan, CEO of Zinnov, and I’ll be your host today. We have a special guest with us who needs no introduction. Please welcome Jim Heppelmann, CEO of PTC. Jim has spearheaded PTC thought leadership and development of market leading technologies. Under Jim’s leadership, PTC has evolved to become a category leader in the product lifecycle management, strengthened his industry leading technology portfolio, transitioned to a subscription business model, and pioneered the SaaS transformation of the industry. It has resulted in significant accelerated top-line growth and increased profit profitability for the company.
Welcome to this episode, Jim. It’s a pleasure to have you join us today.
Jim: Yeah, great. Well, thank you Pari, for inviting me. I’m happy and looking forward to the topic of Artificial Intelligence. There’s lots of interesting stuff happening there and it’s great to have this chance to talk to you about it.
Pari: Great. And if the last few months is all about Generative AI and the impact and disruption it is causing to the industries. But a lot of the talk and discussion is really about how is it going to change the role of marketing?
How is it going to change the role of sales? And even how is it going to change the role of software developers? Because things like CoPilot is going to do part of the job, but there’s a not a lot of talk about how is it going to impact the design of physical products. And you are a thought leader in this area across your products, be it Creo, Windchill, Vuforia, and ServiceMax.
It’d be great to hear about how are you thinking about it changing your businesses and your customer’s business?
Jim: Yeah, Pari, I think already there are three applications of AI in our products. And then maybe a fourth one we could talk about coming, let’s see. So already our products use AI and Machine Learning for analyzing, for example, IOT data for predictive analytics.
We already have a generative technology in our Creo CAD software that generates 3D shapes, sort of like ChatGPT, but it’s not generating text, it’s generating mechanical geometries for parts and products. And then we have computer vision in our Vuforia suite that actually trains on CAD models and configurations of CAD models to be able then using the camera on your phone or smart glasses to be able to recognize real world objects that match those CAD model configurations.
So those three, Machine Learning, Predictive Analytics, 3D Generative, and Computer Vision are already in products we ship. You know, I think ChatGPT and CoPilot for software, that’s interesting because on one hand PTC writes a lot of software.. So, as you mentioned, we could in theory use CoPilot and chatGPT to write our software, but of course our customers write a lot of software in their products and they have the same opportunity, really, you know, the embedded software… If you could to a large degree potentially be generated by technology like, ChatGPT and CoPilot. So there’s lots of stuff happening and I think we’ll find a lot more in coming years. But, you know, we’re already multiple steps into this AI revolution here at PTC.
Pari: And how does it now with the ServiceMax integration, now you have a view of not just the initial design of the product, but all the instances of your product, what you design or the customers use your product to design across the world with different customers. Is that going to change how you design products? The inputs and the data which is there in ServiceMax, how does it feed into the design of the products?
Jim: Yeah, I mean, I think if we think of the product lifecycle and maybe also bringing the topic of sustainability, we’re thinking about the product itself, is the product a green product? And then, are the processes of making, using, and servicing the product as green as they could be? And what can we learn in the manufacturing that would improve the product? And what can we learn in the service process that will improve the product? So, certainly adding ServiceMax to the PTC suites very interesting because now we can track the installed base of products that our customer has at their customer sites, track each and every serial number and watch its lifecycle history and what service needs does it currently have, and what service needs do we anticipate. And we can do predictive analytics… you know, the old saying, an ounce of prevention’s worth a pound of cure. So preventing a problem is generally far more efficient than fixing it after it happened.
We’d all rather prevent a heart attack than recover from one, right? So predictive analytics can be very helpful to avoid more serious downtimes, but also technologies that allow you to do remote service. You know, rather than dispatching a truck with a technician, which will consume fuel and tie up the technician for hours, if not a day or more, we can log into the remote product using IOT technologies and debug and do diagnostics and maybe reset something or change some parameters and get that product working again in many cases, not all cases, but in many cases.
So the most efficient truck roll in service is the one you don’t do. And with IoT coupled with AI, predictive analytics, we can avoid a lot of truck rolls. And we could also learn like how much energy does the product consume when it’s being used, and how could we think about ways to lower that?
So I think that there’s so much we can do with AI and with sustainability and having the ability to actually interact with the installed base of products is critical to get the learnings, what’s going on? how can I make it better?
Pari: Got it. So a lot of the new Generative AI technology will also help you in discovering some of the problem, you probably would not be able to just discover it based on rule-based hypotheses.
Jim: Yep, absolutely. And there’s a few other things about Generative AI that are very interesting, again as it relates to sustainability and otherwise, if you think about the carbon footprint of a product, a lot of it has to do first and foremost with the materials in the. Yeah, what materials did you select?
And so there’s two ways that generative can help. First of all, dematerialization, we can produce parts that include fewer materials, but the second thing is to regenerate designs using a different greener material. Let’s say, we have an existing design using a material that’s got a carbon footprint we don’t like. We could ask generative technology, show me a version of that part that uses this other material or this whole range of materials and this whole range of manufacturing processes that I think are greener and more let’s say, carbon friendly or climate friendly. And generative will come back and give us a whole portfolio of choices.
Say, well, this part that you used to make, I don’t know, steel, you could make in aluminium. Aluminium’s more recyclable. You know, this part you made in plastic, of course, plastic’s based on petrochemicals, you could instead make with this other material. But the shape would have to change because, of course, the material properties not only have a carbon footprint, but they have physical properties as well that are important.
And so, a different material might require, generally does require a different geometry. So generative can be like the best friend, a co-pilot, if you will, to the mechanical engineer, to help them understand different approaches they could take to accomplish the same ends, but with a different material and frankly, much less of it because generative design’s very stingy.
It doesn’t waste material where it’s not really required and you end up with very organic shapes typically because, turns out Mother Nature was quite an efficient engineer. So you end up with a lot of designs that looked like Mother Nature might have come up with them herself.
Pari: And what are the… and we’re talking about sustainability and if you look at decarbonization all the industry you work with and that’s where decarbonization is a big priority. Be it automotive companies, large mining companies, oil and gas companies, and when they come and visit your sustainability labs, and you have a great employee experience, customer experience center in your office. So what are the big questions they have for you in terms of innovation, how you can help?
Jim: Yeah, I mean, so first let me acknowledge what you said, which is our customers who are manufacturing companies generally have carbon footprints that you know, are not insignificant. They are consumers of energy, they produce products that consume energy. So they’re coming to us and saying, can you help me? Because to really move the needle, you have to fundamentally reconsider the product.
If you said, I’m only going to care about the product when the design is done, well, you’ve already locked in so many choices then that you can only make changes kind of, let’s say, at the margin. If you really want the entire product lifecycle to be more efficient, you have to back all the way up to the design and you have to say, now as we design this product, I need a lot more insight into the implications of the choices that engineers are making.
I mean, I think we all know that engineers, for example, don’t spend that much money, but they lock in the cost. You know, they make decisions that determine how expensive the product will be when it’s made later. So just like if you want to remove cost, let’s say do value engineering, you got to go back up and reconsider the design.
Yeah. And so just like there’s value engineering, there’s sustainability engineering, where you’d say, okay, we have this product, but let’s back up and rethink it a little bit. Let’s understand where the carbon footprint comes from, the manufacturing process, from the use of the product, from the service that might, you know, happen during the use of the product.
How can we reconsider all of that stuff? How can we design with different material? How can we design for different energy sources, you know, electric instead of, let’s say fossil fuels or something like that. And then how could we put for example, IOT type features into a product so that we could remotely monitor it and you know, service in a much more efficient way, so that everything doesn’t later require a truck roll. I mean, some stuff will, but every truck roll we avoid, moves a needle. So I think companies are asking us, PTC and you know, probably our competitors help me here because, I know I can’t move the needle if I don’t bring this knowledge upstream into engineering. So PTC helped me with technologies and analytics tools and whatnot that would help me to better understand what’s possible.
Pari: Yeah. That’s amazing…like design for manufacturing is almost a design for sustainability.
Jim: Absolutely. That’s going to be a big practice because, you know, what’s happening now, particularly with public companies in the US and in Europe for sure is there are more and more regulations saying that a company must provide reporting and transparency on its climate footprint.
And then in many cases, the company must make commitments to improve that carbon footprint over time. So when a CEO and a CFO get beat up by their investors around their climate footprint, they turn to their engineers and say, guys, help, we need to make some changes here to our product concepts so that I can report a roadmap to an improved climate footprint to my investors. And so this is really becoming a hot topic now in boardrooms and therefore in engineering departments.
Pari: Interesting. I think the impact what PTC could create here is going to be massive. It’s not being talked about that much because really when you solve it in design that could take care of lot of the other issues in the overall value chain of the product.
Jim: I mean, if I design a product that uses a material that has a bad common footprint, what’s the factor going to do, other than use that material that has the bad carbon footprint. So it really is, at the end of the day, all about the design.
Pari: You talked about sustainability. You talked about broad AI use, but how do you see AI playing a role in productivity of design engineers?
And if you at globally there is a massive talent crunch for digital employees. And the same thing with people using your products and your service is what you’re launching. How is AI going to play a role in improving that, or is this kind of skills required to use a PTC range of products going to be different in the future compared to what it is today?
Jim: I think AI will ultimately be in every kind of business. Maybe social software too, but I’m really more on the business side of things. So I think it will become pervasive and, you know, it’s already showing up in lots of places, but I think we’re probably in a phase now where the best role for AI is sort of the co-pilot idea.
You know, a helper that helps an engineer make good decisions and comes up with ideas and so forth. But still, in the end, defers the final decision to the engineer. I think we’re probably some distance away and I’m not sure it’d be a good thing anyway if we took the engineers out of the process and just let AI make products for us, we might not like those products.
But, certainly AI can be very helpful and, I think it can bring great productivity. You know, like on one hand you worry about losing jobs, but on the other hand, the real world we’re in, you can’t find anybody to hire. So it’d be nice if we didn’t need so many jobs in the real world here. So I think that tools that make engineers take another step function forward in productivity would be very welcome our customers.
One of the thing is that AI done right, you know, in our world, AI needs to be blended with physics and simulation. But here we can’t just borrow product concepts from other products.
We actually have to make sure the physics works. You can’t just say, well, here’s a design that’s like that design. You have to say and I’ve simulated it to make sure that given the loads and the material properties and the geometry, this part won’t break, for example when used in the product. It’s a bit complicated, but the ability for generative, for example, to provide great suggestions to an engineer, if nothing else, simply inspire them, but potentially to solve the problem for them, I think it’s a powerful idea.
Pari: So these are some of the advantages, right? But what are some of the concerns in your mind as you deploy AI and your customers deploy AI widely across your product suites, how they use your products?
Jim: Yeah, I mean, the first, most obvious concern is intellectual property boundaries. When PTC thinks about using ChatGPT and CoPilot for our own software coding, we’re a little concerned. We’re a little concerned that it might generate code that violates somebody else’s copyright.
And it might exfiltrate our best algorithms out into the public domain and help other people generate code. That’s like some of our best code. So I’d say that’s the first concern. Okay. The second thing is, as it relates to code, for sure, you know, generative technologies like ChatGPT and CoPilot, they copy the bad ideas and the malware ideas just as readily as they copied the good ideas. So, you know, what’s happening now is malware sources are training ChatGPT and CoPilot on malware so that it might generate for you an algorithm that works fine, but it also contains some kind of cybersecurity.And if you’re not very, very careful to study what it generated, you might just roll that into your code and there you have a problem. So I’d say intellectual property boundaries. And then learning bad ideas while you’re learning good ideas in the AI algorithms is also for me a concern,
I mentioned for a generative 3D we can a little bit solve the latter problem by making sure we bring in physics. We’re not just copying a shape or copying a shape and making sure with simulations that the physics will work. So it’s probably going to protect us from bad ideas, but I still worry about intellectual property.
You know, if my customers use generative design and it generated shapes that were built into products, and then those companies were sued because the product violates a patent somebody else had, they might come back to me and say, Hey, what’s going on? We use the tech, the shape you generated for us! So we have to be careful with that kind of thing.
Pari: As a product creator, the kind of liability you might have to sign off as you build out Generative AI could be very different compared to you would have to pick because engineers now use you as a tool to design, but you are now going to provide part of the design.
Jim: I think software companies are very worried about this co-pilot technology that generates source code, very worried about it. In fact, many companies, including PTC, at this point don’t allow our engineers to use it to generate production code. A lot of companies are doing pilot projects and so forth, but I know at PTC, Amazon, and Google and many other software companies are telling their engineers, don’t use that because the risk of accidentally copying copyrighted code or going the other direction, exporting some of our top intellectual property out into the world of ChatGPT, it’s just too risky. Again, I think generative design for 3D is different because the physics element of it. But ChatGPT is really just about copying ideas from other sources basically.
Pari: Very interesting. So, other challenge we’ve been reading about this whole job loss and it could potentially create inequality, because already there is a huge issue around the money getting concentrated with less number of people. With AI coming in, how do you see that impacting jobs equality in broad society. Have you thought about those issues?
Jim: Well, I have, and I do think that AI, let’s say in general, now going beyond manufacturing and engineering use cases, I think in general it will eliminate a lot of jobs. Now, is that good or bad?
Then we have to have that debate because, a lot of companies right now are struggling to hire people. And so like if you could eliminate the need for people that you can’t anyway hire, that’s a good thing. And then I think if you project population demographics into the future, you also say like, this labor shortage isn’t going to go away for a long time.
Like here in the US, there’s a profound labor shortage. Unemployment rates are super low, even in a bad economy they’re super low. And we’re not producing the population growth. And frankly, we don’t allow enough immigration either. So it’s a catch 22. I can’t hire the people. So a technology that might lessen my demand for people maybe is a good thing.
So, I don’t know. I think the world’s in a place right now where the population growth is slowing down and tools that reduce our need for mundane workers are probably a good thing to allow us to maintain the standard of living that we have. That’s my current opinion. Now, let me keep thinking about this because, we’ll see how it develops, but my current thinking is it’s probably okay.
Pari: And do you think it’s going to be a deflationary trend? Technology in general has been deflationary, but is AI improving productivity of people, and we are now facing a major inflation issue. But do you see this as a deflation?
Jim: I do see it being deflationary. I belong to a CEO group that meets quarterly and we just had a meeting last week and, you know, CEOs of all kinds of tech companies.
And listening to some of their stories, it’s just shocking, like anybody who’s in the publishing business, like one CEO I was talking to summarizes law documents, you know, legal briefs and stuff like that need to be summarized so that you can search through the summaries and find the one you want and so forth, you know, in legal publishing, let’s say, and they’ve looked at ChatGPT and said, wow! What it takes for a human to read a document and summarize it, it’s hours and hours and hours of work and ChatGPT can do it like three seconds. So yeah, they’re saying like, wow. I mean, we almost have to immediately shift to using ChatGPT and having the worker just review what ChatGPT generated, because we’ll be disrupted if we keep doing it manually what ChatGPT can do in seconds. So I do think there are profound, in some cases, profound productivity benefits to AI that companies will have to consider, less they risk being disrupted by them.
Pari: So Jim, so final question. So what are you excited about personally, around AI in the next three, five years? I know it’s a long time. Three months ago there was some new technology. We didn’t hear about that much, but what are you excited about in the future?
Jim: Well, I mean, personally I love computer vision and I love generative design. I love it all, but those two, three dimensional and graphical. I think with 3D design, we need to make it much more accessible. I mean, ChatGPT goes viral, because anybody right now can fire up a web browser, Google the word ChatGPT, find a hyperlink, and you’re trying it. 20 seconds from now you can be trying it.
Whereas in the world of CAD, where would I start. So I know, at PTC we plan to put generative design in our Onshape technology, which runs in a web browser. And we now have, kind of restarted the free access to Onshape program. So, you know, we hope to make generative design as readily accessible for people to try and experiment with as ChatGPT is.
And we’ll do that with Onshape and of course, maintain what we have in Creo, but try to democratize the concept. Much more so with Onshape and then computer vision this ability to study CAD models and then find those objects in the real world and then transfer knowledge about those objects that you have in the digital world in a PLM database for example, onto the objects in the real world. That’s just fascinating. I mean, I never get sick of seeing those demos. And the technology’s moving so fast too, where that whole database of PLM information can be automatically generated into augmented reality artifacts and delivered out in the physical world right on top of the products. I mean, it’s just fascinating. So I’m very excited about that type of stuff as well.
Pari: Thanks, Jim. PTC has always been not just driving innovation, but setting the vision for the industry, and kind of guiding the industry to achieve that vision, similarly around AI, especially around the physical world, AI use of designing of the physical world.
I’m pretty sure PTC and you would continue to set the vision and guide the industry towards the future.
Jim: We’re certainly going to try to keep doing that. It’s a fun time and there’s lots of people at PTC including me, who love technology and love thinking about where could we go with it.
Pari: Thanks for sharing your vision around how AI is going to play a role in not just your business, but your customer’s business, and even more broadly in business software. And also how is it going to impact the society overall going forward.
Watch this episode with Pari Natarajan and Alan Trefler, about digital transformation's evolution in autonomous enterprises. The leaders discuss how decentralized decision-making can increase efficiency, and ensure business growth.
In this episode, Rajat Kohli and Alyssa Fitzpatrick discuss how partnerships in the global technology ecosystem can revolutionize innovation, profit and growth. Using real world examples, Alyssa talks about her partnering journey over the years.