Evangelos Simoudis says this is not the first “AI spring” he has witnessed.
In fact, it’s at least his third. “The optimist in me sees these advancements in every one of these AI springs,” he says. But “the pessimist says that marketing is always ahead of engineering, and in every one of these springs, marketing has overpromised, and engineering, in a sense, has underdelivered to these promises.”
Simoudis is the Founder and Managing Director of Synapse Partners, where he invests in early-stage startups that develop enterprise software and AI applications, and also advises big companies and governments on the use of data and AI to address strategic problems.
He’s also the author of several books, including Transportation Transformation and The Flagship Experience, published in late 2023. We spoke in April 2024 about AI, universal basic income, Silicon Valley innovation labs, and more.
You can subscribe to our podcast, “Innovation Answered,” on Spotify, iTunes, Stitcher, or Google Podcasts.
Scott Kirsner:
Welcome. You’re listening to the Innovation Answered podcast.
Innovation Answered is the podcast from InnoLead, the web’s most useful resource for corporate innovators and change-makers. If that sounds like you, we encourage you to subscribe to the podcast, so you’ll catch all of our future episodes.
I’m Scott Kirsner, CEO and Co-founder of InnoLead and in this episode, we talk AI, employment disruption, UBI, corporate innovation, and the future of the automotive industry with Evangelos Simoudis, an author, investor, and advisor based in Silicon Valley.
At Synapse Partners, he invests in early-stage startups that develop enterprise software and AI applications and also advises big companies and governments on the use of data and AI to address strategic problems.
He’s also the author of several books, including Transportation Transformation and The Flagship Experience, published in late 2023. He is also a close observer of the corporate innovation and startup ecosystems in Silicon Valley, and he writes a great blog called “Re-Imagining Corporate Innovation with a Silicon Valley Perspective.”
…
Scott Kirsner:
I’m curious to ask you, what is the current vibe in Silicon Valley right now? What are people excited about? What, if anything, are they anxious about?
Evangelos Simoudis:
This is a very good question and I’m glad you added the word anxious because, obviously, everybody’s excited about AI, and the various applications, both small and large. But they’re also anxious about both the state of the funding environment as well as the implications of AI, as well as implications on talent.
We can take those in any order you want. It’s a lot to unpack, but there is both excitement and concern. I’ve been in Silicon Valley since 1990. So, this is not the first time that we’ve seen this bringing together of both anxiety and excitement. My perspective is colored by the fact that I’ve seen three other AI springs, as I call them. In the 80s, the early 90s, the early 2000s, mid-90s, and I will say that the optimist in me sees these advancements in every one of these AI springs.
The pessimist says that marketing is always ahead of engineering, and in every one of these springs, marketing has overpromised, and engineering, in a sense, has underdelivered to these promises. Even though engineering has made tremendous advances in every one of these AI springs. With regard to your question, I think we will see a tremendous advance, again, significant, tremendous work —whatever adjective you want to use. In three to four years, I do not know what happens beyond that. I am not that big of a futurist, but we will see advances that will have major implications in how we do certain things. And to your earlier point, it could very well be that certain professions and certain occupations become obsolete.
Scott Kirsner:
It seems like you’ve been around the block a few times here, and that you’re not painting a picture where, 10 years from now, you and I are playing guitar around a campfire because we don’t have anything to do and the AIs and robots are running the economy.
Evangelos Simoudis:
This morning, I was with a gentleman from Sweden, and we were talking about the problem and I mentioned to him an experiment that was run by one of the Swiss cantons. For a year, they said, we’re going to guarantee you a comfortable monthly income, you don’t have to do anything, just stay home. Stay home, we will give you the money, and within a couple of months, they saw that there were major, psychological problems. These problems emerged because, for most people, their self-worth is associated with their ability to produce. Now, as I was saying to this guest this morning, there might be some sort of a 10%, maybe 15% of the population that can keep themselves occupied. You talked about the guitar, maybe you say, “Okay, I’m going to go learn a new instrument, and I’m motivated, I have the desire, I have the questions in my mind.”
But the majority of the population will not have these motivations. They have much more constrained thinking. I’m not trying to be derogatory, here, it’s just how we are brought up. If we start, you’re developing in a different way, from our very, very early formative years, maybe by the time that we are 30, 40, or 50, and we’re given such an opportunity, we will approach it very differently. But in the way that particularly Western global communities come together, I think that transition will be difficult.
Scott Kirsner:
You could imagine a scenario though, where similar to the Works Progress Administration during the Depression, you have governments that say, “Hey, we have a list of 100 things we want to get done from cutting hiking trails through the mountains, to beautifying parks, to painting murals to helping be a crossing guard as little kids walk to school in the morning. You could imagine potentially creating jobs that better society in some way, not just saying, stay home and watch Netflix?
Evangelos Simoudis:
I do not know and the reason I’m saying I do not know is that, in the last 70 years, not only haven’t we had major wars, the West has not, but we have, we have gotten accustomed to a lifestyle that making that transition that you’re painting, I think will be hard for most. I never say impossible. If you have to bring food on the table, there are different needs that can motivate behaviors.
Scott Kirsner:
Right, but we’re mapping out the future right here on this podcast, right? I do think you would have to have a menu of lots of different things that you would do to get that universal basic income. I don’t quite understand the experiments that are like, well, we’re going to give you this universal basic income and you don’t need to leave your house, you don’t need to do anything productive.
Evangelos Simoudis:
The point that I’m making to the organizations we advise is that this type of change require a lot of forethought. Unfortunately, in most cases, whether it is government or corporate, I don’t see that forethought being put in place for the type of changes that we envision. The picture that you’re painting is not a three-month study and then we put together something in motion.
Scott Kirsner:
I also think if we talk about corporations for a minute, most corporations are doing what you refer to with AI, they’re these little point experiments of getting rid of 20 percent or 30 percent, or 40 percent of our software developers or our customer service people, they don’t really have a vision of like, what are we trying to become? Are we trying to become a different sort of organization?
I use the metaphor sometimes of, “Do you want to be a fully-automated vending machine Amazon style of organization where the fewer people you employ, the better?” That’s one vision and maybe you want to map your path out to that versus a hybrid organization where it’s humans augmented by AI, or maybe it’s a purely AI-free organization. Your amazing winery in Sonoma Valley and it’s all about humans picking the grapes, making the wine, bottling the wine, and AI is not part of the brand or part of the story.
I’m curious if you agree or disagree, maybe you’ve seen some organizations where they actually do have a clear vision of here’s where we’re trying to, here’s what the destination looks like for us.
I see more and more short-termism. I was talking to a few investment analysts, and they say the CEOs are thinking about two-quarters or three-quarters ahead, they’re not thinking about five years ahead…
Evangelos Simoudis:
I will say that I have written above this point, I see more and more short-termism. I was talking to a few investment analysts, and they say the CEOs are thinking about two-quarters or three-quarters ahead, they’re not thinking about five years ahead because they’re thinking if my average tenure is going to be two and a half years, why do I need to think five years ahead?
I’m painting with a broad brush here. But even Wall Street, you look at stock buybacks, companies are being rewarded for sending money to shareholders as opposed to investing in in R&D, regardless of whether they need it. Think of Boeing and how much money they have sent to shareholders. I’ve read somewhere recently, over $60 billion over the last 10 social years. Sixty billion dollars from Boeing, by the way, is over three programs. Seven or eight class programs. Here’s a company that has been gutted with regard to its manufacturing ability,
Scott Kirsner:
When you look at not just the quality and manufacturing problems, but the problems they’re having keeping up with SpaceX in commercial space, if you believe that there is going to be a significant space economy going forward, isn’t that a place you need to win?
Evangelos Simoudis:
Exactly. The automotive industry is the same thing. Analysts are questioning why are we investing in electrification? Why don’t you keep building gas guzzlers and send the money that you gain as profits back to the shareholders? This creates some interesting dilemmas in terms of what corporate vision can be and who sets a vision and why.
Scott Kirsner:
What’s your view on the future of corporate innovation and R&D in light of that short-termism? Is it like some companies have a long-term view and are long-termists, and willing to invest over time and other companies are not and you just maybe need to acknowledge what kind of company are you working for?
Evangelos Simoudis:
Yes, I think that that very much is the case. These days, you have to think about that globally. You cannot think of it as just us or just the European Union or whatever. There’s no question that corporations in many industries are challenged because digital technology is in both hardware and hardware-related areas like robotics. Digital technology is evolving very rapidly and it’s impacting more and more corporations and more and more business processes. That oftentimes makes it hard for a corporation to think of how can I potentially change my business model while I continue to produce given my existing successful business model?
I’ve often written that established corporations need to be ambidextrous. They need to know how to keep producing under their successful business model, and then at the same time, need to be thinking, “How do I disrupt myself with a new business model?” I think more recently, we’ve seen this in the automotive industry. We’ve seen a situation where you need to be producing the vehicles that you have and offering them through your distribution system, which is mostly dealer-based. On the other hand, you have to be thinking about how to create a new business that is going to consist of these automated or autonomous vehicles that are going to be available, maybe just as a service, or if I’m going to offer them directly to the public or through dealers? In a B2C model. We’ve seen incumbent automakers struggling with that. These are not easy transformations.
Scott Kirsner:
That’s a great segue to your book, which came out in December 2023. It’s called The Flagship Experience: How AI and Software-Defined Vehicles Will Revolutionize the Automate Word of Customer Experience. Would you read a short passage from the beginning of the book?
Evangelos Simoudis:
Sure, and I have to apologize for my accent. But this is why I have the audiobook, it is narrated by somebody who has a much smoother accent than mine.
“The automotive industry has entered a period of radical transformation. Automakers that fully participate will become customer-centric corporations, change the way they conduct business, and tap new revenue streams. During the 135 years of its existence, the industry has experienced two transformations of similar magnitude, both of which were vehicle-centric. The key drivers of this transformation are the transition from fossil fuels to renewable energy to address climate change, are changing mobility patterns, and disruptive technology innovations in battery semiconductors, software, and AI. The context for this transformation is provided by new mobility, a global movement to redefine how we transport people and goods. New mobility changes the rules for the automotive industry because it brings together social awareness about the environmental impact of transportation modalities, vehicles based on new architectures and power trains, digital platforms that have managed transportation, and new business models to monetize it.”
Scott Kirsner:
So let’s talk about software-defined vehicles. Are there any cars today that really fit that description? Or is this more a vision of where we’re heading?
Evangelos Simoudis:
The term software-defined vehicles is obviously an overused term, and often a term that each company tries to fit to its product. There isn’t a universally accepted definition. In general, we refer to software-defined vehicles as those which are software-centric. Their functions are enabled by software that is high-level software or software that is more akin to what we see in enterprise software and in enterprise systems, as opposed to what we see in today’s vehicles, which tend to be low-level microcode-based software.
Today’s vehicles, like the non-software-defined vehicles, have a lot of software in it. And I talked in the book about how much software they have. Software-defined vehicles are software-centric, they are compute-intensive and intelligent. That intelligence is exhibited by the driving automation that they provide. Even if it’s not fully autonomous, the management of the energy that they do, most of the time, is about battery and electric vehicles, and also other services that can be built around the vehicle. All of these vehicles, another one of their characteristics, and particularly because of the software, the type of software that defines them, can be updated over the air. Which also means that they’re their capabilities expand over time. So the vehicle that you buy today may not be the vehicle that you’ll have six months from now.
Scott Kirsner:
Given that you were a visionary about the future of the auto industry, what do you drive today? Or if you have more than one car, what are your cars?
Evangelos Simoudis:
So let me let me answer your previous question. I’ll get to that. The short answer is I don’t drive a lot. I tried to use other modalities other than me driving so I drive a BMW 750. Which is not a particularly fuel-efficient vehicle as you can imagine, but I don’t drive it more than 2000 miles a year.
Scott Kirsner:
I assume you live in the suburbs of the Bay Area somewhere?
Evangelos Simoudis:
I do. So I take the train a lot. Many times I take a combination of train and ride-hailing. I try to walk, at least, a mile a day. In the software-defined vehicle environment, they have to care because there will come a time, and I know based on plans that I’ve seen from other automakers, nonincumbents, we call them newcomers, there will come a time that we will buy a vehicle that will address 70% of our needs and the other 30% will be filled during our use of the vehicle. We will also make, and I’m saying this in the book, we will also make a big distinction between the owner of the vehicle and the user of the vehicle.
Today, the distinction is only within our household, my wife drives my car, and I drive her car. But imagine, and this is something by the way that Musk started alluding to, Imagine that I have a car and I feel comfortable for whatever reasons to say, “You know what, I only drive 2000 miles, it’s okay. If the vehicle is driven 10,000 miles a year for the other 8000 miles. I will find a way to make money out of it.”
Scott Kirsner:
Turo is already an example of that, right? You have Turo and other car sharing that say, “Yeah, I don’t drive my car much, maybe one or two days a week, let me rent it out.”
Evangelos Simoudis:
But here’s the difference, the difference is that if you come to drive my vehicle and you have specific requirements, you will be able to customize my vehicle, if it is truly software-defined in the 70%-30% in a way that you want to. If you want the vehicle for a day, for that one day, you will have your customer, quote-unquote, customized vehicle.
That’s where we’re heading. That is a software-defined vehicle. Now, going back to your question, we’re not there yet. Companies like Tesla, Nio, the newcomers are coming closer to that definition. There are, as I explained in the book, there are different levels of software-defined nests. You can call it that, based on the architecture that the vehicle utilizes. Based on that architecture, the vehicle is more or less customizable. We’ll get to that point. We’re not there yet. We’re starting to make strides. The OEMs, newcomers, and incumbents are introducing, they’re talking about software-defined vehicles, but they’re not articulating why the consumer, why the user should care about that software-defined vehicle. I know why the OEM, the automaker, should care. They haven’t articulated why the user should care. That’s what I think is one of the reasons why we’re even seeing slumping sales right now for software-defined electric vehicles.
Scott Kirsner:
Okay, you’re mashing up two things there with EVs and software-defined vehicles. I want to ask about a third thing, which is autonomy. I do think that there’s customer demand for autonomy and yet, human expectations and human experience of autonomy are going to be challenging. The example I would use is my dad is getting on in years, he’s definitely starting to think about at what point am I not a safe driver? So he got the Tesla and bought the FSD. His partner gets anxious when he uses the FSD because she doesn’t think it’s good enough and it does some things that scare her. So if his partner is in the car FSD is not on. If she’s not in the car, FSD is on.
This is a long way of asking where is autonomy today and how do you think about that, not just the technology development, but our embrace of full autonomy, whether it’s for our personal vehicle or a robo-taxi that we get into one day in Singapore?
Evangelos Simoudis:
First of all, I’m glad that you’re making that distinction, because many folks do not make the distinction and they only think of autonomy in technology terms. I think we’re making good strides on the technology front. Even though, we were talking earlier about AI, it is taking us a lot longer to realize the level that we want it to be at and it’s costing a lot more money. As we’re seeing, whether it is in San Francisco, whether it’s in Shanghai, whether it is in the corridor between Phoenix and Dallas, we have real-life examples of autonomous vehicles, Robo taxis for logistics, being used. The problem is user acceptance, user acceptance in the use case that the vehicle is operating. User acceptance of a robo taxi in San Francisco may be very different from the user acceptance robo taxi in Shanghai.
As I’ve said, both in my second book, and in this new one, I think new mobility, whether it is expressed as autonomous vehicle transportation or software-defined vehicle mobility, is going to be very regional. How it is going to be adopted in Southeast Asia is going to be very different from how it’s going to be adopted in Europe, and how it’s going to be adopted in France. Within Europe is going to be very different than how it’s going to be adopted in Germany. I think the model of having produced one vehicle platform for the whole world and everybody drives the same thing in the same way with new ability doesn’t hold.
If we’re thinking that I’m going to develop an autonomous robo taxi, and I’m going to see global deployment with 2 million vehicles, I don’t think that’s possible.
Scott Kirsner:
I do wonder about what the economic assumptions are when people think, as you’re saying, “Oh, great, Uber one night buys a fleet of robo taxis and dispatches them across the United States or across the world and kind of discovers, oh, 100% of people do not want to get into a robo taxi, or 100% of people don’t want to sell their personal car and rely entirely on Robo taxis.” Which may be another underlying assumption here.
Evangelos Simoudis:
Again, going back to how we started our conversation today, one of the investment bets, if you want to call it we’ve made, we invested in a startup called Divergent 3D in Los Angeles because they have developed a particularly innovative AI-based manufacturing system that drives additive manufacturing. They are applying it very successfully to both the automotive industry and a couple of other industries. But their idea for the automotive industry is that we will have this type of system, this kind of manufacturing system, that will enable automakers to have micro-factories, meaning I’m going to have a factory for Southern California, and it is going to produce cars cost-effectively. So, I will not need to have an assembly line that spurs 200,000 vehicles a year in order to make it profitable. I can make it profitable to 10,000 vehicles a year that have the characteristics of the Southern California territory.
Scott Kirsner:
If you think about the last three years, I mean, you’ve been an observer of these corporate innovation labs and R&D centers that have popped up around the valley and some in San Francisco. What’s been happening to them over the last three years as we’ve come out of the COVID pandemic?
Evangelos Simoudis:
We have significant shrinkage in a number of cases. We’ve seen centers closed down. In other cases, we have seen them being reduced in scope and obviously people. There are very few companies, based on what we had pre-pandemic, that maintain, in maybe some cases even expanded their role, their innovation outpost and innovation centers.
Part of it is because, over the 10 to 12 years, corporations in many industries saw that the early-stage startups that they have been investing in, do not actually help them in the way that they envisioned. In some cases, the corporation does not have the patience to see the startup’s product or service being incorporated to their own business model. In some cases, they cannot retain the people from the startup if they happen to acquire it. They also came to realize that, with a few exceptions, the investments that they made, if they do not produce the financial returns in the timelines they envisioned…when we have periods of quick exits, then maybe they could go to their shareholders say we invested 100 million and we produce 200 million or 300 million. Maybe that was an important metric.
But by and large, I would say that there is a realization that these endeavors did not produce the expected results. So we see a pullback. We’re already working with a couple of corporations that have asked us to help them divest their portfolios of startups. To me, that’s just an indication of what is the state of their thinking.
Scott Kirsner:
I also think I mean, you touched on short-termism earlier, you touched on the challenge of really successfully investing in startups that return money while successfully acquiring startups that you can integrate. I think these innovation centers in the Valley also had the challenge of hiring the best talent because, frankly, the best talent was getting hoovered up into Apple, Tesla, Amazon, and Facebook and not Acme Company’s lab in Palo Alto, right?
Evangelos Simoudis:
There was always this tension between, hiring somebody who I know nothing about and doesn’t come from my industry, versus putting in the position within the outpost somebody from my corporation.
A lot of times, inappropriate people for that past innovation position, were moved to these outposts because their corporate management trusted them. Low risk, but the environment was completely foreign to them, which meant that their effectiveness was relatively low.
Scott Kirsner:
Let me ask you one very last short question, which is if you’re advising me, white collar or today purple collar professional, about what I should be doing as we enter this age of AI, that’s what I’ve been calling it, you probably would call it like the third coming of AI or the forthcoming of AI if we go back to you know, 1956 and the Dartmouth Summer Conference, which I think is the very origin of the term, but what would your advice be to a white collar professional in terms of getting prepared and getting up to speed for this era?
Evangelos Simoudis:
I would say definitely make sure you understand it and understand its implications for your both current and future professional endeavors. Don’t see it as a panacea because panacea, which could drive you to desperation and despair, as well as a panacea to say, “Okay, that’s all I need.” We’re not there yet and every time we have tried to see it as such, we were always disappointed.
Scott Kirsner:
Love it. Evangelos, it has been so great having you on the podcast. I really appreciate the time.
Evangelos Simoudis:
Thank you. I always enjoy conversations where they have some, in a sense, expected questions and some unexpected questions and this one definitely fulfilled that.
Scott Kirsner:
You can find out more about Evangelos Simoudis at SynapsePartners.co or Corporate-Innovation.com.
You can learn more about InnoLead, sign up for our email newsletter, or find out about our upcoming in-person event, Leveraging AI, taking place May 16th in Silicon Valley, at innolead.com.
We have a great group of speakers lined up from Amazon, Airbus, Black & Veatch, Mastercard, Swisscom, and Crown Equipment, and we’re holding the event at the PACCAR Innovation Center. It is an event you do not want to miss if you’re involved in AI strategy and execution in a big organization.
To listen to more than 60 of our earlier podcast episodes, search for Innovation Answered on your podcast platform of choice.
Thanks to David Swope and Hadley Thompson for their help on this episode.
