Apple Podcast Link
Entrepreneurs Paul Lam (Bot Auto) and Ali Arab (General Autonomy) discuss building safer, more practical autonomous vehicles—Bot Auto focusing on full-service L4 autonomous trucking and General Autonomy advancing AI safety systems for edge-case control. They share with Jim Barrood candid startup lessons on responsibility, capital discipline, and the balance between innovation, commercialization, and safety in real-world autonomy.
Paul Lam: I'm the Chief Strategy officer and head of finance for Bot Auto, also, part of the founding team.
Ali Arab: Great Ali. I'm founder of General Autonomy and executive director at the company. We are still in the middle of stealth mode from Prese to seat. At the same time as a research scientist, I do research in safety systems control and ai.
Jim Barrood: Wonderful. Alright, Paul, let's circle back to you. Tell us how you got to where you are now. You can take it from high school or college. Let's learn about your journey.
Paul Lam: Sure. So, let's start with where am I today? Bar Auto is a L four autonomous trucking company in us. I started as a philosopher, So, that's my undergraduate degree, and decided to do a little bit of economics.
And then, So, I did my PhD and I dropped out from University of Cambridge across the other side of the pond. I joined investment banking as a sales trader. What I really did was running with coffee and making sure that I remember what every sales and what every traders really like. During the tea break, not So, much about like trading stocks and all those sexy things that prompted me to think, Hey, I want to do something more useful in finance, or do I want to continue this?
I didn't have many choices, but. I found this opportunity like doing transportation business services and that's investment banking. So, I jumped on the other department. What's really interesting back then was that nobody think about technologies. Like people are thinking about business services, payments were being seen as a financial institutions and a lot of technologies were being seen as like testing, inspection, utility services.
So, if you think about D-M-V-G-L, aplu, these are all the companies that provide testing services. So, there wasn't a tech team. Anything that we call a physical AI today, it's basically just checking the smart meters and making sure that yeah, the electricity goes to your home properly, right? Nobody talks about it as a physical ai, AI in electricity, but that also, opens a brand new door for me.
Because I started to realize, hey, 10 years ago, everybody is thinking about using technologies in a commercial world, using it in FinTech in, in e-commerce. But very few people thought about like, how do we use technology to make the world that we live in better? And when I say the world I mean in a physical world, the transportation, the electricity, the energy, the city, the water, the utility, like everything.
That we count today as public utilities. So, that opens up the door, and that's how I have developed my career over the last 10 years. Like my motto is, make AI useful, make internet useful. And the definition to me is it has to be in the real world.
Jim Barrood: Got it. Okay.
Ali Arab: Go ahead. I will take the opposite approach from Paul.
He started from right now and then went back. So, start from the teenager. So, I was always on the wheels. I was a skater, professional roller blade, aggressive skater. And the journey for me to be always on the wheel. It's like just my whole family know that like just either it's me on my skates or it's my skates on my shoulder to go to the skate park.
So, it was either this or that. But moving forward I was not into engineering or like a very good student in reading at high school. I was only good at math, So, I got accepted to college because of my math and. Very luckily it was robotics engineering back in 2007. No one knew what robotics engineering will look like and everyone was telling me you should go study civil engineering.
Like you have a job after, like robotic you, you want to go fix toys, what you want to do. So, it was like conversations like people from electrical engineering, civil engineering, like more established engineering programs were like telling us and making fun of us. But I found a challenge because.
I saw right now a lot of the robots in the industry are like just automation. They just do repeated tasks. But the vision was there. I started reading books. It was one book that I read during my undergrad back in around 2010 by Dr. Ilan Norberg, professor Ilan Norberg from Carnegie Mellon, and it gave me the vision.
Robots are becoming autonomous, robots are going to go to people's houses, robots are going to be on public roads. It gave me that vision and since 2010. I was always focused on autonomous robotic systems. My first publications on autonomous mobile robots are around 2011. First patents in Iran that I had about autonomous robots.
It's around that time and I realized this is a very challenging technology. So, I got my masters in robotics, my PhD in robotics. All from different aspects, from building parts for robots to writing codes at Bell Labs for safety of autonomous mobiles, robots on the factory floor, and building a safety framework for neuro, one of the autonomous startups.
So, I'm very proud of like that trajectory that brought me to here to found Gen Auto with the belief that I have enough disparity to help making autonomous vehicles safer.
Jim Barrood: Got it. All right, wonderful. So, back to you, Paul. Tell us about Bot Auto. Tell us how that got started and how you've been involved and what the vision is for the future.
Paul Lam: Sure. So, Bot Auto started in Houston, Texas and we're a purely US-based a hundred percent L four autonomous trucking company. What we do is develop the algorithm that makes the truck drives itself. We also, own and operate the truck, I see Jim, you're sipping something. I suppose that's coffee. So, I always love to use the analogy that we sell a cup of coffee, not coffee beans.
So, the differences really being if I'm an L 4 autonomous driving software company, what really I'm selling is coffee beans. For you. Consume that cup of coffee. You need to get your own coffee machine. You need to brew it. You need to get, hopefully, the right water with the right mineral contents to optimize that cup of coffee taste.
And if you are starting a cafe, you also, need to have barista. You also, need to get pastry to make it work. So,, what we really do is not just selling software or licensing software as a service. What we really do is to guarantee the outcome. We sell the cup of coffee. So, our product is basically hauling freight for any shippers around the world.
We work with the freight brokers, So, you may like to call it our autonomous driving is our internal IT function. When a customer comes to us, you don't come to buy that almighty. Autonomous software stack. Instead, you're trying to just have us hold us accountable to deliver your good from point A to point B.
Just like any other logistics company, we started in 2023, July 27th, I believe. So, a little bit about more than two years. So, today we have raised more than 60 million. We have developed our technologies and. Our video is online YouTube, that you can see Our trucks is running fully without a human from a commercial yard at Houston, and we are building that route out to San Antonio.
And further So, next year we'll start our continuous commercial operations as well.
Jim Barrood: Fantastic. And what about you, Ali? Tell us how Jan Auto what sort of segment of the market that you're working in?
Ali Arab: So, I consider ourselves as a technology provider. So, we started as commercializing a deep.
Research from like a prestigious NSF grant, that it was part of my PhD funding, and that was based on controlling autonomous systems during the critical and disastrous situations. To make sure we are able to mitigate disasters, to minimize the damage. That was the research that I've spent like the past decade of my life on.
So, we believed having that amount of knowledge for that long time, we can build tools and solutions that the whole autonomous driving ecosystem can use it. So, if I want to simply explain it is. Like microscopes when they realized like IES exist. So, they started looking deep and like just having this kind of enlarged vision of just those small components to find what kind of behavior they have to find a solution for fighting against them to make a penciling or make any drugs that it can stop them.
So, I believe for autonomous driving there exists. Situations that needs that amount of granularity and zoom in. So, that's why I call it like just we have to zoom in on the H case scenarios. Right now, a lot of people talk about H case scenarios. We have data. We can't find h case scenarios. We can generate HH case scenarios.
So, when you have the H case scenarios, you have to zoom in to build safety Solutions for behaving properly. You make sure. You are minimizing the damage. So, that's very deep technical. We are building Solutions step by step. We have raised less than half of a million from a federal state, and. Awards and grants and mostly non-dilutive.
So, we started looking into seed investment In the past two months, we engaged with couple of VC and other autonomous vehicle companies that they are interested to partnering with us to build Solutions and technologies that they can, use it for faster safety assessments. So, that's basically the opposite of what Paul was saying.
Jim Barrood: No, I think that, and it's really important, right? Obviously those edge cases, right? How to be how to ensure safety, right?
Ali Arab: Like this, So, this is like very edgy and sharp.
Jim Barrood: Exactly. So, Paul, how do you view that? Obviously you guys are, have, overcome some of these challenges, but what keeps you up at night as far as safety and things like that?
Paul Lam: Yeah, no, absolutely. I think safety is everything. And again, this is more a market positioning problem, right? Going back to my cafeteria example, right? Starbucks don't make all their pastries. They partner with the local bakeries and to get the croissant right? And it's really important to understand what is the line between commercialization, safety, efficiency, and all the balances.
So, there are two doctrines that. Or at least me personally as part of the founding team that I hold. The first one is you have to use first principle. And to me, the first principle is not just technological. The first principle is also, about value creation, right? What do the customers really want? What do we face as really the business problem today facing our sectors?
You and I both drive. Driving is a verb. Driving is an action. Driving is not a Source of outcome or driving is not a source of value creation. To certain people, driving would be, if you really enjoy driving and cruising, that would be a different matter. But most people drive for a purpose. So, you and I might be driving our kids to school and driving is a means to an end and end is getting educated, driving to work, and that's productivity.
So, ultimately for us, when we build a business, we need to ask ourself, is autonomous driving really a product value physician? No, not really. It's really, is it RoboTaxi? Is it autonomous logistics? Is it autonomous school buses? So, that gives us the focus of what customers really want. But today, if I go out to FedEx to USPS, to anyone to say, Hey, I have this great technology that will be L four autonomy, people would say, did I say I want it?
Did I say drivers are a pain that because I don't have an autonomous driving software, I cannot deliver your mail, I cannot deliver your parcels. Not at all. There hasn't ever been available positions like that. So, this is really important for us to understand that where the lines are. And that's number one.
And number two, this is a real conversation. So, bought auto just launched our Marsh insurance program, and this is about safety. You can never optimize safety if you are not ultimately responsible for the thing. So, think a little bit about it. If you're going to a hospital, you're going into a surgery, you're getting on a surgery table.
The hospital is basically a, I don’t know, a gathering point for all the outsourced doctors and surgeons over it. That nobody is a resident there. You panic. Is that, am I coming to a hospital getting treatment or am I really coming to a place where there, there are a lot of outsourced staff helping us and this is, this sounds a little bit crazy, but this is exactly what we're seeing today, that in, in a lot of autonomous driving.
Ecosystems, we're trying to partner with everyone, yet nobody is holding the ultimate responsibility. So, it's like I'm selling you the software and then you have a truck manufacturer, and then the customers will buy a trucks from someone and they activate it. Guess what? Everybody is trying to optimize their very local maxima and try to claim that this will be a safe platform.
Yet nobody is really owning the ultimate responsibility. And if I. Take that as a step forward. I think this is one of the reasons why Waymo has been pretty successful, because ultimately they asked a very fundamental question, do I really need the customers to understand how many lidar do I use? Or trying to get into that debate whether camera is better or LIDAR is better.
That's great for engineers. The customers do not need to know that. All they need to know is this safe. Are you responsible for giving me the product, moving me from point A to point B? How much am I going to pay for it? So, I think there is a lot of reflections here. Don't mean to go too deep here but I think safety is really important.
And all I want to say is you could only have safety by having really great technologies. And this is why what Ali is doing is instrumentally important. Because you need specialists who focus a thousand percent of their time on safety. That is inevitable, and it's important to the ecosystem. But on the other hand, you really need to build a business model where the interests are aligned, that there is no more hazards, there are no asymmetric inflammation.
There is no adverse selections, and the only way to do it is that in the very early time and the early stage, when you develop the business, you need to have that holistic approach cookie cutting and trying to split the sector. We don't even have autonomous driving. Being fully completed and trying to split the responsibilities sound a little bit crazy to me.
And I, I think it's really important to have the integrated approach if you want to be responsible for safety, responsible for efficiency, and hold ultimate fiduciary duty to all your customers.
Jim Barrood: For sure. Ali, talk to us about how you view that as well, about the whole industry. But also, do me a favor and do us a favor, explain what L four means and how that, you know what, when we talk about Waymo, what does that mean?
And we talk about the robotaxi that's coming from Tesla. What does that mean? Give us a snapshot from your perspective. Sure.
Ali Arab: Yeah. So, I will start from the easy one, which is L four and Robotaxis. When autonomous driving started becoming like more, no, So, like standard com committees, like S-A-E-I-S-O started writing like more formal definition for the whole ecosystem.
So, levels were, I think, initially defined by SAEA, Society of American Engineers. And level four means you have a full autonomous driving feature without the need of it. Driver in the vehicle to engage to take over the steering wheel. Only within certain operational design domain. They call it ODD, and that ODDs are different for any manufacturer or like any autonomous vehicle startup, I'm sure like Paul, they probably have like certain ODDs for now, they don't operate fully autonomous.
They might have remote doors or even still safety drivers behind the wheel. So, I don't know what exactly is, but right now. All of the level for autonomous startups, they have a remote operator and some of them they have a one-to-one remote operator. some of those that they have a better.
Engineering structure. They have embedded like a one to end remote operator door that one remote operator can manage multiple autonomous vehicles. And that was like one of the very important aspects that it helped autonomous vehicle ecosystem to scale. Because I remember until 2020 3, 24, still, they were like a one-to-one remote operator.
Now we can see they are getting scaled. Because if you had like a remote operator still like paying the same amount of money as an Uber driver to sit there and just watch the car driving around. So, when they made this one to end remote operation, they could reduce some of that operational cost.
That was very important. So, I think level four still have that remote operator engage. And the limited ODD and level five means the systems can make decisions at any situation.
Jim Barrood: How far are we from L5 How many years?
Ali Arab: Pretty far. Okay. If I want to give you a very good example that I always give in my talks.
If you look at automotive industry from the day that the first car was mass produced in Detroit area by Ford, it took almost a century that 80% of Americans had access to personal cars. One century. So, if we say from the day that the first autonomous vehicle was serving publicly, probably, I don’t know, five years ago by Waymo or earlier or later around that time.
I'm not saying it will be one century, but learning from history, it will take a while. So, at least couple of decades.
Jim Barrood: Okay. Paul, tell us what the vision is for bot auto and explain. Do you have oversight now and. How does that work?
Paul Lam: Yeah, So, our vision is to be an L four autonomous trucking company and basically haul freight from point A to point B without the human driver involved.
So, if you watch online, we have already achieved our human list drive from Houston. So, no drivers, no human at a backseat. We do have remote monitoring, but a remote monitor and a remote controller do not control the truck for safety. It only comes in. If there is any efficiency problem, for example, the truck safely goes by the road and requires some health check and stuff like that because it's really important.
We need to remember in the states, unlike many other countries, unfortunately, we don't have a hundred percent internet coverage, So, you cannot use the remote control to guarantee safety. The truck itself has to be able to fail safe, and that is the basic requirement. Remember, we're a trucking company. We have nowhere to hide, right?
It's not like the case where in aviation, if the plane doesn't work, you blame the airline. The airline blame, the software engineer and the software engineers. I don't even know where this comes from, right? So, we own the results, and this is really important. So, this is why we have the highest bar among industry because we do good it.
So, that's our vision to really be serving the logistics sectors, removing the drivers So, that we get to the efficiencies. We're available 24 7 and we have to be safe Now in terms of our vision, like going forward I agree with Ali very much. I think it would take a long time for us to get to L five but I dare to challenge every audience here do we really need L five?
It's like human life. Is it a problem that a person could only be one occupation at a time, or is it not? Do we really want to be a generalist on everything? Do I or do a business really require L five to be commercialized? So, think about a product definition as a logistics provider, as a truck capacity provider.
Is my shipper comfortable that I will be able to drive the truck anywhere, anytime anyway? I like, probably not. If today you are shipping whatever important documents of yourself to FedEx, you probably want to make sure, and when you look at the tracker, you want to make sure the FedEx is going in the right direction.
So, that's a defined ODD, not from a technology perspective, but from a business perspective. You just only want to go the shortest route you want to go on the highway. So, does the truck need to be able to drive in all So,rt of scenarios and all sort of situations without definition? I doubt it. And that is not a technology response.
It's more from a product and a commercial response that why do we need a truck that is L five ultimately? And that's the question. So, I. I have a lot of faith in humanity and I think technology breakthroughs is a doctrine of humanity advancement, and I think human is capable of making of those breakthroughs.
But I think L five is not going to come too soon. It's because there isn't such a big demand. For us to do that. Robo Taxi is running great with that specific ODD, knowing going from where to where doing L five is not necessarily really building a lot of value for the customer. So, I'm really coming in from that perspective.
Jim Barrood: That's a great point. And this has been a great conversation. We usually do just one things questions for our guests. So, we have a lot of entrepreneurs in our audience. What's one tip for entrepreneurs who are looking to start an autonomous vehicle company or related venture? Ali?
Ali Arab: My opinion is start with the customer need as Paul was mentioning.
And I, I just want to add one more thing to what Paul said we definitely don't need the level five. So, there are a lot of. Specific needs that it can be automated with mobility and transportation. So, if you can find, like those specific needs is always easier to start if you want to build a full stack solution, similar to what Paul was saying, but to just add one more item, like level five five, from my opinion. And the same thing that Paul mentioned is not going to be the same way that is defined right now. So, it'll be when these level fours or scale to the mass population. So, like most of the people have access exactly what Paul mentioned with limited operational.
We don't need like a full autonomy for every single vehicle. And one more thing, health. Mental and physical financial support is very important for entrepreneurial. Great. Never that
Paul Lam: great Paul. Yeah. At a professional level understand the pain point, understand your customers. That's really the strike.
I think you want to turn the game into a finite game and constrained maximizations. So, whenever somebody said, Hey, I want to do a startup in autonomy. You haven't thought through this yet. So, it's the same comment that I had. If all you have in mind is L five autonomy, you haven't defined the problem yet.
You have only defined the Solutions, right? It's autonomous, great autonomous for what, but as soon as you try to define the pain point and what customers you want to serve, what naturally you'll come to L four and I think. That's really important is that to find out the beachhead market, find out the pain point that is really painful for your customers and use it to crack through this very big journey.
And at a personal level, I think one thing that will be very useful is to understand capital is a responsibility. And I think especially over the last 10 to 15 years. In the eyes of a venture capitalist, they're playing the game of a law of large number, right? So, that works. But don't forget, as an entrepreneur, you're one of those numbers, right?
So, if you fail, that's a hundred percent for you. That's 1% for the vc. So, it's really important to understand. Every entrepreneur has to be responsible for the decisions that they have made, and capital is not something to be celebrated. Capital is something. As a duty of care and it's a responsibility. So, if there is a way that you might be able to bootstrap the company, you can take non-dilutive funding.
That is great because there is certain work that you want to do that ultimately you do the right thing without being. I hate to use the word like kidnapped by the J curve of the revenue growth or the capital return, because as soon as that happens, you are on a ticking clock to commercialize the product.
So, really think about capital very carefully. And capital is a duty of care, is a responsibility. It's not something that you want to be bragging about in TechCrunch the moment when you do that. This would go downhill for you very fast and quickly.
Jim Barrood: Got it. Thank you for that. Ali, what about you?
I know you've been raising money. One tip for entrepreneurs,
Ali Arab: We haven't raised any private money, So, I think that was a very good vision from Paul because we tried to push in the past two, three months, you saw me like presenting and pitching. And it was the first time for me reaching out to private investors.
So, it was very, interesting learning curve. So, it is very different than like just non-dilutive and government type of support that you are looking for. So, it's a completely different angle. It's great to hear this from Paul here right now. We don't want to push the private investment unless it's a hundred percent aligned with what our vision is to build safer autonomy.
If someone is interested to partner with us and invest in us to help them to build safer autonomy, we definitely would take that. But right now we are not like just despite looking for fun to make sure we will be able to deliver what the vision we had initially.
Jim Barrood: Got it. Okay. Now we're on to lightning questions.
What's one belief you had early on that has totally changed? Paul?
Paul Lam: So, many things actually changed for me, but one that has changed is I thought everything can be done with ai. That changes because. When we do the autonomy, we realize our customers will actually send us a check.
Jim Barrood: Got it. Now what about a TV character or movie character that most resembles, the founder style at your company?
Paul Lam: There's a love hate relationship. The more assimilate you are to somebody you also, love, but you also, hate, I feel like Elam Musk really resemble who this team is, like really doing everything yourself out from first principle, try not to outsource too many things.
Jim Barrood: Got it. What about your favorite productivity app?
Paul Lam: Right now, I hate to say it, but chat BT really takes two, two hours of my time every day I talk to chat with more than my wife.
Jim Barrood: Got it. All right. Ali, back to you. What's one belief you had that's totally changed?
Ali Arab: I was, I remember I was doing the National iCorp, which is a program by NSF and.
I remember I put 300% of my effort there and they gave me the trust, the process award, and that trust is changed for me. I don't think there is a. That process. So, the process itself, it changes. So, I don't think there is a process that you can trust and say okay, this is a process that is, if I go step by step, I will become a successful entrepreneur.
So, I don't think such a process exists. So, that was something that changed in my mind.
Jim Barrood: That's great. Interesting TV character, movie character that resembles your founder style.
Ali Arab: It might be a little bit funny, but as a teenager growing up, I loved Jim Carrey. He was very funny and when I learned more from him that like he wrote himself a check a year before he even had a carrier and then he cashed it out when he got the first like actor job.
So, that was very interesting that he trusted in himself. It was inspirational, but not always work.
Jim Barrood: Got it. Alright. What about your favorite productivity tool or app
Ali Arab: think LMS are right now and in general, like foundational models are very helpful as a person that English was not the first language and I had a very difficult learning curve and I struggle to write my PhD thesis.
So, right now, Chad G is making life easier So, I can compete with native English speakers or even better.
Jim Barrood: Got it. Alright, wonderful. This has been really great. We usually end with a poem or a saying or a quote. Paul, what do you have for us?
Paul Lam: Water can make potato hard, but it can also, make something else soft.
So, it really depends on. How you use it. And this goes back to life, to capital, to building a startup. A lot of things come in pairs and they sound like an oxymoron. There's no one way that fits all right? You have to be hardworking, but you all still need to take care of yourself. So, figure out your own combination.
There's no one set of rules.
Jim Barrood: Got it. Thank you…Ali?
Ali Arab: This is something that is not mine, but, during the recent discussion that I had with one of mutual friends that works in autonomous driving for years now, he said Entrepreneurship is an art. And you have to be an artist to make a entrepreneurial trajectory successful.
And this code that I'm giving you, it's from Marcus Weldon, the ex-president of Bell Labs. He gave in one of his talks about the relation of art and technology. I'm reading it to make sure it's correct. He says the art challenges technology and technology inspires the art. And it's originally by John Lester, who was the producer of the animation cars.
And I watched that movie multiple times. It gave me a lot of interesting challenges that was found at that time. So, art basically challenges technology. It's very interesting. So, they give us ideas to go work on it.
Jim Barrood: That sounds really interesting. That's really thoughtful. Alright, this has been a great conversation folks. Good luck with your ventures and we look forward to seeing more autonomous vehicles on the road.
