Eventbrite CEO Julia Hartz on Live Events, AI in Logistics & Healthcare, AI Euphoria | Sep 11, 2025

0:13Welcome, everyone, to the Informations TITV. My name is Akash Pasricha. It is Thursday, and today on the show, we have got a busy lineup for you. Julia Hartz, the CEO of Eventbrite, is coming on the show. She is going to be speaking with our editor-in-chief, Jessica Lesson. We've also got the CEO of Hinge Health. That is, of course, the health tech company that went public earlier this year. It is making a big play into the wearable space, and I am excited for that conversation. We're then bringing on the CEO of Augment. He is also the entrepreneur that previously sold his business to Shopify for$2.5 billion.

0:49And we are also going to start things off with a discussion with our senior finance editor who wrote a column last night about how he thinks we could be at the peak of AI euphoria, especially in the stock market. I'm excited for that conversation. But before we get there, I want to highlight for you an op-ed that Vinod Khosla published in the information yesterday. The headline is a question, are AI valuations bonkers? It really is a great piece because it gives us a detailed history of how exactly we got to this point where valuations have gotten so high and so rich. And yet, Vinod makes the point that the lower valuations will not necessarily be the ones that will be good investments.

1:32He also explains the story from the point of view of all the limited partners and all these venture funds, what metrics they are looking at. and how he sees venture capital as a whole changing over the next 10 years. It is a fascinating piece. I will link it in the show notes. And with that, I want to get to our first guest. Well, the markets have been a bit crazy lately, to say the least. Oracle shares jumping yesterday were just the latest data point to that effect. IPOs have also been making big swings lately. And last night, the Information Senior Finance Editor, Ken Brown, put out a column that said that he thinks we could be at peak AI euphoria.

2:11And I want to bring on Ken to talk a bit more about what he meant. Ken, welcome to TITV. It's great to have you here. Hi, Akash. Okay, so lay it out for us. Are you calling the top or what was the point you were trying to get across? Oh, you know, this one is going to come back to haunt me for years to come, right? No, look, it was just watching Oracle, the shares go up 36%, creating hundreds of billions of dollars in value because they said that they're going to have a whole bunch of revenue years down the road. And, you know, so they're not this is not in their earnings. This is not in any official document.

2:46This is years down the road. And then when you think about it, where's that revenue going to come from? It's going to come from businesses that are losing tons of money that need to raise tons of money. So it's crazy from the start. And then the deeper you look into it, It gets crazier. And so to see this thing go up the way it did, you know, it just made me react. And that's what I wrote. Now, it wasn't just Oracle, though. I mean, you were also pointing out that, I mean, the energy sector was also getting all excited about this. Yeah, yeah. So like a whole bunch of stocks around data centers were up yesterday.

3:20Companies like Constellation Energy and Vistra and GE Vinova, which is, you know, that makes gas power generators, among other things. They were some of the top performers in the market yesterday. Again, it strains credibility to see that a company like G.E. Vernovers have done great because they make these gas turbines that can power data centers, but they have a backlog up until 2028, 2029. So it doesn't quite make sense because Oracle says they're going to be making all this revenue in 2030. there may not be enough power around to get these data centers off the ground. Okay. So I want to dive a little bit more into the power piece of this, but I guess one thing, broadly speaking, is what I've been thinking about is, okay, if we are at peak AI euphoria, what is the catalyst that starts to sort of pop the bubble?

4:13And this is, look, I mean, people have been studying this question for decades now. What causes bubbles to pop? I mean, the energy question is kind of interesting because one thing you and I were talking about was if they can't get the energy to build these data centers, I mean, all these projects that OpenAI is investing in, I mean, they could be for naught, really. Well, and, you know, some of it just gets at the speed, right? So, you know, the AI is moving at a million miles an hour, and the energy industry is moving at five miles an hour. And so that's just a problem. And so, you know, they'll build the power over time, but not as fast as, you know, the tech world moves.

4:52And so, yeah, that could be an issue. I mean, what makes things pop? You know, it's always an accumulated bunch of factors that make things pop. And there's a huge amount of tailwinds that are going to keep pushing it forward. But, you know, all I was pointing out was, you know, there were these transactions and these reactions that just don't make any sense. Right. And, you know, things can not make sense for a long time. We've seen that for years. And so So it's going to be an accumulated group of factors, maybe energy, maybe some companies not being able to raise the money they say. I mean, OpenAI is going to burn$100 billion.

5:32And so maybe they can't do that anymore. And then that slows things down. And it's going to be just a bunch of things and people are going to start to like sober up a little bit. Well, what about all these renewable energy sources that we hear about? And, you know, we hear of even longer off bets, like we've had nuclear companies on the show, for example. I mean, do you get any sense that these newer emerging technologies could at all satisfy the AI demand? Well, look, 90 percent of the new U.S. power generation last year and this year was solar and wind. That has been solar especially is the cheapest form of energy and the quickest to deploy.

6:07And when you pair it up with batteries, it can be effective for running data centers. So that but the problem is the Trump administration has been pushing back against renewables and putting tariffs on the gear. And so it's hard. And then, you know, the gas generation is just there's plenty of gas, but the generators, the turbines are they got to be built. So the renewable energy would be the quick way to do it, and people are using it. But it has headwinds that it didn't have last year. And is there also sort of a debt component to this story? Because one of the comments that we had an analyst yesterday on the show, he was talking about how Oracle finances buildings capacity for its cloud business.

7:00And he pointed out, he said, look, the company's burning cash right now, so they might look to raise debt. How should we be thinking about the debt side of the story? What should we be watching for? Yeah, yeah. You've heard me rant about this, right? Like, you know, the tech world usually doesn't use a lot of debt, right? It uses equity and VC money and all that. And this is a capital intensive business AI. And so there's a lot of debt. And the problem with debt is you got to pay it back. And so there's strict deadlines. And it's, you know, if you're not generating cash, you know, you can keep borrowing, but no one's going to keep lending you money if you're not paying it back.

7:37And if there's no sign, you're going to pay it back. And so that can cut you off. And so Oracle is going to have to do that. a bunch of these other data center operatives, CoreWeave is buried in debt. It may all work out, right? They may grow fast enough. They may generate cash and they may be able to pay it back, but that's a real risk because that gets at the timing of this, right? You don't have forever because you have to pay back the debt. Right, right. Okay. Well, yeah, once time comes into the picture, right now, there are no deadlines really. It's just, you could put out five-year projections like it's nothing, but you are right that once debt comes into the picture, then somebody is coming to you and saying, hey, where is it?

8:15Where's the money? So we've all had car loans and credit card loans and all that. And it's the same for big companies. Right, right. Well, big cars. All right. Well, Ken, thank you so much for coming on the show and talking about it. That is Ken Brown, our senior finance editor at The Information. Okay, let's get to our next guest. Well, Hinge Health is newly navigating life as a public company. The company debuted its shares on the stock market earlier this year. and it is going pretty well. Shares are up about 40%. The company has a digital platform that helps people with musculoskeletal disorders, and it is also moving into the wearable space.

8:50The company is now worth more than$4 billion. And I want to bring on CEO Dan Perez to talk about what he is building. Dan, welcome to TITV. It's great to have you here. Thanks for having me, Akash. So for people who aren't familiar, just talk to us about what the disorder is that Hinge Health is trying to solve and what you've built as a solution. Well, great question. So look, overall, our aim is to use technology to scale and automate the delivery of care. And if you think about healthcare overall, despite all the technology, all the money that goes into healthcare, it's a multi-trillion dollar industry in the US, it is remarkably manual.

9:25It is very much a services industry. When people complain about healthcare costs, when people talk about how it's a relentless drive up, it's because it really hasn't been automated. We're tackling orthopedics, essentially, back, joint, and muscle pain in healthcare. And for most employers in health plans, it's the number one or number two largest cost driver. Think about knee replacements, back surgeries, knee arthroscopies, MRIs, twisted knees. These things add up, and it's about one in six healthcare dollars for our customers is spent on orthopedic care. And we're using technology to focus on physical therapy initially.

10:03We've automated away about 95 % of human clinician hours associated with PT. And we want to keep marching in healthcare, continue to use technology to automate more aspects of healthcare. And so your platform, is it like pre-recorded content that you sort of package together and personalized plans? Or are people actually talking to a physical therapist digitally through telemedicine? Great question. So we have multiple options. So every single person who we onboard into our program has their own care team. So we firmly believe in having, you know, human powered care teams as well. We also have their own physical therapist.

10:37We also have health coaches. We have access to doctors. We have access to nurses. And our technology, however, is able to automate most of the interactions as they go through their program. And that makes it a lot more convenient for the member. That makes it much lower cost. Like automating interaction, what does that mean? Well, when you come onto the program, we're able to create a custom care plan for you. And you're able to do your exercise therapy, whether it's at 2 a.m. in the morning or 2 p.m. in the day. You don't have to worry about co-pays. You don't have to worry about setting up an appointment.

11:05If at any time you want to speak to a physical therapist, you could speak to it, a physical therapist. But most of our members could just go right through using just our software as they guide them through their exercise therapy or the physical therapy right from home. And we use computer vision. So we use the front-facing camera on your phone to track 100 unique landmarks on your body. Okay, so they set up the camera, and then the camera kind of studies how you're doing the movement. And then it coaches you saying, well, lift your arm up that way or this way. Exactly. We'll take you through your rehab for your back pain, for your elbow pain, for your knee pain.

11:39And we know exactly what your range of motion was. We could track every rep, every exercise, and give you live feedback. We have higher fidelity than what you will see in an in-person physical therapy clinic in terms of the data we're able to collect. We're able to help you track your progress over time. And so the number of exercise therapy sessions that people do on Hinge Health in a year is well over 30. and I think it'd be very hard pressed for the average person who starts in-person physical therapy to do 35 plus in-person physical therapy sessions. So our experience is much higher. I'm gonna jump in here because one of the most fascinating parts of your business is that you are selling to health plans and to employers.

12:20And one of the questions I wanted to ask you is, we've seen this AI boom and a lot of healthcare companies are trying to take advantage of that. As you're talking to health plans, How are they thinking about AI? And has that changed the nature of your conversation or your pitch to them? Or are they saying, you know, we don't really know about this? Well, it's a great point. Look, most of us expect our health insurance to kick in for health care. Even with a$150,$200 PT appointment, we expect maybe we'll pay a$25 copay. But we expect health insurance to kick in. And so, you know, as a technology company playing in health care, we have to sell to enterprise customers, employers, and health insurance companies.

13:02And they are cautious around AI. You know, technology is moving a lot faster than the regulations are moving. And health plans, health insurance companies, and employers tend to be more cautious. You know, and rightfully so. So they're not buying it. When you say, hey, we've got AI-powered solution, they're not immediately. No, they are buying it. They want to learn more. It's not like, you know, if you're building an AI tool in design or you're building a coding software tool in AI, you could launch a bunch of features. Some might stick, some might not. In healthcare, the stakes are a little bit higher.

13:35And so you want to understand, well, what is your approach to AI? How does your AI work? And there's going to be more questions. And if you're building an AI company or an AI company in healthcare and you're trying to sell to enterprise customers, there's going to be a higher performance barrier for you to demonstrate the safety of your product. We've been very lucky to do that. Our computer vision, The computer vision is actually a subfield of AI. And we've been working on this well before even ChatGPT was launched. So we knew that AI was coming to healthcare, and it needs to come to healthcare.

14:04Because if you do not automate healthcare, we will not lower costs in healthcare. So are they asking for more data than that? If the bar is higher, are they asking for more evidence that AI is not? Or whether, and AI is one type of technology. But yeah, any way that your technology is delivering care, they want to better understand if a human is not in the loop or if the AI or the technology is making decisions on behalf of a patient's care, what have you done to safeguard this? What are you doing? What sort of oversight is there? What is your track record? Now, we're very lucky that we have over now a million and a half people who've been treated by our program.

14:48So we have a very, very good track record. If you are building an AI company in healthcare that's going to actually treat patients, not just some back office software that's just automating operations, you should better understand how might this go wrong, right? How would you ensure that people are going to be safe because the stakes are simply higher in healthcare? So you've also got this wearables device, which is so fascinating to me. I kind of want to ask you about the decision to do something in wearables because, look, your public company, your margins are great. They're like software company margins right now.

15:23Wearables is a really hard business. And so why do you need to do wearables at all? Well, look, we are committed to automating healthcare delivery itself, right? There's a lot of technology in healthcare, usually on the periphery. Most healthcare delivery is still one-on-one with the provider. I firmly believe that software will soon automate all non-touch aspects of healthcare. Think about like, you know. So AI is replacing non-touch doctors. That's what you're saying. Oh, cognitive aspects of healthcare. Any aspect of healthcare that involves, you know, just thinking about it. Look, here's your symptoms and I'm interpreting your symptoms.

15:57That's pattern recognition. That is an LLM could do that very well. I'm evaluating your labs and trying to interpret your labs and what this means and what might be the implications of your labs. That's cognitive. That could absolutely be done by AI. But there's a lot of touch aspects of healthcare that still require moving things about or connected hardware. And if you're committed to automating healthcare, you're going to have to think about connected hardware as well. So you is actually our connected hardware. It's about the size of a half dollar. It's a pain management device. Yeah, I see.

16:30You can put it, you stick it on your neck or your shoulder, whatever. Exactly. And it delivers electrical nerve stimulation directly through the body. Gives our patients a non-surgical, non-invasive, non-addictive way for their pain relief. We have about 15 patents. We're going to be filing five or so more. It's FDA approved as well. And it's FDA approved as well. And FDA cleared, which we're really excited about. And we've invested quite a bit into this device, and we're investing in more hardware. And I think if you're a healthcare company, and particularly if you're unlocking enterprise reimbursements, you could still have robust margins so long as your outcomes are really good, because the price points in healthcare are actually a lot higher than price points in other areas of the industry.

17:12I just want to go back to, so software will automate all non-touch aspects of healthcare. It is. Absolutely. Right. So that's sort of the rationale for, you know, building a wearables device. But I guess my question is, so what do you see the role of the doctor then going forward if software is doing some of that cognitive work that you say they do right now? It's a great question. If I was a pre-med student and undergrad right now considering a career in medicine, I would think the advice I'm getting from the old guard or from 20-year veterans would be somewhat stale advice in many ways because the world is changing very, very fast.

17:57So what's your advice then? What do you advise them? Well, to be eyes wide open, the world is moving towards automation. The AI, Hinge Health and digital health companies, we are not happening to healthcare. The future is happening to healthcare, and it's happening a lot faster than it's ever happened before. Healthcare in the past has been somewhat ignored from technology's automation. Technology's automated and improved and streamlined a lot of other industries and has not really touched healthcare. And now healthcare is coming for healthcare, and AI is coming for healthcare. And I think patients will soon be demanding more AI because AI is showing that it can often be a little bit more accurate than human doctors in terms of interpreting symptoms.

18:34And so I think doctors, their roles will change. And where they're adding most value will change. And I got to let you go, but riff with me here for a second. Let's think. So what is the doctor patient experience? It'll change, but how? What does the future of the doctor visit look like? We're going to see progressively more automation in healthcare. I don't even go to the doctor's office. I think you'll still go to the doctor's office. It's going to be really, really hard to automate a lot of physical aspects of healthcare, these unstructured physical tasks. Imagine something as low risk as an ingrown toenail.

19:15We do not have a robot that could do that. Your family medicine doctor, an internal medicine doctor, or a pediatrician is still going to have to do that sort of procedure. There are plenty of procedures. You know, we don't have a robot that could draw blood anymore, you know, or at all yet. And so you have a technician in your clinic who's going to have to do that. So there's many touch aspects of healthcare. It's going to take us many, many years to automate. But look, LASIK is a surgery. I think that's 95 % delivered by a robot today. And it has to be a surgeon, I guess. That's what I mean. Be a surgeon.

19:45Robotic surgeries, robotic surgeries are coming for us right now. So then specialize in getting the ingrown toenails. So maybe that's, you know, I don't know. So those aren't the most glamorous parts of healthcare, but I was not being automated. I don't have a crystal ball, but I know it's changing. I still firmly believe there is a role for human providers in healthcare for a long time, working with technology. I think doctors who don't understand AI and are not embracing technology, I think, will struggle in the future. Doctors who do embrace technology, who do learn how to embed technology into their day-to-day practice, will be augmented and will become some of the best clinicians.

20:22and I think that's true of designers. If you're a designer and a really good designer, you use AI, you will become a great designer. If you're a designer, not so good of a designer and don't embrace AI, you will be rendered obsolete. So that's going to be the case, not just for doctors, but for design, for eng, for product management, for sales roles. It's going to change all of our jobs. It's exciting. Dan, thank you so much for coming on the show. It's a fascinating business and I do want to talk to you more about the wearables business next time you come on because I think there's a lot to talk about where that business could go and what percent of your business it could become.

20:57But that is for another discussion. That is Dan Perez, the CEO of Hinge Health, his first time on TITV. Okay. Our next guest sold his e-commerce fulfillment company to Shopify for$2.5 billion. It was a big exit, but that was three years ago. Now, Harish Abbott is back in the founder seat building an AI logistics company called Augment. The company He has raised$110 million in five months from big name capitalists like Redpoint Ventures and 8BC. I want to bring on Harish to talk about where he wants to take this company. Harish, welcome to the show. It's great to have you. Thank you for having me.

21:34It's great to be here. You know, I was listening to your previous segment. I think I need one of those devices. I just fed my back this morning. Oh, my gosh. Well, what exercises are you doing? Give us the secrets. I think it was, I was doing some bench presses. and I think I just, I did not position myself well. Oh man, it's, you know, you got to keep the feet planted on the floor. You know, that's the key. What is it? Arch, keep the back arch? I don't know. I gotta ask my brother. He's the lifter in the family. Okay, but let's talk about another big lift, which is the business of logistics. It's a meaty topic.

22:10Before we get into that, I do want to ask you, you sold your company for$2.5 billion to Shopify, okay? why are you doing them? You don't need to do this. Why are you doing this now? I think like, you know, I love building, you know, and you kind of like do things that enjoy, that you enjoy most. And this is what I enjoy most, which is building. I think two is AI is probably the most transformative technology I've seen in my lives. And I was like, this is like one time we can marry with an industry that has been so resistant to technology. It could be even more than 2.5 billion. It could be 3.5.

22:45Who knows? I mean, I think it is like less about money is not going to change the lifestyle. Right, right. It is more about like, can we have a meaningful impact here? Right. That badly means it is so vital. It's just the joy of building, you know? So, okay. So you talked about AI being an opportunity and I'm with you, but a lot of times when we talk to BCs, they say, hey, it's not about the product. It's about the problem. And so what I want to know from you is What is the problem that got you so energized to come back and be a founder? It's not easy. It was, you know, with any amount of money, what is the problem that you're addicted to right now?

23:23Yeah. See, logistics is very large and it's also extremely fragmented, right? Like in the US alone, there is 1 million truck driving companies. There are maybe tens of thousands of brokers and freight. There are hundreds of thousands of shippers. It's an extremely fragmented industry. Right. multiple people have to touch something multiple companies for like that coffee mug that you have on your table to get where it is at maybe 15 to 20 companies right so but when a company when an industry is so fragmented but in order to move goods they have to trade information yeah but the industry is so fragmented that the only way that is common among all these people is emails and phone calls and text.

24:07Okay. All of those methods are both asynchronous, but they're also bandwidth. And this is like, hey, what's the update on where this freight, where this cargo is, essentially? Now I can plan my labor. I can schedule my appointments. Then I can schedule my next thing that needs to go out of this warehouse. Everybody needs to know the updates. Everybody needs to know where a truck is at, where the inventory is at, how much inventory is coming. is a container laid on a port, you can plan the next thing. And imagine there's a chain of 20 companies that have to coordinate for that little water plastic you have on your table.

24:48And one of them is late and there's a whip-saw effect. And so what that does, it creates lots of inefficiency, lots of waste, because I have my labor plan, but I don't know if that container is going to come on time. Right. And this is massive industry. So when you multiply that across all the participants, this industry is very large. So that's what got me excited. You know, like, could we bring this waste down? Could we make this industry more efficient than the status quo? Okay. So that's the problem. Now, you know, your business and the product, I think the first product is called Augie is the name of the product, short for Augie.