Episode #78: Three Breakthroughs: 2024 Election - Polling Is Dead, Prediction Markets Are In

Frequently Asked Questions

FAQ

• Where is the transcript for this webinar?

  Sam:
  Doors are open at your right. Watch your step as you approach the platform into the land of tomorrow. Welcome to the Tech Optimist, an Alumni Ventures podcast. Enjoy your stay.

  Naren Ramaswamy:
  When I saw this, I was like, “This is absurd.”

  Sam:
  That is Naren Ramaswamy, senior principal at Alumni Ventures.

  Mike Collins:
  I think there are going to be two types of CEOs. There’s going to be AI-amplified CEOs, and former.

  Sam:
  You know him, you love him—that is Mike Collins, founder and CEO of Alumni Ventures. And that’s me. My name is Samantha Herrick, and I am the guide for this podcast. Welcome aboard.

  All right, welcome back to the show, everyone. How are we doing? We all hope you are having a wonderful day so far, night, morning, wherever you find yourself while you’re listening to all of us here at Alumni Ventures.

  Today, on the Tech Optimist in our Three Breakthroughs episode, we’ve got some really interesting breakthroughs today. Again, before we hop into this episode, we’re going to take a few seconds for an AV ad and then a disclaimer, and then we’ll hop right into the rest of the show. Enjoy.

  AV Ad:
  Do you have a venture capital portfolio of cutting-edge startups? Without one, you could be missing out on enormous value creation and a more diversified personal portfolio. Alumni Ventures, ranked a top-twenty VC firm by CB Insights, is the leading VC firm for individual investors. Believe in investing in innovation? Visit av.vc/foundation to get started.

  Sam:
  As a reminder, the Tech Optimist podcast is for informational purposes only. It is not personalized advice and it is not an offer to buy or sell securities. For additional important details, please see the text description accompanying this episode.

  Mike Collins:
  Hello and welcome to this week’s episode of Tech Optimist. I’m Mike Collins, I’m the founder and CEO of Alumni Ventures. I’m here with Naren Ramaswamy, who’s one of our teammates here, one of our great young VCs. And we’re here to talk about this week in technology and innovation in venture capital—and what’s on our mind. What are big announcements? What are big breakthroughs?

  And there was an election last night. And so, I think we’re going to talk about that ’cause that impacts our ecosystem for sure in a big way. And without taking the political angle, our job as VCs is to hit the ball where it lies. And we have a Republican government for the next four years. And so, I think we had done good scenario planning here at Alumni Ventures, and we’re kind of prepared either way it was going to go.

  I think there are some takeaways—I think we can expect generally good things for the venture ecosystem from an M&A perspective. I think there’s likely to be a lot more activity on that front, which tends to be good to open up liquidity vectors for companies.

  I think there are certain segments that are going to have more runway and freedom to do things, such as AI and blockchain. The markets have ripped—frankly, crypto, public equities—and I think it’s really this sentiment, which may or may not prove to be true, that there is going to be more liberal perspectives when it comes to M&A and probably a loosening of some of the kind of regulatory headwinds that the industry, as perceived, has experienced.

  All that is true, and we’ll probably dig into all of that stuff more. It’s still all pretty fresh, I think, for a lot of us. But I will tell you—here is my takeaway, my point number one: I think polling is dead and prediction markets are in. And I was frankly a little skeptical of this. I said polls are not perfect, but they’re directionally pretty close. And prediction markets are probably biased because it’s all men, and are people really making an investment decision? Or is it just like betting on their home sport team?

  I think I’ve probably flipped on this, which is I think that they’re now big and deep enough that they are providing real information.

  Sam:
  Okay, so now what we’re going to do before Mike and Naren really get into this conversation about how polling is dead and prediction markets and what they are, I’m going to sort of explain what prediction markets are compared to traditional polling.

  So they’re both two distinct methods for forecasting election outcomes.

  Real-time responsiveness: Prediction markets can react swiftly to new information or events, providing real-time updates to forecasts. For example, when Hillary Clinton’s email leak occurred in 2016, betting odds quickly adjusted to reflect this new information. In contrast, polls take time to collect and analyze the data, which can result in a lag between events and their reflection in polling numbers.

  In prediction markets, participants are putting their own money at stake, which can be seen as a more genuine reflection of their beliefs about the outcome. Polls, on the other hand, do not involve any financial risk for respondents. Polls typically draw from a broader demographic, potentially offering a more representative view of the general public sentiment. Prediction markets are limited to those willing and able to place bets, which may not be as representative of the overall population.

  Polls can focus on specific regions or demographics, providing a more nuanced understanding of the election landscape, whereas prediction markets tend to offer a more aggregate view, although some do allow betting on state-level outcomes.

  Accuracy and reliability: The debate over which method is more accurate is still ongoing. Some researchers suggest that prediction markets are at least as accurate as polls and quicker to incorporate new information. However, others argue that polls provide more accurate snapshots until very close to the election.

  There are potential issues with prediction markets as far as regulation and manipulation. Prediction markets are susceptible to manipulation, especially when participation is limited. Large bets from individual actors can significantly shift the odds. Polls, while not immune to manipulation, are generally less susceptible to this issue. Prediction markets also face regulatory challenges in many jurisdictions, including the US, which can limit participation and potentially affect their accuracy.

  Complementary approaches: Many experts suggest that both prediction markets and polls provide valuable insights and can be used in conjunction to improve election forecasting. Some models incorporate data from both sources along with economic indicators to create more comprehensive forecasts.

  In conclusion, while prediction markets offer advantages in terms of real-time responsiveness and financial commitment, polls provide broader demographic representation and regional specificity. Both methods have their strengths and limitations, and their relative accuracy continues to be subject to debate and research in the field of election forecasting.

  Mike Collins:
  I think that’s a big deal. I think people vote in their self-interest, they vote with dollars, and if the market is liquid enough and deep enough and there’s enough good information, I think there is enough data now that prediction markets, which are an innovation…

  And I was actually listening to an interview with the founder of one of the big ones. Kalshi—is that the name of it?

  Naren Ramaswamy:
  Kalshi. Yeah, it’s actually a Sequoia-backed company. Friends at Sequoia had sort of seen this a few years ago and said, this is going to be like polling. This is just one example of it, but prediction markets are, like you said, voting with money rather than just hearsay. This is another example of technology underpinning what has been a speculative, hearsay, laborious manual job of a pollster, which is interesting to see.

  Mike Collins:
  Yeah, I think there’s a couple of other takeaways too, which you can Google it, but there’s actually a pretty good interview with the founder that took place last week. And literally their job for three years was to get this thing approved from regulators, right?

  Naren Ramaswamy:
  Yeah.

  Mike Collins:
  So I think there’s the layer of this, which is, oh, prediction markets did better than polls. I think he articulated and brought me around a lot to the idea that these prediction markets are just the form of what goes on in the commodities market, right?

  Naren Ramaswamy:
  Exactly.

  Mike Collins:
  And people are betting on weather, people are betting on government supply reports. And the difference between kind of gambling, and hedging, and where’s the role in government there? There’s a role, but I do think—again, putting on my opinion hat—I do think people should have the freedom, with complete information and full disclosure, to do with their money what they want within reason.

  So I thought he created a really compelling argument about his business and the value that it served. And yes, I mean, here’s a perfect example: if I was a venture capital firm that was invested in some particular segment of the economy where I could be looking at different outcomes depending on how an election went, should I not be able to hedge my position for the benefit of other stakeholders in that? I thought it was a pretty damn compelling argument.

  So again, I would encourage people to check out that interview. Not an AV portfolio company, but I think the whole idea of, again, technology and innovation—here’s an entrepreneur who’s been through huge ups and downs, grindy, grindy to get this thing legal and approved, and is definitely having their day in the sun today, for sure.

  Naren Ramaswamy:
  Yeah, definitely in the limelight now with the election. And I’m just fascinated by how this is another, ultimately, piece of software that has come to the limelight. And to your point about it being another investment vehicle, you can think about some venture investors ultimately taking bets on companies. This is something similar and provides an avenue for folks to do that in an informed manner. So fascinating.

  Mike Collins:
  It just broadens the spectrum of your hedgeable tools, right?

  Naren Ramaswamy:
  Yeah.

  Mike Collins:
  And one of the anecdotes he gave was: okay, if you live in the hurricane or tornado belt and you can’t get insurance anymore, you can, in a way, hedge yourself on weather. So again, still pretty crude.

  Naren Ramaswamy:
  It’s the smart thing to do.

  Mike Collins:
  Not the best way to do it, but it’s like these markets tend to be good. And again, obviously, we need regulation, full disclosure. So I appreciate the fact that he did the pain of going through the process of doing this the right way. And a lot of times, when you have markets and innovation, if you don’t do it the right way, bad things happen and it kind of ruins the whole idea, right?

  Naren Ramaswamy:
  Yeah.

  Mike Collins:
  Kind of thing. So it’s like, again, I think this perception that VCs and entrepreneurs don’t like regulation, I think, is wrong. And I think, generally, what we want and what entrepreneurs want are clear rules and a reasonable process for companies to meet those reasonable rules. That’s all. And that grows it. And that protects people, that grows the market, that allows innovators to be creative and to bring new things to society.

  Naren Ramaswamy:
  I mean, regulation has historically lagged behind technological progress. And here, to see a founder actually put that in front and say, “I’m gonna be methodical and responsible about this,” is great to see.

  Mike Collins:
  100%. Okay, Naren, what do you get?

  Naren Ramaswamy:
  It’s a good segue with the point about regulation, I think, because I wanted to talk about brain-computer interfaces and brain chips. And we have, obviously, the FDA process in place to approve those. But I think this is fascinating because—I mean, we’ve been thinking in our lifetime, maybe we can put a chip in our brains and help understand it—but this is coming way sooner than we think.

  It’s timely because people have been talking about Neuralink, and this is Elon Musk’s venture, but we actually have a V portfolio company called Synchron, which has now finished the first of two major FDA clinical trials. And at the end of that, it’s shown that the device itself was safe.

  What’s cool about this brain-computer interface is that it allows people with paralysis to access technology in a way that they couldn’t before. Last time we talked about voice agents—instead of keyboards, you use your voice—but here is another way to access technology. You can do things like play games, scroll on an iPhone, interact with apps because… And this is not a super invasive surgery. The cool innovation is that it’s a small, minimally invasive surgery that goes into your neck and allows you to interact with technology around you.

  And we live in such a tech-first world that the gap between folks that can access technology and those that can’t is just widening. And here comes this cool technology that—in a matter of years—is going to be in our hands. And this is not just for folks that maybe are paralyzed and not able to access technology, it’s also for others. It just unlocks a different behavior.

  The same way that the iPhone introduced a full-screen touchscreen and multi-touch—that was a new behavior—and brain-computer interfaces are going to provide another new behavior that’s going to enhance the quality of life for, beginning with millions, but then eventually billions of people. This is going to be the future of how we interact with tech.

  Sam:
  All right, to provide a little bit more context on brain-computer interfaces and brain chips that the guys are talking about, I found a video that Penn Medicine released about a month ago called “Thought Controlled Prosthetics: A Brain-Computer Interface Breakthrough.”

  Now, this video is really cool. It’s from some of the researchers from Penn Medicine and it features a few of the people from Precision Neuro, one of the portfolio companies within Alumni Ventures. They actually were on episode 27 of the Tech Optimist, if you want to go back and hear their story—really cool story.

  But this video, along the lines of just brain-computer interfaces, really talks about how their technology is directly implemented into these patients and how it works and what this breakthrough means for the sector of innovation, technology, and medicine. So we’re going to play this video now after we’ve heard a little bit from Mike and Naren. And so, it brings both ends to a point where there are the breakthroughs about some possibilities of what this technology could really do, and then there’s this video which actually provides a real-world example of what it’s doing.

  And so, I think marrying them both together really helps sort of round off this subject. So here’s that video and enjoy.

  Bijan Pesaran:
  The way that we translate our thoughts into actions is absolutely mysterious. There’s no doubt that the motor cortex is involved. There’s also no doubt that it is not able to control your movements on its own. And how those things come together is, in some sense, the heart of free will.

  Iahn Cajigas:
  My area of expertise with functional neurosurgery allows this unique window where we can now research the intricacies of hand and arm function so that, ultimately, we could drive computer interfaces that would help patients with paralysis.

  Speaker 8:
  Is 60… 60 is the lowest?

  Iahn Cajigas:
  Yeah.

  Speaker 8:
  Okay. I’m saying for each, so minimum…

  Iahn Cajigas:
  I had a profound experience while I was growing up. A friend of mine was paralyzed from a diving accident. And so, I’ve always been thinking about this idea of restoring function with devices to give patients some independence back.

  Patients that we’re recruiting for this research are undergoing deep brain stimulation for movement disorders. They have Parkinson’s disease or tremor. We place a long-term electrode deep into the brain for their movement disorder that they’re going to have as their therapy moves forward. And just before we wake them up, we place this electrode on the surface of the brain that’s about two centimeters by two centimeters.

  And then we’re going to start the experiment.

  And it has 1,024 contacts.

  Ben Rapoport:
  It’s hundreds of times higher resolution than the next most readily available electrodes.

  Speaker 8:
  Is Precision ready?

  Ben Rapoport:
  And it provides the world’s highest resolution electrical picture of her motor cortex as she moves.

  Iahn Cajigas:
  So we can see a very, very high-resolution window into the brain. And we’ve created special gloves that reconstruct the movements of the hand. And so, in the OR what we see is this full 3D reconstruction of their actual movement synchronized with the deep activity in the brain and the cortical surface activity. All are a window into how the behavior is being executed.

  Speaker 8:
  All right, so I’m going to ask you to reach out to me three times.

  Bijan Pesaran:
  Each person is different from every other person, and until we study more people, we just don’t know what that range of variation is going to be.

  Iahn Cajigas:
  When you think about moving the hand or actually going to move the hand, different frequency bands carry different types of information related to movement.

  What we get are these really interesting time waveforms. You can see how the activity on the brain surface evolves over time. So it’s not like a musical orchestra where certain frequencies go away, certain frequencies come up. We see interesting patterns of spirals and waves traveling from one direction to the other. One of my colleagues likened it to watching the weather on television. So really, our job is to listen and know what patterns in the music relate to the behavior.

  Bijan Pesaran:
  When we are successful in deploying these kinds of technologies, what it will demonstrate is that we understand the brain and its inner workings. And that’s going to have a foundational impact on who we think we are and how we explain our behavior.

  Ben Rapoport:
  The technology that we’re working on to provide thought-based control of external devices to people who are paralyzed—that has never been possible before. That’s only been really dreamed of.

  Iahn Cajigas:
  And so, I imagine a world where folks with paralysis will be able to interact with things in their home. I imagine a world where we can take somebody who’s an amputee and restore the movement with a prosthetic. But it’s incredible that we’re now at this cusp of new technologies that are going to facilitate us understanding how the brain works and potentially translating that to our patients.

  Mike Collins:
  The human brain—the interface between the human brain and silicon brains—is just a super fascinating and rich investment area. And I know we have several things. A lot of times, the first step is to deal with someone that has some health issue going on—paralysis, a brain disorder, or tremors—something that this technology helps address. But for most of these companies, this is just a first step into narrowing the gap between these biological computers and the communities. And its sensors, human brains, computer brains—more sensors equals value creation, right?

  Naren Ramaswamy:
  Yeah.

  Mike Collins:
  And so it’s a fascinating area. Again, also really cool stuff still going on in the area of CRISPR, which is technology to really get into the brain and snip DNA problems. So that’s kind of repairing software, if you will. So this whole class of areas is just so exciting. And again, over the next five to ten years, I think we’re going to see just another area of life-changing advancements as some of these things come online and give us ways to interact with other forms of intelligence, other sensors, and other good things.

  Naren Ramaswamy:
  It’s just another vector in the whole physical-digital universe that’s going to be created. And at the intersection of biology, hardware, and AI, I’m quite excited to see what’s going to pan out over the next, call it, five or ten years.

  Sam:
  All right, let’s take a break. We’ll be right back. Don’t go anywhere.

  AV Ad:
  Exceptional value creation comes from solving hard things. Alumni Ventures Deep Tech Fund is a portfolio of 20 to 30 ventures run by exceptional teams who are tackling huge opportunities in AI, space, energy, transportation, cybersecurity, and more. These game-changing ventures have strong lead venture investors and practical approaches to creating shareholder value. If you are interested in investing in the future of deep tech, visit av.vc/deeptech to learn more.

  Mike Collins:
  And I think the final one I want to talk about, which is going to sound a little odd, but I do think it’s an interesting story and I’m going to touch on this superficially. People can Google it and Wikipedia it. But basically, the story is there was an AI bot that was created to talk to another AI bot. And out of that came the idea between these two AI bots to create a meme—and a meme coin that’s associated with it.

  Naren Ramaswamy:
  Of course.

  Mike Collins:
  Which then used another bot to really start posting at huge volumes on X and that value was created. And so this was the first, to my knowledge… And then Mark Andreessen gave it… Folks can look up a kind of goat meme—I think, what’s the exact name of it, Naren?

  Naren Ramaswamy:
  Yeah, it’s [inaudible]. Yeah, the token is called goat, G-O-A-T. And that was the first. When I saw this, I was like, this is absurd—

  Mike Collins:
  Are you kidding me?

  Naren Ramaswamy:
  Yeah. But no, go ahead, Mike. I’m curious to hear your perspective after seeing this.

  Mike Collins:
  Well, and then Mark Andreessen got involved to grant them 50 grand to run this experiment. And beyond the meme, goofy, weird dimension of this… And oh, by the way, the market cap of this coin—

  Naren Ramaswamy:
  Hundreds of millions.

  Mike Collins:
  Hundreds of millions, I think now, actually, right?

  Naren Ramaswamy:
  Yeah.

  Mike Collins:
  So again, I just think weird AI experiments are worth watching. I think that’s kind of the thing. This is odd, this is strange. Our job is to not dismiss it. Sometimes it is just that—this is just weirdness. Other times, it’s the idea of AI having discussions with other AI and coming up with things—that is interesting. And this is the first example of this. And there is going to be much, much more of this flywheel perpetuating stuff. Like AI writes code, code gets checked by another AI, gets suggestions by a third AI—that these kinds of loops can exist, I think, is… And this is just a really, really early weird example of it, but I think it points to where the world is heading.

  Now, I will give you a much more practical example of how this works in my life now, which is: I use all the AI tools every day extensively. Different AI tools do different things better than others and are beginning to segment into areas of competence. And so, if I want to create a very simple work bot to do a very simple workflow for me, I’ll use another AI system to basically draft my initial concepts, and then I’ll take that very rough work and I’ll give it to a different kind of AI system to elaborate on that idea.

  Naren Ramaswamy:
  Critique it, or add more ideas.

  Mike Collins:
  Critique it. And then, I take that into a reasoning model and I’m saying, take this work and make it better. And I take that work, and then I could take that work and write a paper on it using a different tool. Or say I wanted to then take that paper and listen to it on my ride home from work, I’m going to use a different tool to do that.

  So the idea of AI scaffolding the idea—this will get me to this to get me to this to get me to this—I think is in this family of ideas and innovation going on. And I think today, having a family of AI systems that are kind of customized to do different things well… You and I were talking about this AI system dropped by Stanford called Storm, which is, in the simplest terms—and this probably undersells it—but it combines a really good kind of web crawler and an LLM.

  And like, oh, it hasn’t really combined that way before. And that allows it to do certain things better than other AI systems. So you get Anthropic and Claude’s really good at certain things. OpenAI and Gemini from Google—there are choices being made in the design of them, it seems, that allow them to have strengths and weaknesses.

  And I sit here and try to understand those and then orchestrate what I’m trying to achieve in the world and in my day by using the right tool for the right job. In the same way, in the internet era, when do I use a Word document? When do I use PowerPoint or Excel?

  Naren Ramaswamy:
  Different applications.

  Mike Collins:
  When do I use email versus text? I mean, we all use all of these internet-based tools for different things. And so, it’s just adding a new set of instruments, if you will, to do things. They just happen to be incredibly powerful.

  Naren Ramaswamy:
  It’s like having your own advisory board. Every person has their own advisory board for anything. Going back to that crypto example, what really stood out to me was not the ability to generate the idea itself—that’s fine—but the fact that it was able to market it, to really hype up the value of its own coin. These are all manual tasks again, but it makes a marketer’s job easier.

  We’re already seeing a lot of LinkedIn posts that you can tell somebody has drafted straight from ChatGPT, but the productivity of that person has significantly increased as a result of that. So we’re heading into that future where every person will have this advisory board, and they’ll spend maybe 10% of the time that they spent on some of these tasks that five years ago took up most of their day.

  This is going to be huge productivity enhancements across the board. People are talking about how automation will cause job losses. The other side of it is that everybody can do 10X more. That’s the tech optimist way of looking at it. Not to diminish the concerns of automation, but the reality is going to be somewhere in between—somewhere between extreme efficiency and some short-term job loss.

  Mike Collins:
  And I do think it is a fairly democratized tool for white-collar workers. I think people have the choice to lean into this. Whatever job you’re doing—and I say this to my board—I think there are going to be two types of CEOs. There’s going to be tech-amplified, AI-amplified CEOs, and former CEOs.

  I think that applies to every job. If you want to lean into these tools and get really good at them, you will be orders of magnitude more productive and an order of magnitude ahead of your non-AI-enhanced competitor. But in that world, what we’ve also seen in our society over the last 10 or 20 years is winner-take-most. Markets become bigger and global, but it becomes a more segmented society.

  So I do think that, on the net, it’s great and it’s good, but there are going to be local dislocations that are pretty profound. So again, lean in, people.

  All right, Naren, it was a good one—big week. Say that every week, it seems. Good discussion. We’ll do it again. Thank you.

  Naren Ramaswamy:
  Yeah. Thanks, Mike. See you.

  Sam:
  Thanks again for tuning into the Tech Optimist. If you enjoyed this episode, we’d really appreciate it if you’d give us a rating on whichever podcast app you’re using and remember to subscribe to keep up with each episode. The Tech Optimist welcomes any questions, comments, or segment suggestions, so please email us at [email protected] with any of those. And be sure to visit our website at av.vc. As always, keep building.