Vir Sanghavi about Tilt.vote and how it can improve debates in democracies
Vir introduces Tilt.vote, a platform that uses AI to rank comments based on a “Logic Score” rather than just popularity or rage bait. We discuss how the AI detects fallacies, the future of AI agents in the workforce, and how civic tech can help save democracy from polarization.
If you enjoy our podcast, please consider making a donation. It really helps us to sustain the podcast.
Top Keywords
- tilt 0.023
- debate 0.014
- logic 0.010
- logic score 0.009
- logical 0.009
- score 0.008
- platform 0.006
- arguments 0.005
- saying 0.005
- mean 0.005
- best 0.004
- argument 0.004
Transcript
Welcome on another episode of Democracy Innovator podcast, and our guest of today is Veer Sangavi. Is it right, the pronunciation? Yeah.
Okay. Yes. It is. Thank you so much for having me on. I'm very excited to be here.
Thank you. And work welcome, and thank you for your time. I discovered you, I mean, that you are building this platform that is called tilt. Vote. Would you like to tell us something more? Yeah. Sure. So tilt. Vote is
a primarily social platform where users can post debates, vote, reply to debates, post takes, and really just have a place to voice their arguments and what they believe in a structured meaningful way. Right now, social media is kind of an echo chamber where everyone yells out into the void, but the best ideas get buried beneath comment sections. There's no method for people with the opinions that they want to express to rise to the top, and it's really just based around a culture of virality and rage bait. Tilt is intended to fix that, and it's necessary because everyone has an idea, but not everyone has a place to express it. And that's where I think Right. It's so important in today's democracy because the at the core of a democracy is everyone gets a voice, and I think Tilt is the platform to surface that.
And, when you had the idea about this platform, like,
Yeah. I had this idea a few months back, and,
I've been very grateful and blessed to have this idea. I've just been building ever since. And, no. But how did it happen? Like, do you remember if there was, like, a moment you said, okay. I have to make this platform, or, like, because you told me you also have experience with debating. Right?
Yeah. So from sixth grade through now, tenth grade, I've been participating in competitive debate. I've won some pretty crazy titles in middle school, like, I won this, state championships for Texas, and I've finaled nationals in The US twice. Obviously, it was middle school nationals, but it gave me enough exposure to the debate world to understand that real life was not as civil, structured, or, as easy to explain your opinions in as debate. Because the way debate works is people with the best ideas and the best presentation will generally win. And that was a void I saw in real life because as I told you before, whenever I wanted to express my ideas, the best way to do it was social media. And social media was just me yelling to the wall, And that was very meaningless. So after a few weeks of just being frustrated because I wanted to say things, I wanted to have what I said have impact. I just didn't know where I wanted to go. I didn't know what to do, where to put my opinions, and that inspired me to build Tilt.
And, also, we have to say you are quite younger. Right? And, because of maybe someone listening to the conversation, doesn't know. But, you're how old are you at the moment? I'm 15 years old. Okay. 15. And so you already had, some experience debating. How does it work like a a debate? Because you you you said that, the representation
Yeah. Yeah. So debate is, based on different, events. There's congress, PFLD, and there's a bunch of interp and, like, presentation based events where it's not actually debate. It's more presentation. But with regards to debate, I participate in congressional debate, and the job of a congressional debater is to go into a room full of anywhere from 12 to 25 people and represent your constituents. You have different bills or, resolution that you want to either pass or fail. You give speeches, telling the chamber whether to vote to pass or fail. You get questioned. You argue with other people in your speech. You reference their speech, and, say what was, like, inconsistent with it or why you support what they said. And it's done in a timely manner where you have to present in front of the room and, speak to everyone at once, and that's where presentation matters as well. Other forms of debate include, like, public forum in Lincoln Douglas. Is a two on two format of debate. There's, again, resolutions, and you have to either go in the after NED. Lincoln Douglas is, is one one v one. And, I haven't done LD, but what I've heard is it's more based on ethics, values, and
impacts. So there's Okay. A lot of different forms. Yeah. No. Actually, I didn't know about all the all those forms. And, and so, like, TiltBot, in some way, tries to represent, these, kind of debating, like, in in, it adopts a specific form of the one that you described, or it was a way, for you to, I don't know, invent a platform that could, try to end the meeting?
So tilt that vote is not meant to embody any form of debate of debate because, debate has a very, hard entry point for any beginner. It's kind of you have to learn the procedures, the, rules, the, like, expectations in any debate round. And for most people who don't have experience in debate and still have ideas they want to get out, having those rules and restrictions of any specific debate event would be very limiting. So Tilt is designed in the most, user friendly way that it could have been, which is users post a debate, users argue, and
AI will automatically push the best arguments to the top. And, and now there are some users, right, in the platform?
We're currently sitting at over 4,000 users and are growing every single day. Like, I'm very excited that people are also can coming back. We recently had day 7% or day 711% user retention, And that means that people are recognizing Tilt as a place to express their ideas, which I'm very grateful for. And, have you discovered anything, like, during,
because, you know, sometimes you can have an idea, then you actually I mean, you realize the idea, so you have a sort of product. And then, you see that maybe, I don't know, users are using the platform in a different way from when you first had the idea, maybe. I don't know. Like, have you had any surprise? Like,
Yeah. For sure. And I think that's for me, it was that the original AI integration part part of Tilt was supposed to be summarization. You see these long threads of comments and, arguments in any like, under any news article, for example. And you see that people, have to, like, scroll down and read through all of them to really understand what's happening. And I thought that summarization and AI summary at the top would be what's most popular. But looking at, AI request logs, it's that people are barely using the AI summarization feature. What's more shocking to me is that logic scoring that's done through Tilt's AI is way more popular, way more requested from any IP address. And I think there's a few reasons for it. Right? Like, there's people want to see how they're arguing. People want to see how others are arguing. People want a consistent, fair judge, and I think that outweighs any summary. It was really a shock to me at first that because I thought that the summary would be, something that everyone wanted to see after seeing the, like, length of these debates. And while it is true that for longer debates, the summaries are more called upon, it baffled me that people wanted, to see the AI logic scores more than AI summaries. And,
how does it work, the logic score? Yeah. So
the logic scoring system for Tilt's AI is built on very, transparent criteria. There's four, parts of any logic score under any given debate. All of these parts are quantifiable to some degree, and I'll go into them one by one. There's a fact check, which, great and all of these, four criteria grade your take on a scale of zero to 100. 100 being the most desirable and zero being the worst possible. So there's a fact check which looks to, real sources, real like, does a web search to see if what you're saying is logical. Then there's, a fallacy or no fallacy check, which looks to common logical fallacies. Some of these include, like, ad hominem, straw man, false dilemma, appeal to authority. And all of these different logical fallacies can undermine the strength of your argument. And that's why we check for, logical fallacies going from a scale of zero to 100. Then there's relevancy within the debate. Because if I can if I post the best argument, one without fallacies for a debate about Israel versus Palestine, where, the actual debate is about pineapple on pizza, then, it shouldn't be scored highly. And that's where relevancy comes in. Looking at the context of the rest of the debate and rating your take from zero to 100. Lastly, there's logos versus pathos, logic versus emotion. How much of what you're saying is based on, a consistent chain of reasoning and chain of thought as opposed to just yelling? And, all four of these scores are averaged out to produce a logic score. It's all visible under any debate at dot dot vote. And then the AI also gives a brief rationale for its logic score. For example, in a debate about how Israel or the debate said Israel will win the war against Palestine. What happened is the top the user with the best pick said it's not a matter of winning or losing. It's a prolonged conflict that cannot be resolved through military victory alone. He went along with it for a little while, and then he said, however, Israel could win, could take control of Gaza if they wanted to because of their military superiority and alliance with The United States. The AI's rationale for giving a 92 out of 100 was overall, because the argument is well grounded and verifiable facts directly addresses the debate and shows no clear fallacies. Its reasoning is evidence led and the presentation is well structured. So for each tape, it's, graded and then, given a rationale for its grade. Recently, I was also able to, put logic scoring on replies. So whenever I reply to an argument, that will get logic scored as well. So by default sorry. The arguments with the highest logic score will always get pushed to the top of the debate. Say there's 50 comments under a debate. So the way the debate page is structured is there's the title, then there's vote bars. Do I agree with the debate's topic? Do I disagree, or do I am I neutral? Then there's a place for you to post your take, and then there's a whole chain of arguments underneath. So because of this AI scoring system and one more factor that I want to talk about next, the best arguments get pushed to the top. So the next, part of the, like, argument scoring and ranking factor is a rank score. A rank score is compiled from the logic score, which is an average of the four criteria put from zero to 100. And the rank score is the logic score, timed the amount of net upvotes or downvotes. So one thing about tilt is that it's also user driven. Each argument can get upvoted by users or downvoted. No. So say, an argument has a logic score of ninety two and three upvotes, then the rank score would be 92 times three. The way that or the reason that I did this was because they AI says an argument is the best, 100 out of 100, but every user hates it. Then it doesn't deserve to be at the top because it doesn't, match with, like, what really does make sense in a user's eyes. And seeing how Tilt is supposed to surface the strongest and, most logical arguments, I think user, intervention is an important part of it. And that's what compiling the rank score came from. And, how how is working at the moment,
in your opinion, like this, I mean, having because it's it's interesting the fact that, usually, if we think about, I don't know, Reddit, people, can, how do you say, like a poster, and that post goes up in the, in the in the feed. But here, there is the logic, I mean, the four, scores plus the like or dislike. And how is it working?
Yeah. So what I've heard from users is they're loving it. Like, everyone that has spoken to me about it tells me that a common flaw with Reddit is the noisiest arguments will always rise to the top because people who like, the loud minority will post the craziest arguments and just upload each other's arguments. The quiet majority will just post something, but won't be there to, like, scream into the other side's face by liking their side's comments. And that's where Tilt's logic score comes in. Because even if I have an opinion that the rest of the world disagrees with, I still have a chance for it to be seen and for it to surface just because of the logic score. So it it does a great job of bringing value and, attention to every argument, and that's what I think is so powerful about it.
I just had an idea.
Yeah. Go for it.
What if, like, let's say you also use some Reddit conversation, to test it? So, like, to analyze the comment. And let's say, at the moment, there is a certain, so there is one comment that has received more upvotes, and so it is at the top at the moment already. Like, maybe using your algorithm, then that comment will not be at top anymore because maybe, and also we start maybe some conversation.
For sure. So I've actually recently built a REST API for Tilt to be embedded in any website with just one to four lines of code depending on the structure you're using. And what that allows you to do is literally bring Tilt's entire consensus engine over to your website and for you to be able to post arguments, create debates, and really just run them on your website. If I can share my screen, I can show you what it looks like. Is that alright?
Yes. But consider that, a lot of people will just listen to the conversation.
Okay. Sounds good. I'll talk in a way that allows everyone to understand. And, actually okay. Yeah. You can share. Okay. Perfect. So here is the API reference. You can see that this kind of consensus engine will always, with just a few lines of code, be integrated at the top. You can see that you can agree both neutral, go disagree, post takes, replies, and even generate a summary. And it's really so powerful because any any debate can now or any, website will now be able to boost their user engagement and fairness uptake sharing. This REST API allows not only for Tilt to make money long term through contracts, but for a small blog or, as you say, Reddit, to embed it into any one of their subreddits or threads and, automatically surface the best arguments. I believe that it can be integrated into Reddit, Discord, Slack, WhatsApp, iMessage, everywhere, and it's gonna happen very soon. I'm working very hard to create those contracts and talk to those potential clients. Okay. And,
how let's say that tomorrow, everyone will start, I mean, if not using that app, but, like, debating in a certain way. So, actually, using logics in a certain way. So, so how do you think that the work could be? Like, do you think that at the moment, you said about the noise, that there is, on traditional social networks? Like, I I I don't know. As an example, you were debating, so I suppose that, maybe you have, friends that also like to debate. And, is it easier for you to I I don't know, like, maybe it's an hard question. Like, in comparison maybe to other people, could be other young people that are not used to debate.
Do you think that this logic and this clarity helps you to Yes. I get what you're asking. I I know where this question is headed, and I'm very excited to answer this because this is one that I've spent a lot of time thinking about and working to answer as properly as possible Yeah. Just in the way I've built the site. Let me tell you this. Everyone has opinions, and everyone wants to see it rise. So during the last twenty twenty four presidential election, between Trump and Kamala Harris, there were so many people who wanted to advocate for a certain side with their friends. They would say, oh, tell your parents to vote this guy because or vote this person because blah blah blah. And then the other side would say, well, that doesn't make sense because yada yada yada. And, so many people who wanted to argue during that time with their friends in a civil way have came to me and said, look. I love what you're building. I love the AI summary because if I add this during the election, then my friends and I could have really seen who was winning and who was right. And, the way that they would have been able to seen it see it is logic scores, and that's why I think it makes sense for everyone. Have you had,
was it, maybe last, interview that I done? I mean, we were talking, about how it could be that people use the same word, but, they think about a different concept. So it was with Richard Bartlett, and, he cofounded Lumio some time ago. It was 2008, I think, or so. Yeah. And and, basically, we were saying that, if we say, I don't know, capitalism, I mean something, you mean something else. And, otherwise, maybe we think about, a concept. We have a concept, but then we express it in a different way. So let's say, I don't know, for me, like, to have, freedom, stability, peace, and so on, the perfect candidate is, I don't know, that persons. And but and for you, and you also want peace, and stability freedom, but for you, it's someone else. Or maybe you want to do it in another way. How does it work, this with debating? I don't know. Yeah. So
one part of tilt is that arguments that are very short or very weak don't get logic scored. If I just go into a debate and say, like, five words, capitalism is a good thing. Then, obviously, I can manually call for a logic score. I can say I want it to be scored, but it won't get scored highly. And one way that Tilt is working to reduce people saying something and meaning something else is through our AI rewrite feature. Once so if you're writing an argument and you've surpassed, like, a 100 characters, you can call the AI rewrite feature, which will restructure and, rewrite your argument in the way that will allow the world to best understand what you want to say. The reason we kept it at a 100 character minimum is so you can't just use AI to draft an argument. You actually have to say what you want and then have AI clean up the noise. That AI rewrite feature is also embedded into debate creation and replies. And that's why I think that occurrences of people saying one thing but meaning something else are happening less and less. Okay. So you think that,
because AI can help the person to express in a better way what he or she means, then this doesn't happen. I'm actually curious to see, because also, this is something, like, I don't I don't know in in The US, but many times, I mean, I mean, when I was younger, it happens to me, I don't know, like, to talk, with my mother and to say, oh, but this is not logic. It's not logical what you're saying. And so we were arguing, and they thought, okay, what I'm saying, it is logical. And, so, I mean, she is wrong. Yeah. Yeah. Then I then I discover, like, that there are several kind of logics. So what I say can be logical from my point of view with the logics that I'm using. So I'm not not very I'm not an expert, of this field. But then, also another person can be right using another logical system. And this makes everything more complex than what it is. Because then, saying as an example that, what I say is logical, it's also a way to, legitimize what I'm saying. And also to say that that basically you're wrong. Yes. So I don't know if you have any thought. Of course. Yeah. So
first of all, if I had if you had tilt back then, I would have loved to see you arguing with your mom on tilt, But, it would have been a hassle to get any mom on tilt just to argue with you. But in terms of logic and saying that one side is logical in one way and another side is logical in another way, I think that's true, but to a certain degree because logic has a very strict, definition to it, which is that if your reasoning for your claim makes sense, if it's grounded in reason, then what you're saying is logical. So if I'm arguing something and I have a reason for it that's logical in my eyes, but maybe not in yours, and you're saying something that is logical in your eyes, but not in mine, then as long as it has verifiable, rationale to it, that's what makes it logical. And that's what Tilt's AI is meant to embody. Because if I say an opinion, again, that the rest of the world doesn't have, as long as I can provide proper reasoning for it, it's considered logical. K.
Yeah. Thinking that also, like, I I really think that AI can help, from this point of view because, also can help us to discover, what are the differences in the meaning that we give to to words or concepts.
Yeah. I mean, to be honest, I think that the next generation of civic tech is all going to be AI driven, because AI allows human voices to be represented and scored in a way that is not capable for humans to do at scale. Going into this going into the rest of the twenty first century, I mean, I'm sure life five years from now will be very different from what life is today. AI is going to automate, revolutionize every every minute of our lives. We're going to be looking at something a that has AI in it, interacting with some type of artificial intelligence. We might have even reached AGI. So for any platform to have weight in the coming century and to outcompete its AI using competitors, you have to integrate AI because that's what allows users to enjoy and interact with your platform in a meaningful way at scale.
Yeah. I'm also thinking that in the future, we will see, a lot of AI. I mean, also in the latest year, we had a lot of AI, but, it was not visible. It was behind, behind such a networks, basically. And, and probably it was the one that was creating the noise, that there is now such a networks. And, and then I was curious, like, you build a platform all by yourself? I don't know. You have, you are more than one person. No. How does it work?
And, Yeah. So there there's no way I was building this platform without the help of AI. I'm sure AI has, built around half of this platform up to not a little bit less. And, really, it's I work on a tech stack of Visual Studio Code and GitHub Copilot, which is then deployed to Vercel, via GitHub. And, I've also wired a live Supabase backend that constantly fetches and, constantly fetches and puts all the data that Mhmm. Is needed on a page to that page. So it's also built on next.js with React. And, like, obviously, this this is a full stack development setup. It's very complicated for anyone to know. And just as I was saying before that, like, oh, AI is going to revolutionize this. AI is going to revolutionize that. It's really, been instrumental to my ability to build in my ability to build this kind of platform. I'm sure I could have built something similar, just not as capable without the help of artificial intelligence. But, yeah, aside from that, it's just me building. I've had a few friends try to get involved, but,
none have stuck around. You were debating about how to make the platform? That's why. Yes. Okay. So different ideas or not. Yeah. I saw that by the way, it's it's hard to do something, with other people. It's harder sometimes than, Yeah. And using
yeah. Please. Go ahead. Yeah. No. Sorry. I think it comes from a motivation point of view. Like, if I'm staying if I'm going to school from 08:30 to 4PM and then coming back and working on this from 04:30 to 5AM and then sleeping from five to seven, there's going to be very few people that are willing to put in the time and effort that I'm going to to make something like this happen and succeed at scale. So if I want to get anyone involved, like, if there if someone wants to be a cofounder with me and comes with this level of hard work and motivation to make something like this happen, I want to work with them. But if someone just wants to work with me for, like, a resume boost or, that kind of thing, then it's like, sure. You can if you help out, you can take the credit for it. But I don't want you as, I don't want to rely on you to build any part of the platform. And that's why I think up to date, it's been just me. It's interesting.
And, yeah, as I was saying, yes, sometimes, it it is harder with other people because you have, to to agree about everything. I mean, if you're like, if you have a cofounder, well, of course, if he's, like, an employee, you can, like, tell him what to do, but that is a second step. And, also, I really think that with AI and other technologies, maybe we could, I don't know, in the future, we will be able maybe to coordinate without having a boss, let's say. Maybe having a leader because a leader, like, in some way could be, like, the one that is, oh, I have this idea. I have this vision in my mind. I want to. I know how to do it, and, please help me in doing that. Mhmm. But maybe without, like, having a boss that tell you, you have to do this, and, you just have to obey. Yes. And, also, like And yeah, please,
please. Sorry. It won't necessarily be only not having a boss. It'll be not having employees. In the future, only the most skilled workers will stay around. And my estimate is by 2035, 90% of employees in a corporate workspace will be gone and replaced by AI agents. Even recently, I don't know if you've heard of it, but a company in China tested an AI CEO instead of their real CEO, and, they got delivered significantly better results. So there's a benefit to having AI agents that I don't think humans will be able to outcompete in the next few years unless, and big unless, they're more skilled with AI than just AI by itself. If they can deliver more results while working with AI, then AI can by itself. It's imperative that they stick around. And that's why, like, you're saying, oh, it won't be it won't necessarily be a boss. It'll just be a leader or a visionary in the future. I think it won't be employees that follow that leader or visionary. It'll be a team of AI agents that do it. Yeah. I'm thinking, like,
then what remained to humans, is, I'm thinking maybe the production of data, but then, so if I use the platform in some way, I I contribute with my it could be a tilt or another platform. I, ex express my ideas and thoughts. Yeah. And then I don't know. Like, AI is evolving very fast and, actually don't know where it could lead. Like, you know, there are people that say that it will lead, like, to the, the best possible world and other people that say that maybe this will lead to the destruction of the world. And then there are other problems like, climate change and, and worse. We we have AI, but still we are having worse. So will be nice to fix that small problem, like, that are called worse. Now it's a big problem.
And just because you brought this up, I wanna kind of go back to talking about a certain part of Tilt, which is its AI. You're saying that every time you interact with the platform, the AI gets trained even farther. And I think that's what in the next few months, few years, few weeks, whatever it is, is going to make, Tilt more valuable because every time someone interacts with Tilt, our AI is actually given more data to train on. And over time, it will really help, make the biggest change in democracy by allowing the loud or the most logical voices to rise over the loudest. AI being trained on this new data is very imperative to the success of any major social platform, and, I'm doing the work it takes. I'm trying to do the work it takes to get there. Yeah.
Having the the, this is a nice idea. I mean, several, I think, civic tech, software are, trying to do this, or at least they aim to do this to to have all these, flow of conversations and so on that that, is, I don't know, aimed to fix a problem or, like, and and then using that data to train an AI so that the AI can fix problem, instead of us, like, without us, basically. Yeah. That is also another interesting thing. Like, having AI that could help us in, let's say, fixing problems. But then it's also very political in some way. But but maybe we we are like, it's too complex maybe also for us.
Yeah. I mean, eventually, what I want to see Tilt AI do is, suggest solutions to problems. After it's trained on enough data, I want for Tilt AI to be able to say, like, this could be a potential solution to this conflict or to prevent global warming in this, field in this region. Here's the actionable changes that can be made. And, not only for Tilt, I want to see other platforms doing that too because AI is capable of it. And it's just not being used in the most productive way possible yet. Yeah. Unfortunately, also, other technology doesn't are not used in,
in the best way possible. Like, if you think also Internet that that is, like, the basic layer that, allows us to communicate and so on, often it is used, for the, for the noise and also to polarize, because sometimes, like, let's say you're a person that have power and, you want to have, and there are other people. If you want to keep that power, it's better if the other people fight with each other. Because if everyone is, you know, like, friendly to everyone, then they could say Yes. Maybe we don't need a boss or, like, and so there are, like, some, structural changes that should be done maybe to use technology in the best possible way. Because at that time, otherwise, what happened is that, technology, or anyway, like, you know, you know, like, it's, how do you say, like, I know in Italian. Yeah. Like, to keep power is better if people are not united together.
Yes. Yes. Divide and conquer. Yeah. Exactly. Exactly. Thank you. And over time, I think that with any platform that is meant to surface the is meant to push democracy forward by allowing people to see where they stand together, there will be state level challenges. People in power won't want to see it succeed because, as you say, it can undermine their power. I don't know what where it will be. It could be in The US. It could be Italy. Someone at a certain point is not going to be happy with, pushing the loudest or bringing people together because it can undermine their power. I don't know if you know about it, but that's what the British did to conquer India in, a few hundred years back, where they divided it up and conquered it just and then exploited the entire country. And if you're stopping leaders from dividing and conquering people to stay in power, then there will be challenges. And I think AI is the perfect tool to address those challenges because at the heart of it, it's not someone making the summary or it's not a person grading the logic, it's AI. And when when AI is doing it, it's very hard to pinpoint blame on someone. Yeah. Absolutely. I I share it,
as I said, the long term, hope. I'm also, also share it like, I hope that in the future we will able to do this. At at the moment, what we have seen and is that sometimes AI still resonates reasons as humans Mhmm. Because they were trained on human data. There was this example. I don't know. I I probably they fixed it. I don't know. But there was this paper, and they were saying that if you say today, I, if you give me a precise answer, I give you, like, $1,000. The AI was replying in a in a better way. Or, like, sometimes, there are, like, differences compared to how women and men are treated, by the AI because there are some, prejudice, that are, like, embedded in the in the AI system. Mhmm. But, yeah, I really hope that in No. No. Sorry. Go ahead. No. No. I I was just saying that, yeah, I really hope that in the future, we will be able to fix, those kind of problems, and I have an AI that is, really neutral.
Yeah. Yeah. For sure. That's what Tilt is working towards. That's what many other companies are working towards, and it's gonna happen soon. So I'm very excited for what the future holds. Yeah. I'm also very excited. And,
I mean, you told me something about the future, how are you imagine it? Mhmm. And, also, like, if you like to say any something about democracy and how you imagine the a possible world when you grow up, like, in, I mean, now I'm 33. So, like, a little bit more than the double of your age. So, like, in fifteen, twenty years, as we said, yeah, probably we'll have incredible AI models, or maybe the the world will will, will see, I don't know, the third world war. Yeah. I hope not. Yeah.
If you I think that I think that in the future, democracy is going to be, like, democracy ten years from now is not going to be the same democracy it is today. Just like the industrial revolution, like, completely transformed every aspect of society from when it happened to after it was done, I think the AI revolution is going to be just as, if not more significant. I think that voting will be done with some sort of AI embedded into it. I think that candidates will speak to the best they possibly can using an AI teleprompter instead of a physical teleprompter. I think that, I think that every form of, like, potential debate or civic disagreement will be moderated and controlled by AI. I know that right now already candidates are using AI to kind of map out their speeches, direct policy decisions. OpenAI gave the US government access to chat GPT for $1. So, and Claude did the same as well. And I think that as these companies continue to build even more powerful models, as you say, in, fifty years, we might have World War three or we might have the best utopia we could have imagined, Every little even the smallest aspects of our society are going to be transformed. In day to day life, this might look like, a humanoid serving you coffee at 8AM in the morning or, maybe let me think. Maybe, like, Ubers with human drivers are going to be obsolete. Maybe, like, robo taxis are going to be the only way we get around just because they're so much safer and cheaper. There's so many every aspect of society that I look at has potential to be transformed by AI, manufacturing, civic tech, transportation, whatever it is. And with regards to what the future holds in my eyes, I think it's one that is going to be increasingly shaped by civic tech platforms because they're what is going to guide the world's policy decisions. And if they can be done in an ethical and, fair way that really bring gives everyone a voice, then we're looking at a better world fifty years from now than without these civic tech platforms.
Yeah. I think in diet, yeah, I really hope that we will be able to use technology in the best, possible way because all the other times that we invented amazing technologies, then we use them to kill each other. Unfortunately.
Yes.
So, yeah, I really hope that we will be able to to do it. And and, also, like, I haven't asked you anything about yourself. Like, where do you live? Or, like, if if you want to share something, I don't know if you have brother, sister's
hobbies. Yeah. I live in Houston, Texas. I have two younger brothers that are, I forget. They're growing so fast. They're eleven eleven and 12 years old. They're going to turn 13 and 14, and it's like, they're a big part of my life. They kind of helped validate Tilt at the very start. Like, they tried to find flaws in the idea. Even my parents, like, they're they're to support me, to help me, and, like, I'm very grateful for that. Thank you. And,
and do you have any other question or things you would like to add to this conversation?
Yeah. I'm curious because this whole call I've been saying, like, what AI looks like in my eyes. Where do you see AI transforming the world twenty five, fifty years from now? I don't know. That's why I asked, do they have people? Okay. Awesome.
No. I mean, yeah, I have some ideas. Then I think it's very hard to to have our ideas, and then, because in some way, we are influenced also by by the world. Sometimes also in a way that is not very easy to recognize. Now I'm not, I remember I've read about how I don't remember which movie it was if, there was some movies about Mars and exploration about, maybe it was interstellar, maybe it was another one. And, I remember that, sometimes, you know, there is a movie about space. But maybe there are reasons why there is a movie about space. And, and it is because, maybe, they want to move more money, related to space research. So people dream about space, colonization and so on. And so they are they also accept that a lot of money are dedicated to to that field. So sometimes, we are not, it's hard to think with just with with our mind. You know, there is like a famous, series, like, how it's called about the future. I think it's on Netflix. Like, it's very, very famous. Also, I don't remember now the the the the title, and they and they know that it's all about AI, about the future, and, how it will looks like. And a lot of friends, I know that, they watch it, and they told me, like, hey, Ale, you should really watch it. It's, it talks a lot about what you're saying about, but I don't want to watch it because I don't want to be influenced, because, it is true that I I do think that, in five years, we will have probably an AI that can like, everyone can create, like, an application, just saying, hey, Alexa. Hey. Please make this up for me. And the AI will will do it. And, yeah, we will have, probably these kind of things. But it's not, I think we should never be, think that it will happen. Because we are like, we can change it, always. Then then I I hope that it will be very interesting to have, an AI very, very smart. We also know that AI in some ways is very, very stupid. Yes. So I actually don't know how the future will look like, but I think that sometime we should have an idea and then destroy that idea.
Yes. For sure. For sure. I'm like, I can't reiterate this enough, but I I can't wait to see what the world looks like in five to ten years. The world, before COVID was so much different. And then when I found out about CHATGPC in 2022, it was like a gold mine in school. CHATGPT ing essays and everything. It was middle school. But, it's radically transformed every aspect of my life. Whenever I have something that I'm not sure what to do, chat gpt.com. Whenever, like, I'm just very excited.
Yeah. Yeah. I'm also very, like, curious to know how, but I cannot really experience it because when I was 15, there was not such GPT. And Yes. And and so I I mean, I I was very nerdy, so I was maybe using tools and exploring computer science more than other friends, but, there wasn't.
So I I I completely get it. Before, like, AI came along, everyone older than me at eighteen twenty was talking about, like, crypto and blockchain, and those words sound like alien to me. So, I get what it's like, being there before you're able to experience everything. But, luckily, we're at the point where if we adopt and start working with AI super well now, we'll be able to reap its benefits,
dozens of years into the future. Yeah. I feel we should put a lot of effort, to do it, to do something good. And I yeah. Because as I said before, it's it's up to us. Like, we can change, so probably there will be a third world war, but we can change it. We can actually decide that we don't want a war. And, and we can create tools that maybe could avoid, to have new wars.
Exactly. I think AI for bargaining would be great. Like, just president Singh what they're able to give up. AI bargaining with AI on the other side and coming to some sort of decision could be a reality just a few years from now. Yeah. From a technical point of view, I think there are a lot of things that can be possible,
from, but then there are other, problems. I will say political problems. There are, like, yeah, a lot of other problems. And, but hopefully, we will be able to solve them using, some civic tech tools, some AI that is, trained to do that. Trained by us and, tools developed by us. So thank you a lot, Veer.
Yeah. Thank you. Of course. It was great to have this conversation with you.
Same for me. Thank you again. Thank you.
