Can A.I. Sway the 2026 Midterm Elections? | Chris Dessi Explains

How can voters during campaign seasons know that a video is fake or real when it comes to these A.I. deepfakes? They can't. Hey, it's Natasha. Can I ask you for a quick favor? If you've been enjoying these conversations, hit the follow button wherever you found us. If that's YouTube, Spotify, Apple Podcasts, wherever you found us, that way you never miss an episode. And if you have 10 more seconds, can you leave us a review, maybe even rate us? That way new people can find us. As always, thank you so much for spending your time with us. Hey guys, welcome to the Talking With Natasha podcast. Each week I'm sitting down with founders, trailblazers, and some names you definitely know. Matt Friend. The Watch King. Neil deGrasse Tyson. Darna Gard. The Scarlatta Twin. Cal Penn. How are you? But we're skipping the small talk. We're taking it to the next level. It's live, thrilling, fun, and honest. Let's get to mucking. It's ChocoClock. How are family gatherings, though? Like, does she... Painful. Almost like getting a colonoscopy. So come hang out and let's get into it together. Hey everybody, it's election day and one of the biggest issues on the ballot is artificial intelligence. So my next guest lives and breathes all things AI. Please welcome Chris Dessie. Welcome. Yeah, it's so great to have you. So great to be here. Shout out to everyone. Already our live chat is popping off and hey, we have a studio audience, a live studio audience today. Hey guys, great to see you. Look at that. Fresh-faced, smart, ready to conquer the world. I love it. Ready to take on AI and talk all things artificial intelligence. So Chris, I think first tell our viewers and everyone who's joining right now what you do with AI because you work with it every single day. Sure. I mean, I've been in tech for 30 years on the revenue side. I've managed teams. Built agencies, sold agencies, built other businesses, written books. So I've always been a student of it, but from a capitalist perspective, you know, doing work on behalf of clients. And now my current company, Torque AI, we consult and we implement AI within organizations. And I do executive coaching for one-on-one for people that really want to implement AI throughout their lives. So you've seen AI through so many different facets then and how it's changed the world I feel like in the last couple of years, even the last year, it's escalated so fast in its development and what we're learning AI can be capable of. When your mom and grandma start talking about AI, then you're like, okay, everybody's got it. A shout out to my mom because she uses ChatGPT so well. She creates pictures and then she does real estate and she's She's able to do staging. I mean, it's, it's, it's. Fascinating. It's a wonderful tool. It's a wonderful tool. When it's used properly. When used properly. So what is the biggest misconception about these tools like ChatGPT, Claude, for Americans? Oh, for Americans, the biggest misconception is that it has your best interest top of mind. It doesn't. It's a machine. It's a program. Also that it will be objective. It's not. It doesn't. It's a program. Cornell did a study. Where they realized that some people were asking chatbots how to vote. And so Cornell said, hmm, how to vote or who to vote for, who to vote for, who to vote for. And they got, you know, like you just hear that. You're like, oh, goodness, I didn't even think of that one. Right. And what they realized was that OpenAI was leaning one way and Perplexity and Claude were leaning another way and that it became a problem and that you shouldn't ask. AI how to vote. And right now the consumer version of it will not give you an answer. Cornell study, they did it through the developer version of it to just to see what would happen. And there was only like 10% of the time where the chatbots agreed. So it's, it does not have your best interest because it's referencing different pools of data, right? It's all about data, right? But we rely on ChatGPT for everything. We shouldn't. We ask it relationship questions. We shouldn't. Why shouldn't we? Because it's not your data yet. And this is why voting is important. Because if you are not paying attention to the manner with which we are regulating this and monitoring the technology, then it becomes a slippery slope. There was just a recent story where... I'm going to mangle exactly what it was, but generally in discovery during a lawsuit, the attorneys asked to have a printout of the person's ChatGPT searches as admissible in court. And what they discovered was they were asking ChatGPT how to lie during that specific lawsuit. That's a problem. And that's a little bit scary for people. It's like, oh, it was just a hypothetical, right? Or you're asking relationship advice or you're confiding in ChatGPT. Right now, that data is not your data yet. But this is why we want more privacy laws implemented so that that does become your data and they don't have access to that. Who controls it and who regulates it? That conversation is happening actively before our eyes. Yeah, on Capitol Hill, we've seen tech leaders come out and talk about and say, Hey, you know, I guess, like, stop us. They're almost saying, like, stop us from doing this. We're on a crash course, basically. It is, it's really. Doesn't it make you feel good? It doesn't make you feel good. It's a little, like, of a, it's mind bending. It's a little mind bending. It's crazy, yeah. But who should be in charge of controlling this? So I can only share where my perspective. Okay, so I'm not going to say this should be or that that should be. When it comes to something this complicated and this all-encompassing and completely pervasive within our society now, I believe there should be one source of truth. I believe this should be coming from the federal government. And I only say that, well, not only, but I'm saying that because if we really dig in and look at, if you have 50 different sets of laws from 50 different states and you're a small to mid-sized business and you do business in different states, imagine The regulatory compliance hoops that you're going to have to go through by making sure that you're following those guidelines in different states. And that's not even the worst part. The worst part is big tech companies can handle that. So they're like, yeah, sure. 50 laws, 50 different states, that's fine. If there is a definitive fine, right now the fines are like, I think they're listed like $15,000 and $25,000. For a multi-billion dollar tech company, that's nothing. For me, I don't want to pay a $15,000 penalty if I happen to be doing business in New Jersey as opposed to in Westchester, New York. So that's where, from a capitalist business perspective, it's only going to be easier to kind of grease the wheels for our economy for that to come down from the federal government. It is really important that we get a handle on regulating this technology because it is evolving faster and faster every single day. We were just talking about a video that came up on a timeline. I think we saw the same video. And for a second, it took me a few seconds to realize it took me longer than I would like to admit. You're like disappointed in yourself. I'm disappointed in myself. It took me a little too long before I'm like, there's no way that this is a real, actual topic. The voices were manipulated. And then I thought, oh God, this must be AI. It must be a deep fake. How can voters during campaign seasons know that a video is fake or real when it comes to these AI deepfakes? They can't. That's my biggest concern. However, most of the laws, there are laws in 30 states currently, those laws seem to center around labeling. If they post a deep fake video, they have to say in the small print, by the way, this is a deep fake video. But I mentioned earlier, I'm like, that's like putting like a Surgeon General warning in Braille. Like nobody's going to read the darn thing. Nobody's like their brains are already being impacted by what they saw and thinking that it's real for a second. If it annoys you and gets you angry. That should be an indicator that you should be like, oh, okay, that's a problem. You said something in the first few seconds. Yeah. In the first few seconds, if you're furious, pause. Because they're not only manipulating what that politician is saying, they're manipulating the manner with which they are saying it. So, for example, there was a candidate in Texas, a Democratic candidate in Texas whose name is escaping me. I apologize. They, there was a deep fake video of yes. Yeah. Reading his own posts, but there's a way, you know, it's like when I was a little, little boy, my mom would say, it's not what you say, it's how you say it. Right. I mean, the guy sounded and then there's like another voice in the background. You're being manipulated. There was like an AI voice, right? Yeah. I mean, it's like, it's like 101 classical conditioning, which is a bit terrifying. The reality from a political strategy perspective is that it does not definitively move the needle. So from where I sit. You know, we're leveraging this technology for business, online, for marketing. We're not trying to dupe anybody. We're just trying to scale. We're trying to give an executive the ability to do the work of 30 other executives and to be able to do great work in this world and contribute to society. So when you see that politicians have spent, I think it was like $137 million on deep fake videos, that's what they call AI slop. They're just making this difficult for everybody. And I wish it would stop. I really do. They spend a lot of money. This was one of the most expensive midterms. I think it was like half a billion dollars spent in Ohio. Right. And then this is something for New Yorkers, actually, for the tri-state area. The Upper East Side was a guinea pig because in the primary, there are these AI super PACs. And there was about $21 million involved in that primary from those different super PACs. And that was... Again, the politician that it's going through is, I'm losing the name. Point being, after that, okay, after they're raising this money, they're dedicating the money to these different elections, what the research is telling these AI companies is that middle America does not want to hear from a candidate that is leading with. AI, because there's just this perception that it's bad, that it's like, whoa, pump the brakes, people. So it is interesting that AI super PACs tried to get involved. There was some money involved and then the money went quiet. It's interesting how we perceive AI here in America versus different countries like in China. They treat AI in a totally different way than what What we're doing here in America. And about eight hours ago, NVIDIA came out saying that they are going to reduce their spending in China and reduce the access that China has to their technology. Why is that? Because it's a matter of national security. Yeah. This is a big deal. Yeah, it's far-reaching, and that's why AI is on the ballot. Yeah. That's why. It's not an actual bill, but who you vote in is going to have big repercussions on this technology that affects us. And look, as a New Yorker... What New York has done thus far is fantastic. OK, so basically what we're doing is we're passing laws, even though I'm like, gee, I want it to come from the federal government. Federal government isn't moving fast enough to say this is the law. This is what we're going to do. So instead, they're now suing some of the states that are going against what their intent is. I mean, it's kind of like the only analogy that I could come up with is it's as if. The fire department is suing you if you're trying to put out a fire of your own home with a fire hose. Like, it's bizarre. Yeah. So this is a situation, it's unprecedented. We've never experienced anything like this that is so pervasive. This is not just, you know, impacting financial markets. It's not just impacting national security. It's not just impacting, it's impacting everything. Yeah. And this is not a red issue or a blue issue. No, fully bipartisan. I mean, it's like 80% of Americans polled believe that everything should slow down. Like that seems to be the global sentiment is like, let's pump the brakes. Figure out how to leverage this technology properly, figure out what the regulatory compliance issues should be, and then we can proceed. But can it be stopped? I don't think so. That's what I'm saying. It just seems like it's a little bit too late. It's a little too late now. Yeah, but we can wrap management around it, okay? So it's not too late to implement laws, to have rational discussions around these laws. I'm very encouraged that it is a bipartisan issue. I think, you know, the... Seventy percent of people saying poll that said that they wanted to be further regulated was originally about 44 percent of Republicans, which is now like 54 percent of Republicans, where typically, you know, the Republican vote is, you know, less regulation for for capitalism and for growth. Everybody's realizing this needs to be regulated. Yeah. We've got some good questions I'm seeing in the live chat. So I just want to pull one out here. In what ways should schools integrate AI to support both teachers and students? It is interesting because New York City, when they rolled out school for a while, did not know how they were going to address AI and then decided to put a pause on it for a year so they could figure it out. What is your take on that? Ah. I don't think administrators from education have the time to say we're going to hit the pause button. They need to come out with a stance is really my position. Like you have to communicate to parents and say, this is the manner with which we're doing things and then go ahead and do it. You know, I think my example was my daughter's a freshman at Bucknell University. It's the first thing during the convocation that the president addressed. They led with AI really. And yeah. And so it's not out of the realm of possibility for smart people to get into a room and say, okay, This is the way it's going to be. Let's allow it for this situation and not that situation. My biggest concern, and I think I mentioned it on air the first time, is for individuals that might not have, you know, two parents at home. You know, if a parent's, we have latchkey kids, like, you know, both of my parents worked and my mother was a career school teacher. I would come home to my grandparents. I was lucky, but not everybody's that lucky. So what's going to stop a child from using ChatGPT to do every homework assignment? Yeah. They need to have some sort of rails to run on. They should. It shouldn't be so on the defense. They need to be ahead of it. Yes. And that's how I felt. Okay. We got to sneak in a quick break. When we're back, we're going to talk about the really sexy topic of data centers. Of course. Yeah. I can't wait. I'm so excited. All right. Stay with us. We've got a lot more after the break. All right, welcome back, everyone. Talking with Natasha Live, my guest, Chris Dessy. Great to have you and talk all things AI. It's such a fascinating conversation and one that... It's embedded into our daily lives, and today is election day, and AI is not on the ballot, but it is in the back of our minds, and it is very important that we pay attention to the speed at which this technology is progressing. So we're talking about this, we talked about a social media bubble, and it just felt like you had Instagram and TikTok and likes and the algorithm so much, it felt like the social media bubble is going to burst. Do you think that at some point, maybe when is that, that the AI bubble will burst? So I'm not sure about an AI bubble bursting. I think it's just going to continue to evolve. What I'm seeing is executives that are doing real work beyond just, oh, let's play with this chatbot, are holding human interaction and human engagement in higher regard than I had previously seen. Meaning, previously meaning when AI first came out. When everybody, AI is taking all the jobs. We don't need human beings. Then we had AI slop where everybody was using AI in order to create marketing material and sales material. And then we're like, we need some human beings in this. And then there's a tell, like you'll read something and you're like, oh, AI definitely wrote this. There's no way I know this person and I know how they write. This is not them. I should earlier, I should have mentioned my undergraduate degree is in psychology and there's a psychologist, Dr. Klein, who came up with something called a pre-mortem for businesses. And you get in a room, you have a project, everybody's excited about the project and most large businesses will lie to themselves. They get addicted to the success. They think that they're super smart. They know better than anybody else. And this is great. And it's a fatal flaw. My dear friend, David Kidder actually has a company called pre-mortem and they do this for fortune one hundreds. I do it on a much smaller level, but we get in and we say, what could kill this project? It's six months ahead project that we all love is blown up. What happens? What are the outside factors? And you start from there and you just kind of remove your brain from we're plowing down this one path and we're going to implement that. That's the way that I do it when I, and then I'll start, I say, we, the only way to win with AI is to delete it. If you're running a business, right? You start with deleting it. You run a pre-mortem, you figure out where the chinks in the armor are and you go, should we fill that with AI or should we just put a human being there? What are the ways that we can solve for that Those business owners. Those are the people that are winning right now. They're not just saying, oh, okay, everybody use AI. It's let's examine the business problems and then let's implement intelligent AI throughout the organization. And so that's why you think that this bubble is not going to burst. I don't think so because like, well, I mean, from where I'm from a business perspective, that's what I'm saying. I think the smart, the smart business owners are absolutely implementing AI from top to bottom throughout their organizations. You have to, it, it makes workflow so much better. Yeah. And think about a small business owner can punch above their weight. Like if you're a solopreneur and you like, you can service. probably three to four times more clients than you've ever serviced before and do it at an exceptionally high level. Like this is, that's not bad work. That's smart work. It's not about working hard. It's about working smart. That's what it is. Yeah. I want to talk about data centers because data, data centers came up as a huge topic in the midterms. And back in July, Governor Hochul decided to pause permitting on any new large scale data center. So you think that's a smart move? Super smart. I think pumping the brakes and being like, whoa, whoa, whoa, whoa, whoa. We don't have any idea how this is going to happen. We don't know how the communities are going to benefit or what the pros and cons are. Are they really going to benefit, get a tax benefit? Who's responsible for paying? Is it the big tech company? Right. Okay. Well, we're going to bring jobs in. Great. Are there people that are going to work at that data center in perpetuity or is it just temporary? And are you going to hire? Local tradespeople, is it going to be, you know, the local construction company that could stimulate an economy? But if you don't understand how many people are going to be in the data center, you're like, great, you're going to build this monstrosity, the size of Costco in our backyard, and we don't know how we're going to benefit from it. That's why I think it's smart. Yeah, it feels like it's happening too soon without getting the full picture. I want to take a step back, though, because when you think of AI, you don't necessarily think of an electricity bill. Yeah. And then these data centers, as big as Costco, what is the connection between AI and electricity? So it's not a direct, you know, because of data centers, my electrical bill went up, right? The past five years, I think in New York, it went up 58%. It's a contributing factor. And this is why we should all be voting today, right? Like people should be watching this and be like, I'm done. I was on the fence about voting. Go vote. If you have any feeling about this in either direction, right? Because if you don't understand where the money is coming from, who's going to be responsible for any violations? Like, are we going to be? penalizing a big tech company for what's happening in New York's backyard or not, right? And this is why I said earlier, I'm like, it needs to be one set of rules from the federal government so that if we know that they mess up, it's not going to be the local municipality that gets the hit. This is absolutely going to impact the taxpayers' purse strings if those laws are not clearly defined. And that's when people are like, oh, okay, yeah. That data center is maybe not such a good idea. I think right now New York has 30 data centers that are like waiting final approval. These things are big. They're as big as Costco. Like I want, and the best is like even people that are like, we got to build them tomorrow, but they don't want them in their backyard. Right. So it's. What are the pros and cons of having them in your backyard? Tax benefits. Yeah. Right. Like, so, you know, anybody that has, you know, I think PepsiCo and, you know, any of those businesses that, you know, MasterCard up in Purchase in Westchester County, you know, those local municipalities get tax benefits by having them in their backyard. But I am certain that they understood the manner with which that money was coming in and going out and which regulatory bodies were governing that. When you literally don't know which regulatory body is going to govern. Something the size of a Costco. Yeah. That's a problem. It's massive. And it takes, AI is powered by a lot of electricity. It needs a lot of power. Can you give a little perspective or some analogy to kind of just wrap our minds around just how much electricity it takes to power AI? Oh gosh. Not, I'm not even going to speculate when it comes to the electricity. Quantify it. A whole lot. Yeah, there's smart kids back there. What are you doing there? It's a whole lot of electricity. That's one thing I'm not going to speculate on. You know, it's a lot. It's a lot, but it has a huge impact. You know, because we joke about it. We make a picture on AI and we're like, oh, we drained a lake doing this. It takes a lot of power. I mean, we as human beings have a definitive impact on God's good earth, so we need to be aware of it. I mean, you know, like anything else, it just needs to be regulated and we need to understand, you know, who's going to get, who's going to be, you know, accountable for any bad actors. Okay. And that's really what happens. And if big tech comes in and says, you got to build 30 of these in your backyard. Well, great. If things go sideways, are you going to foot the the bill or we're going to have to foot the bill. Yeah. We're getting a lot of questions in our live chat. Matt asks, why is there so much resistance to the bricks and shovels that bring us AI data centers? And is the resistance organic or manufactured by our greatest competitors? Good question. I think that resistance is the dynamic that I spoke about earlier, where there are no parameters or laws saying you need to hire local. Crafts people. There's, there's no, there's no governing body that's saying you can't bring stuff in from out of state and just dump it in our backyard. So there are no guidelines. That's really the answer. Yeah. Also asking a client of mine when asked, is all this build out for real and sustainable? He said it's transformative. He uses AI for research. Not sure if that's a question. I'm not sure if that's a question, but I don't know whether it's sustainable or not. But this is why we want to have these laws implemented. This is why I believe the federal government needs to get involved, because, you know, if you if you really look at the worst case scenario. Right. Let's say, for example, there's a bad guy in your state or gal, and they know that they can line their pockets by building. 300 data centers on farmland in middle America, that's a big problem, right? So you think about things that are heavily regulated, like the cannabis trade, right? There are hoops that people need to go through in order to get to the point where they could potentially make money off of this. AI in these data centers absolutely need to go through the same hoops and the checks and balances And what, you know, as a capitalist as well, you really look at it. I worked at a regulatory technology company for a bunch of years that acquired my agency. Regulation stimulates growth. Mm-hmm. It's basically like having good parents. You need to know where out of bounds is in order to thrive, right? You need a bedtime, okay? You need to know not to stay out past a certain hour. You need to be a little bit scared of the guy that bought you the drink and you think you might put something in. Again, if you don't understand those guidelines. Then it's the Wild West. Yeah. So once, you know, you need to be, you need the regulation implemented. I think we're at that point now where we need regulation. Like we were Wild West, I feel, where people were just saying, what can we use AI for? What can it be developed for? And now we have an idea of it and the regulation comes into play here. Okay, we have about a minute left. Time is flying. Um, if you could just very quickly, like, where do you see AI going, um, just in the next year? You know, where do you think we're headed? I think our tiny little human brains cannot wrap itself around where we're going with this technology. Now, is that, is there a touch of trepidation in my voice? Yes. I have young daughters, I've spoken about them before, my oldest is a freshman at Bucknell, my youngest is a sophomore in high school. So as voters out there, as parents, as brothers, sisters, fathers, daughters, family members, as a human being, I think we all need to be paying attention to this and the manner with which we all use it. And it's Could potentially be the best thing that has touched humanity or maybe not. Yeah. I am an optimist and I really think. By virtue of us literally having this discussion, the fact that we're like, okay, we need to spend time on this and have a long-form chat about AI, to me, that's encouraging. If we weren't talking about it, if it was being swept under the carpet, if there were backroom deals happening, then that would be a problem. That would be a problem. Chris, always great to have you. Thanks for having me. Thank you so much. This was so much fun. Where can people follow you? Go to ChristopherDessy.com. Amazing. You've got to follow Chris. Also, hopefully you post your Halloween costume because that was really good too. I was Dracula, yeah. Thanks everyone for joining us. Until next time, see you later. See ya.