EP 414: The AI Race is towards who owns the search bar on your phone.

This is the Everyday AI show, the everyday podcast where we simplify AI and bring its power to your fingertips. Listen daily for practical advice to boost your career, business and everyday life.

Are large language models a commodity? Right. They're updated like a trillion times a day. Every time you look, someone has a new large language model, there's a new company, new new reasoning models. Does it all matter? Or does it maybe come down to the hardware? Or even more specifically, the search bar where you start your query for what you ultimately want? We're going to be answering some of those questions and talking about this a little bit more today on Everyday AI. What's going on, y'? All? My name is Jordan Wilson and I'm the host and Everyday AI, this thing is for you. It's your daily live stream, podcast and free daily newsletter helping us all not just keep up with the trends and the topics that are going on in the world of AI, but how we can all actually use them and leverage that information to grow our companies and to grow our careers. Is that what you're doing? If so, awesome, you're in the right spot. So if you haven't already, please go to our website, your Everyday AI dot com. Sign up for the free daily newsletter. Here's what happens inside. We recap each day's podcast interview, pulling out important insights to help you actually leverage what you're learning on today's show as well as we keep you up with everything else happening in the world of AI. So again, your everydayai.com and if you are joining us and you want to know what's happening right now in AI news, technically we got a pre recorded show debuting live. We still have today's AI news in the newsletter, so make sure you go check that out. All right. I am excited today to talk about the AI race and it's actually towards who owns the search bar on your phone. Might not just be about, you know, who can update the model and get the highest benchmarks on the LMS leaderboard. It might not be about that. It might be more hardware or the search bar. All right, don't worry, you don't got to listen to me ramble on about this. We have a guest for today, so I'm excited to chat with Andrew Amen who is the CEO at 923AI Studio. Andrew, thank you so much for joining the Everyday AI Show.

Thanks for having me, Jordan. Pleasure to be here.

All right, I'm ready to geek out about AI, but before we do, Andrew, tell everyone real quick about what it is Y' all do at 923AI Studio?

Sure. So we're an AI development agency. We've been around for 12 years, and for the last eight years we've been building AI products. Now, that doesn't mean we've been building Gen AI products. We've been building machine learning time series things that are not generative AI, aka ChatGPT. But in the last two years now, we've done about two dozen projects around generative AI or ChatGPT. But our experience entails machine learning and gen AI, whatever it is. It's a lengthy project of incorporating somebody's problems or workflow inefficiencies and solving for them and creating an AI solution for them and then selling it back to them. Well, basically they're buying it, but the whole idea is that we transfer it back to their team so that their engineers can use it and gain those efficiencies for the rest of the time that that company's in business.

Yeah, and I'm always curious, Andrew, especially those that have been working in the AI space since before the generative AI wave that we're all, you know, being sucked under right now. You know, what's, what's been the biggest change that you've noticed in terms of general appetite. Right. So, yes, you know, AI is democratized now with large language models, but, you know, you noticed anything else, you know, having built in the space, in the artificial intelligence space before this big wave?

Yeah. I mean, the first one that's very clear and obvious is saying the word AI does not mean gen AI. So back then we could use AI interchangeably with ML or NLP or any of those other subsystems. But now people just assume that it's ChatGPT, which kind of stinks for us because we have to retrain them. I would also say the hardest thing today is I'm talking to five or six CEOs that are in the 100 million to billion dollar range of running their companies. They're very confused about what to do. They are receiving way too much information, way too many people on Twitter or friends that they've had tell them to download the latest, hottest AI tool, use it, and how much it's saving in their business. And they don't know where to start, they don't know how to use it, they don't know who to train in their business. And so they're coming to us thinking that we can solve all their problems without even having a baseline of, like, where to start, and thinking that, like, we're the Magic wizard that can just say here's AI and you're done. You just solved your problems. So it's just like the education gap that exists now is so much farther than it was back then. When somebody's like, hey, I got a product, I'm selling it on an E Comm site. Tell me which product I should give to this person because they're in this cohort and use some machine learning to figure it out. And that would be way easier than I have a 300 million dollar business. Can you install AI somewhere?

That's, that's such a good point, Andrew. Don't even me started. I'm gonna go on a quick like 10 second rant here. Hey, if that's you CEOs who are just seeing things on Twitter and, and telling people like Andrew, oh, go infuse AI. Number one, that's, that's not how it works. Number two, most of those posts you see on social media about this hot new tool, they're actually ads that those people are not disclosing. Sorry, that's. That, that's like a, something that grinds my gears when people are, you know, breaking FTC rules and you know, promoting all these tools and in these pods and these groups and you think it's the next big thing and it's. No, it's a trash product anyways. Had to get that rant out of the way. That's all correct.

Rant.

Yeah, yeah, let's, let's, let's jump to the end here Andrew, and talk about this AI race. And is it actually going to be the search bar? What does that even mean? The AI race going to the search bar?

Yeah. If you work backwards from where we're heading and you try to piece together why people are racing to acquire as much information as possible, AKA like Google and Apple, they're all crawling the Internet now and trying to carve out their corner of the data that they can own. And you're also seeing companies like CNET having a very tough time surviving because all of their information has been crawled and they're trying to figure out a place to exist because they're no longer an isolated website that you need to go to. You can just use an AI overview. But if you work backwards from like why is everyone doing this and what is from five years from now, what is it going to look like on everybody's phone and everybody's laptops? Well, there's two very obvious things that I think we all wish we had immediately once we hear about AI and that is a personal know it all assistant that only knows about me that is very private, that is very secure, that can do tasks that I assign it to that, that I say, go do this, go do this, go do this. And that AI assistant goes out and does those tasks in a very trustworthy, private and secure manner. If we start there and let's say that happens in 2026 or 2027 and we peel back to today, what is the thing that needs to change to allow that to happen? And if you think of how humans interact with computers, we all have this information stored in our brain of where our information lives across the Internet. For me specifically, I call it app stacking. So for me specifically, I use bamboo for my 3D printing. I use bank of America for my banking application. I use Uber when I try to order a taxi cab. I use Airbnb when I shop for a hotel. I also use Chase Bank. So there's this whole series of apps that has my information stored in their databases. Only I, only my brain knows where to get that information from and call that server, retrieve the information and then tell me the answer that I was looking for. And we used a series of apps to do that. And back in the day we used to use website URLs. Well, if I have a smartest assistant sitting next to me and I want that assistant to log into these services to retrieve the information from me so I don't have to, how am I going to do that? The answer is obviously the search bar. I simplify it by saying the keyboard, but the input device to these pieces of hardware. Because my brain knows all of this information and I now need to train my, my, my trusted assistant where all this information lives. So I give it access to the apps, I give it access to the websites, I give it my passwords, I give it my usernames and I then say, can you please order me a taxi cab? Can you take me to the airport, book a flight to Boston, make sure when I land in Boston that I text my wife that I've landed. Then I have to pick up my kids at soccer. Can you please put that on the calendar that I'm going to be 15 minute late if I am. You command these things and then the agent needs to go off and do that. But since I'm doing that, from the search bar, from my input device to my piece of hardware, that search bar is going to have a lot of information about me. We need to think about a future in which one search bar no longer this spread out server farm of all these apps and all these different servers that only my brain knows where it is. This one search bar is going to have to know all of this information to retrieve it and take action.

And this is very real, right? Like if you're hearing what Andrew just said, this isn't some, you know, some obscure topic that just a couple C suite people at Google or Apple are talking about. This is very real, right? So you know, I know that as an example Google did pay Apple, I think it was $22 billion or something like that. It was more than $20 billion in 2022 to be the default search engine. And then you've seen, you know, recent news. Now ChatGPT is is trying to carve out by, you know, releasing, you know, their kind of Chrome extension that takes instead of if you're searching for something where you would normally search to Google, it takes you to straight to ChatGPT. Andrew, how hot is this race right now? Especially with all things going on in AI, how hot is the race for that search bar?

Yeah, and I specifically have asked people at these companies like what are you actually trying to achieve? What is your goal in releasing Chrome extensions? What is your goal in releasing AI overviews? Why, why does someone need to search Perplexity differently than they do Google? And how does Google Ads fit in? There's all these like non primary events happening underneath of why these events are taking place. AI overview specifically is a very interesting Internet challenge I imagine for Google because they paste the AI overview at the top, then they put some sort of Google previous legacy best answer in their own text right underneath it, which is very similar to the AI overview if not stolen exactly from the SEO article. Then beneath that are three sponsored links. Then beneath that are people may ask, then beneath that are Reddit threads. Right. It's very confusing. And so the goal of these companies is trying to land grab data so that they can provide the best answer possible when it's called upon. Well, maybe that is no longer a problem. We've all been able to crawl the Internet, not we all, but like Google, Microsoft, I'm sure Apple's trying to do it behind the scenes. Perplexity is using just Google search to do its special tax tactics. But all the companies, they're trying to achieve this dominance in that they have the answer when somebody asks a question. And to do that, once you get enough knowledge from enough people using your system, you then can start predicting what people will be asking and therefore putting it on their phone before they ask, reducing the latency Decreasing the cost. The knowledge gap between today, which is I have to call a cloud service, large language model that is very unsecure, do my prompting and then retrieve an answer to then bringing that all to the device and having that device be secure, private and have a conversation with the actual apps on my phone, securely and privately. That is the direction we all are heading. And I think even today they're trying to be that first mover. So that ChatGPT is the only one that can call Uber apps. Right, for instance, and if they have that locked up, I can only use ChatGPT to call Uber. I can't use my Google Gemini, which is built into my phone, or I can't use Siri, which is built into my phone to call an Uber. So that's what I think they're racing towards.

Yeah, I still think Siri and Alexa and some of those other, you know, dumb assistants are trying to figure out, you know, one plus one or what today is. But, you know, as, as we look forward to, to a more, you know, knowledgeable future with these kind of, you know, kind of AI assistants that you're talking about here. Andrew, how does this ultimately change, you know, the average business owner? The average individual? Right. Like, how does this actually change our day to day? Because, you know, kind of like what I talked about at the top of the show, are large language models getting commoditized and if so, what does this kind of race to search ultimately mean for all these big tech and AI companies? For the rest of us, I think.

There'S two major things that are very clear that's happening in my view is that the bigger the company they are, the more data they have, you're going to see less competitors if they use AI correctly. There are people that I'm talking to in very specific industries that have two or three competitors in that space. And let's say they all share 33% of the pie. The first one to achieve market dominance with AI and can reduce the revenue per employee. So basically increase revenue while decreasing employees and just creating more profit, they're going to be able to reduce their prices to the point that the other two competitors have to get out of that business in that niche. So I think the bigger the business and the more advantages they have to spend money on AI, the earlier opportunity they're going to have to completely dominate that niche. On the smaller side, I think websites are irrelevant. I've said this for a while, that a restaurant website is the most pointless piece of real estate on the entire Internet. I Never need to go to a website, a restaurant website and look at the menu, look at the pictures, look at what all that is. A restaurant only needs to advertise on Yelp. They need to be on Google Maps and they need to be on maybe Uber Eats and all of those other services in which places can be ordering the food and retrieving that food. So if I need to get a place in Boston, I need to be on the top 10 list of restaurants in Boston. I don't need to have my own website. It's pointless to go to Moo, which is a restaurant here in Boston and just look at pictures of steaks. All I'm there is to look at the menu if they have a reservation and if, if I can get to that reservation, take that one step further. It's completely pointless in an AI world when I shout into my Siri or, or Google voice or any of that and say, give me the top three restaurants in Boston. I'm with my wife, I have two kids, I want to be done by 7 o', clock, make sure I get a reservation that only has an hour and a half. Wait. Like all of those lists of things that I specifically need to give me my personal result of restaurant. I don't need a website to then do that next research. I just need to trust my AI assistant. So I think small businesses need to get out of building UI UX servers, websites and all that and just focus on the experience of where can you live when people are searching for the thing that you're selling.

So to play a little bit of devil's advocate here, and I get what you're saying because as a, as a consumer, as myself, as a small business owner, I very rarely go on websites anymore. It is, it is a mind numbing, hair pulling out of my head experience. It's terrible, right? Compared to the alternative of using a large language model and getting the information you need, you know, a little more efficiently, a little more effectively, but. Right. Someone else might be saying, oh well, you know, especially if they, you know, outside of the restaurant example. Right. Maybe they're a nationwide, you know, I don't know, auto service detailer, whatever, I don't know. Right. But you want your information that be gobbled up by these large language models. Right. So are you maybe suggesting, you know, business owners pay less attention to getting their presence out on the Internet because ultimately what's on their website isn't going to matter as much anymore? Even though that is the content feeding all of these big, you know, AI.

Conglomerates Yeah, I don't know many small businesses that won at the SEO game. Marketers have ruined SEO for the last 10 years by creating top 10 listicles, blogs and all of these other words that are just made up words of why you should rank in SEO. So Joe's Pizza Shop on the corner of Boston is not going to all of a sudden win at the SEO game and beat out these crazy pizza places that have been around for 50 years and have SEO results and Boston Globe backlinks and all those other things that exist. So any, even the detailing shop you gave any of those SEO plays I think are just completely gone. For a small business that just has a domain authority of less than 30 or 40, the play they should be focusing on is what is the long tail keyword that people are searching best pizza in Boston for a family of four. If you have that entire sentence and you give up on best pizza and you give up on pizza shop in Boston and you give up on all the other ones that people are ranking for and you focus on best pizza shop for family of four and you go on Reddit and you go on all the places with the communities and you talk about being the best pizza place for a family of four and why you're there, and you focus on reviews and have those people go on Google reviews and Reddit and Facebook at places that Google can crawl those. So Facebook doesn't really count, but Reddit does and Twitter does and you keep talking about best for family of four. You should show up in the AI overview when somebody asks in their chat assistant, I'm driving to Boston, I need a place. Right? That is the the effort you should be putting in now and focusing on what your restaurant's truly good at, why people come to you, and having conversations about that on the Internet where Google can see it.

Are you still running in circles trying to figure out how to actually grow your business with AI? Maybe your company has been tinkering with large language models for a year or more, but can't really get traction to find ROI on Gen AI. Hey, this is Jordan Wilson, host of this very podcast. Companies like Adobe, Microsoft and Nvidia have partnered with us because they trust our expertise in educating the masses around generative AI to get ahead. And some of the most innovative companies in the country hire us to help with their AI strategy and to train hundreds of their employees on how to use Gen AI. So whether you're looking for ChatGPT training for thousands or just need help building your front end AI strategy, you can Partner with us too. Just like some of the biggest companies in the world do. Go to your everyday AI.com partner to get in contact with our team, or you can just click on the partner section of our website. We'll help you stop running in those AI circles and help get your team ahead and build a straight path to ROI on gen AI.

Hmm.

So that's, that's a great point. Right? It's. Where do ultimately people trust? And I think what you said there, you know, Andrew, makes sense, right? Like, yeah, Reddit has become the new search engine or TikTok or, you know, whatever, whatever genre you might choose. I guess it depends on how old you are as well. But, you know, if we look back at this kind of rush toward the search bar, but also toward hardware, right? So, you know, Perplexity, as an example, was recently confirmed to be working on a hardware device. We've heard now for nearly a year how OpenAI is reportedly doing the same. If the rush is toward the search bar, why are companies, you know, these AI companies may be playing the long game with hardware. And how does that all add up?

Google has everything that ChatGPT has internally. If they could figure out their marketing and they could figure out how to deploy this properly. I'm not positive that Gemini is a worse experience than ChatGPT. Now there are benchmarks and people can show me a quick prompt that says, Look, ChatGPT outperforms this every single day and look at the billions of parameters it was trained on to do this. I'm sure Google can also compete with that if you give it enough time. So the commodity being the large language model is not necessarily what they're competing on right now. The hardware that Google can release, and I have the new Google Pixel 9 and I have Gemini built into it. I've never used a voice assistant before, up until this day. I mean, every once in a while and I just shouted Google. And now it's responding to me. But like, every once in a while I'd be like, set a timer. You know, add this to my to do list. Those, those simple commands that I would do that it would definitely. I would know that I'm not wasting my time and need to pick up my phone anyways to start typing. But the other night I was sitting with my daughter and she asked about planets. And I said, I can explain to you that there's eight planets instead of nine. And she was asking about the moons of the planets. And I'm like, we're getting deep. Why don't we Have a conversation with Gemini, and I popped it open, and it was a natural conversation between the three of us of asking about planets, learning about planets, interrupting about the story that it went on and started telling. And we're like, all right, that's enough. Like, we know about Pluto not being a planet anymore. Let's focus on Neptune. And that natural conversation is now the experience that I'm having with a piece of hardware that so happened has a large language model built into it. If Chachi PT doesn't compete with the hardware side of things, there is no reason I'm ever going to log into ChatGPT. Because Gemini, what I'm paying for $20 a month, plus, I get this built into my plan right now with Google, it's like having a service plan. And I truly believe what will happen is that these tokens that we're all paying for right now are going to be agent tokens or agent currency to do tasks for me. And only the hardware can service and control those tasks because it's monetary transaction at that point. So let's say I get 100 tokens for the month, and I know that's small, but maybe the model's only doing a penny, right? And so let's say I get 100 tokens for the month, and I'm starting to talk about planets. I use 13 tokens to talk about that planet. That's my bill for the month. And only the hardware and software that's integrated together can do that. And so chatgpt sit standing there out in the cold, being like, you got to go to my website. You got to type in OpenAI.com or chatgpt.com, then you got to search, then you got to log in. Oh, and if you're on the browser, I forgot that you logged in before. Let's log in again. As before, Google and Apple are going to have a button on the side of their phone to start having that conversation. And being able to do things like privacy controls, logging into apps. For me, figuring out how those apps remember previous requests and making sure that those requests are not like stupid voice bots like Siri is all the time, right? So all of that's built into the hardware. And I really think that if Chachi BT does not come out with a phone, they're going to be left in the dark of, like, wishing they had done the phone earlier to compete with Google and Apple.

It's interesting, right, because it makes perfect sense. But also, you know, some of these companies have decades, right? Decades ahead of Everyone else in terms of, you know, hardware and it's not always as easy. Right. But you know, I'm curious, you mentioned something there, Andrew, about security. Right. Because I don't think that there's any one piece of technology or one piece of software that knows more about you and individual as the search bar that you use. Right. I was, I always joke around with my wife because her searches are all over the place. Right. And it's like, oh, if someone could see these searches, it's just, you know, hilarious. But you know, when we talk about this, this kind of pivot or this eventual how we might be using AI in the search bar on the phone, how do we need to be looking at security maybe a little differently or is it just kind of the same thing as, you know, how we've always used a, you know, google.com search?

Yeah. You got to remember that it's actually one step further. So like your Google search results, sure. They're like really funny because it's all things that you ask for. You like wife's trying to figure out like how do I cut carrots with a spoon? And you're like, okay, well why would you ever search that? Because we have knives in the house. Right. So there are funny searches. But Google does not know about your searches inside of let's say Amazon. They don't know about your searches inside of bank of America or maybe customer support chats you've had with your wealth advisor and that might get merged. And that's the scary part of this, is that once you start combining apps that exist independently, where your information is divided and you can't kind of piece it together with just Google to figure out who this person truly is on other services, that search bar is going to start having that information. And so from the security of that standpoint, I'm very worried. Right. Google's not the best trustworthy long term experience with good. I keep saying Google, my phone's going nuts. They don't have the best track record with users on their platform and securing their data. Apple has been touting it, maybe predicting for this day when they can finally release a large language model or use AI correctly anywhere besides the messaging app on Apple like. But once they figure that out, there is going to be a privacy concern with everything being routed through that piece of hardware. I hope Perplexity and Claude and chatgpt come out with their own hardware and software combinations or even just their software combination that the government says to Google, you cannot own that search bar. Perplexity can Come in and you can pay $20 a month or $30 a month, whatever it is, to use perplexity as your voice assistant, your keyboard, your input to interact with the other apps. Because without that control, Google will obviously try to wall garden any outside searches from Google and try to control that because that data is that valuable. And same with Microsoft. I'll take this one step further. Imagine sitting in a workplace with Office 365. Of course Microsoft is going to monitor everything you do, say, talk about, interact with your work computer because the employer owns that data. So how do you get out of that too? Like Microsoft's going to own that experience with that search for work. Right. It's all this like security stuff that really, really is going to change things. And I, I don't see a solution outside the government stepping in and saying you cannot like limit people like they do with the browsers right now.

So Andrew, we've talked about a lot in today's conversation from this kind of shift, shift going from, you know, commoditized large language models to search engines. Right? Or AI smart search answers engines. We've talked about now kind of the pivot for these companies also wanting to get into hardware and the privacy concerns that come along with it all. But as we wrap up, what's the one most important thing that you think people need to know when it comes to this kind of AI race and where it's headed, at least towards search and devices?

Yeah, I would say that it's not moving as fast as people are saying it's moving. You'll hear every single podcast in person speaking that AI is going at a break next speed. The average person is not seeing that the people that are not in tech are not experiencing it. If they're not logging into ChatGPT, they have no idea how to even use it properly. So while it's moving fast for our technologists and the people in technology, we're seeing the shift just like we saw the shift with Bitcoin or you know, any of the Internet when it happened. There's a bubble that's going to happen in which finally the mass amount of people move to it. But I think for the most part the average person and people listening to your podcast, I listen to a few shows like the average person that is looking at AI should just learn the basics. How does a voice model work? How does ChatGPT work? How should I be using it? How can it help my daily assignments that I get tasked with and how can I safely use it? And if you Start there. You don't have to be an AI expert. You don't have to, like, solve world hunger by, like, prompting. You just have to know how to use the technology and know that when something gets asked of you, you get under the habit of using AI to solve that so that you can do something else with your life. And that's it. That's it. If you just do that now, if you use Gemini, it doesn't matter. If you use Claude, it doesn't matter if you use ChatGPT, like pick one of them and just get in the habit of using it so that you can get better and better at prompting and working with AI.

Great advice and I think that's a great spot to end. Yeah. Just understand what's going on, right? And hey, maybe I'm biased, but there's no better place to do it than I think right here. So, Andrew, thank you so much for taking time out of your day to join us and help us better understand what's going on in the space here. We really appreciate your time.

Thanks, Jordan, for having me. I appreciate it. It was a pleasure.

And hey, as a reminder, y', all, I said every single day we break down, if you miss something, maybe you're running on the treadmill or walking your dog and your dog went off in the other direction. You're like, what did Andrew say? Don't worry, it's all going to be in our newsletter. So if you haven't already, please go to your everydayai.com sign up for that free daily newsletter and we'll see you back tomorrow and every day for more Everyday AI. Thanks, y'.

All. Thanks.

And that's a wrap for today's edition of Everyday AI. Thanks for joining us. If you enjoyed this episode, please subscribe and leave us a rating. It helps keep us going for a little more AI magic. Visit youreverydayai.com and sign up to our daily newsletter so you don't get left behind. Go break some barriers and we'll see you next time.