What’s happening with AI and Search? [AI Today Podcast] AI transcript and summary - episode of podcast AI Today Podcast: Artificial Intelligence Insights, Experts, and Opinion

Go to PodExtra AI's episode page (What’s happening with AI and Search? [AI Today Podcast]) to play and view complete AI-processed content: summary, mindmap, topics, takeaways, transcript, keywords and highlights.

Go to PodExtra AI's podcast page (AI Today Podcast: Artificial Intelligence Insights, Experts, and Opinion) to view the AI-processed content of all episodes of this podcast.

View full AI transcripts and summaries of all podcast episodes on the blog: AI Today Podcast: Artificial Intelligence Insights, Experts, and Opinion

Episode: What’s happening with AI and Search? [AI Today Podcast]

Author: AI & Data Today
Duration: 00:11:24

Episode Shownotes

If you’re ever done a web search you know that getting relevant and appropriate answers on the web can take a lot of effort. Oftentimes, it also requires multiple attempts to get relevant results. In this episode of AI Today hosts Kathleen Walch and Ron Schmelzer discuss the changing landscape

of AI and search. How AI is changing search Already, AI is proving to be a massive enhancement and augmentation to the everyday work and tasks of businesses, enterprises, and organizations of all types. Continue reading What’s happening with AI and Search? [AI Today Podcast] at Cognilytica.

Full Transcript

00:00:01 Speaker_00
The AI Today podcast, produced by Cognolitica, cuts through the hype and noise to identify what is really happening now in the world of artificial intelligence.

00:00:10 Speaker_00
Learn about emerging AI trends, technologies, and use cases from Cognolitica analysts and guest experts.

00:00:22 Speaker_02
Hello and welcome to the AI Today podcast. I'm your host, Kathleen Mulch.

00:00:26 Speaker_01
And I'm your host, Ron Schmelzer.

00:00:28 Speaker_01
You know, the interesting thing about AI, especially now that it's got everybody's attention, and to be honest, it really is like almost as popular as computers were when they first came out, personal computers and the internet.

00:00:40 Speaker_01
Remember when it was that big wave, everybody just learning what the web was and email. We're at that point with AI where AI is everywhere, and some people are not as engaged with it than others, but it feels like inevitability.

00:00:55 Speaker_01
It's kind of funny for those of you that have been through these previous waves. When personal computers came out in the 80s,

00:01:02 Speaker_01
There were a lot of people who were really interested in it, but there were also a lot of people who were not interested in personal computers. They didn't want to use them. They felt complicated. They felt they looked bulky.

00:01:12 Speaker_01
Remember the old IBMs and the Commodore 64s, TRS-80s, all of those, right? They weren't meant for the average person. Of course, Apple and their vision, they wanted it to be for the average person, but it still wasn't. It took a while.

00:01:26 Speaker_01
And, you know, if you remember the early 80s, there were people who were like really against computers. They thought computers were evil. Like i don't like talking to working with computers and the same thing happened with the internet.

00:01:39 Speaker_01
In the nineties when the internet first came out in the web and websites looked terrible you go back and go back and look at some vintage websites.

00:01:48 Speaker_01
who are in experiences very confusing and there are people at the time in the 90s who are like really enthusiastic about the web and they saw the future they're like everybody's going to doing everything online but there are people who are really against it they're like no i want my call i want to call customer support i want to fax information i want to mail things right

00:02:08 Speaker_01
There's still all those people. And it feels like, right, Kathleen, we're kind of at the same moment with AI, that AI feels like it's everywhere.

00:02:15 Speaker_01
There's a lot of people who really want to engage with it, but there's also a lot of people who really don't.

00:02:19 Speaker_02
Exactly. You know, and it's funny, as you were talking about computers, I was reminded, I recently read an article and it was talking about Lenovo and their ThinkPad and how it's a popular laptop because it hasn't changed much, right?

00:02:31 Speaker_02
And people like familiarity and people get comfortable in a routine and people don't like change. especially at large organizations. And I think that that really plays in to AI today, right?

00:02:44 Speaker_02
Because a lot of people, just like Ron said, with the internet and computers back 30, 40 years ago, some people really saw the vision. Some people are visionaries. They want to move forward. They love trying new things. They're early adopters.

00:02:59 Speaker_02
And other people are more laggards. They want to see how it's being used. They don't wanna be first to market they really need a proven model and a proven product and a proven way of doing things and basically hold their hand the entire time.

00:03:12 Speaker_02
Same is the way with ai and so when i was thinking about that laptop and i'm like yeah you know it hasn't changed in thirty years they said it even still has the little red ball i'm like.

00:03:22 Speaker_02
oh my god who wants that but some people just like it right and they're creatures of habit and they really feel comfortable in that.

00:03:28 Speaker_02
Same thing goes when it comes to AI and so it's important to understand that you do have all of these different people that are on different parts of their journey because we talk a lot about how AI is a journey and so you have to understand where everybody is and

00:03:44 Speaker_02
You know, this even goes into sometimes when we talk about those fears and concerns related to AI, because you don't want to diminish anybody's feelings, whether or not they're rational or irrational.

00:03:54 Speaker_02
And we want everybody to move forward and feel comfortable with this.

00:03:58 Speaker_02
So I think that that also goes into play with kind of, you know, where we are now and how AI is being used, how it's being adopted, how it's being embedded in everything, whether or not you may want it to.

00:04:08 Speaker_02
and also with generative AI, how it's being used, how organizations are adopting it, how they're putting policies together, and then what that means going forward for things like search.

00:04:19 Speaker_01
Right. And that's actually why I started this conversation this way, rather than just starting to talk about search, because the way that we interacted with the web for the last 20, 30 years, oh my goodness, probably 30 years, right?

00:04:32 Speaker_01
Google started in 1998, so they're like 25 years old, quarter of a century old. It's kind of weird when you say 25, it doesn't feel like a lot of years, but a quarter of a century feels like a lot.

00:04:41 Speaker_01
But they were based on the idea that there were all these websites and they were everywhere and they were hard to find, which is why you build a search engine to look for the information.

00:04:52 Speaker_01
And what Google built was basically a really big indexing engine that you could type in a search, and it would try to match what you were searching with the website that would provide that information. That was the original vision, and it still is.

00:05:09 Speaker_01
When most people think about trying to find information, They think about searching. And really what you're doing is it's trying to find a match with what you're searching to another website. It's not going to give you the answer.

00:05:20 Speaker_01
It's just that's what it was built for, right? Of course, well, Google realized you can't make money that way. No one's going to pay to search. I'm not paying 25 cents a search.

00:05:30 Speaker_01
Some people tried it in early days, realized that didn't really work to pay like a subscription fee to search. And you can't make it free because that's not a business.

00:05:39 Speaker_01
people realize advertising, which is that, hey, you may be searching for something.

00:05:43 Speaker_01
So maybe I can show you someone will pay to be listed at the top of that search, whether or not you are actually interested in the thing there, because some percentage of people will click on it.

00:05:56 Speaker_01
That's the whole idea of the click through rate and the cost per click and the Bob and all that sort of stuff. But that whole multi multi billion trillion maybe industry came about that. But things are changing, right?

00:06:08 Speaker_01
That's sort of like, that's that old wave. And actually, if you had asked people in the 1990s about how to find a phone number for something, they wouldn't have thought the web.

00:06:15 Speaker_01
They would have thought the Yellow Pages or 411, for those of you who know, right? The Colony Directory Service. I like the phone number four, right? But if you don't know the name of the company or whatever, you'd look at the yellow pages.

00:06:28 Speaker_01
I need a plumber. I need a whatever, right? It took a while, but people kind of now became this. Now, AI is changing everything because AI is not built. Nobody built like a database per se of like websites and you're searching for it.

00:06:45 Speaker_01
Instead, it's trying to give you an answer to your question. And this is basically breaking the fundamental idea of search. Because people, when they're searching, a lot of times they're looking for an answer for something.

00:06:58 Speaker_01
They don't really want to go read a website, right?

00:07:01 Speaker_01
And so, a lot of people have said that this model, especially the conversational stuff, the conversational pattern of AI, the generative AI stuff, really is fundamentally changing the nature of the web and search.

00:07:15 Speaker_01
And well, I think OpenAI has realized this, right? And they're leaning into it.

00:07:21 Speaker_02
Exactly. And if you think about it, we continue to evolve with the way that we as people engage with technology. So at first, you would be able to search and you would type in, it would pick up on keywords. People didn't really maybe type long phrases.

00:07:37 Speaker_02
They weren't getting the results that they wanted. And it was good for a while. But now, especially because of generative AI, people want to have that conversation and be able to engage and be able to dig a little bit deeper.

00:07:50 Speaker_02
I may be typing in something in search, and it's for camping gear. And then it gives me a whole list of different camping gear. And I may go, OK, well, what's the best sleeping bag for camping in the mountains?

00:08:03 Speaker_02
Or what's the best tent that can fit a family of four comfortably? Or what is the best tent to use in summer months versus winter months? And search doesn't really give me that dynamic and that conversational back and forth that I'm looking for.

00:08:18 Speaker_02
So enter now large language models. And what can I do with that? Where I'm able to put in a prompt and engage with this back and forth conversation, depending on how in-depth or how much detail I want to provide in the prompt.

00:08:35 Speaker_02
may depend on the results that I get back, but I can also ask further questions so that if I say, what's the best sleeping bag, it can, you know, give me some results and I'll say, well, I actually meant what's the best sleeping bag for a child or what's the best sleeping bag for guy that's 6'5".

00:08:52 Speaker_02
You know, I mean, those are going to be very different sleeping bags and different results. I can't have that kind of dynamic search right now when it comes to these non AI search tools. And I think that that's really what people are looking for.

00:09:03 Speaker_02
And so we're saying, how is this going to evolve and we've already started seeing it. And like Ron said, open AI is now

00:09:10 Speaker_02
you know, piloting a new type of search and with GPT and, you know, saying how are people really engaging and what's going to happen.

00:09:20 Speaker_02
And so you can ask follow-up questions and you can have that conversation and maybe you can provide some additional context for each one of your queries so that you do get more of those relevant results without having to do hours of research like you had to do in the past.

00:09:35 Speaker_01
Yeah, so talking a little bit, so this prototype, that's what they're calling it right now, OpenAI, SearchGPT, a temporary prototype.

00:09:42 Speaker_01
They're being very cautious with this wording because I think they realize how disruptive this could be to the whole industry that is Search and to a whole company that is built around Search, right? I think we all know who we're talking about here.

00:09:53 Speaker_01
Google, I should just say it because we all know who we're talking about here, right? And, you know, if you look, if you go to their website and you go to OpenAI's site and you look up the SearchGPT prototype, we'll link to it into the search.

00:10:03 Speaker_01
Now, you'll see It's people asking questions about things. They show an example of, you know, show me the best time to see something in the bay, or what's the temperature going to be like, etc, etc. And the system's, as you would expect, it's prompt.

00:10:17 Speaker_01
So it's going to provide answers to your question.

00:10:19 Speaker_01
Now, of course, the usual issue, the traditional problem with LLMs and systems built on top of them is that they're only as good as the data they're trained on, and they're only as good as the data has recent.

00:10:32 Speaker_01
The usual pushback has been like, well, it might be good for search, but only for things that are old enough that they end up in the LLM. It's not going to be good for things that change in a minute-by-minute basis.

00:10:42 Speaker_01
And companies like Google can say, yeah, you can search for ticket times for a Taylor Swift concert. And you type it into Google, and you'll get a list, or you get a website, a link to a website that has up-to-date information.

00:10:53 Speaker_01
And that'll always give you the most recent, because you can update that website minute-by-minute.

00:10:58 Speaker_01
And i'll provide it and it's harder for lm to do that because you can't you constantly shoving data into the lm is hard it's a hard problem but it seems like you know what the combination of agents and the combination of live web browsing and the combination.

00:11:13 Speaker_01
of systems, GPT is like, well, it's not that hard of a problem because if the system thinks that it's information that might have changed, I'm just going to go find that information on the web for you.

00:11:25 Speaker_01
Because that's really the difference between the search style. Well, I would say the LLM style of searching and the big database of websites version of searching. And the LLM, it's going to try to just give you the answer.

00:11:38 Speaker_01
It's going to say, here is the answer. I might say the source of where I found the answer It's not clear if it's doing that, but if I just tell you where it found the answer.

00:11:47 Speaker_01
But the alternative is that this is, I think, the problem with search is that the alternative is you as a human have to go and open up a bunch of websites. Maybe you have 20 tabs and you're like, I'm looking for some answer to some problem.

00:11:59 Speaker_01
Maybe you want to buy something in the case of sleeping beds, and it might be something else. It might be like, I'm looking for advice on places I should visit when I'm in Rome, right?

00:12:08 Speaker_01
well you're not going to so you might it might one might take a trip advisor one might take you to yelp or one might take you to some personal blog who knows right and it's up to you as the human to discern what is what and that's how google builds a business because some people pay to to be the first eyeballs and and they monetize those eyeballs in a variety of ways you know and the whole search gpt approach lm breaks that whole

00:12:37 Speaker_01
monetization method, because there's no ads, at least so far, right? You know, who knows how things might change. And I think this is a huge threat for organizations that have sort of built their business around anything related to search, right?

00:12:51 Speaker_02
Absolutely. So, I mean, obviously, you think about the big search giants, it's a threat for. But then everybody that spent a lot of

00:12:58 Speaker_02
time, money, and resources for SEO optimization, and to play those ranking games, and be listed first, and make sure that they're matching on the key terms that they want to be matching on, and how do they continuously update their content and refresh things so that they stay at the forefront of that.

00:13:19 Speaker_02
And this definitely is going to change that. Time will tell how it changes it, and time will also tell how these new

00:13:27 Speaker_02
search companies like OpenAI, but I'm sure others will follow suit, are going to be showing and listing and kind of, you know, doing their own rankings.

00:13:36 Speaker_02
Because when we are engaging with these large language models, we do want a result, and we don't want it to just basically pull up

00:13:44 Speaker_02
you know, a search result page and then continue to have the human do it because that's really not providing that benefit, nor is it providing that back and forth, you know, conversation that we're really hoping for when we use large language models.

00:13:57 Speaker_02
So, from that perspective, it will be interesting to see how this evolves and what that's going to do for search in general and how people continue to use and engage with it. So it's exciting.

00:14:09 Speaker_02
We've also seen, you know, a lot of times people will have a their default page, maybe be Google. And now we're starting to see people have their default page be something else.

00:14:21 Speaker_00
Whatever.

00:14:21 Speaker_02
Yeah, I was gonna say whatever large language model you choose. So some people are making it chat GPT. Some people make it Claude. Some people make it something else. And it's interesting that that behavior is starting to change.

00:14:34 Speaker_02
We're going to dig deeper into these topics in additional podcasts. So definitely subscribe to AI Today if you haven't done so already. Because these are really, obviously, exciting times for us.

00:14:43 Speaker_02
We talk about AI Today and how it's changing, and how people are really leaning into this technology, and maybe how others aren't. And it is a practice. So like with anything, change takes time.

00:14:57 Speaker_02
going to chat GPT may not always be your first instinct when you're looking for something and, you know, doing search, but eventually it will get there.

00:15:06 Speaker_02
And so when people do things that make it more intuitive for them and easy for them, how is that going to continue to change the game?

00:15:14 Speaker_01
Yeah, it's definitely changing. I would say like, that's the thing when we started this whole conversation about how people's patterns are changing, you know, in the 90s, people had to learn about the web.

00:15:23 Speaker_01
And now all of a sudden, it's the default pattern. That pattern is changing. And that means that, you know, yeah, SEO, the search engine optimization probably won't be a thing.

00:15:32 Speaker_01
15 years ago, it might be the LLM version, I guess, like the making sure that whatever you're trying to sell or do or share shows up in a prompt response. And that's tricky.

00:15:43 Speaker_01
i know that's the reason why people are so hot on these lm companies cuz now they're the new gatekeepers are the new sort of you know gardens walled gardens did you need to get into somehow the tricky problem because of people's that i would say another another reason why i think people are so interested in this is because let's be honest the search experience is actually gotten worse.

00:16:02 Speaker_01
over the past decade or so. You know, it's kind of hard to point as to exactly when it started happening.

00:16:07 Speaker_01
I like to say things started getting bad, you know, definitely pre-pandemic, maybe like 2016, 2015, for whatever reason, things started not looking so good.

00:16:15 Speaker_01
But, you know, it's like all of a sudden Google started making their search results harder, worse, you know, promoting things that shouldn't be promoted, you know, and, you know, the website started to get, you know, more clickbaity, making it very hard to get results.

00:16:29 Speaker_01
Google started actually providing answers to your questions this is before the LM where you would try to directly answer your question in the prompt but that actually made the SEO people upset because, google was basically picking the winner, the answer and if you're let's say have a travel site for example and you ask some basic question that's popping up a response from say trip advisor near another company,

00:16:52 Speaker_01
What the hell you know all the money in advertising so we even talked about how to say cathleen i talked about a version of this problem way back in twenty seventeen i think.

00:17:04 Speaker_01
Twenty eighteen we just started talking about voice assistance and sort of one of the challenge of voice assistance is that voice assistance were like the really really early. version of the LLM because they're conversational.

00:17:17 Speaker_01
They weren't large language models, really. They were based on a related but a different AI NLP technology. They were obviously because we didn't have these large language models in 2017.

00:17:27 Speaker_01
But they were able to do voice recognition and you can build things called intents where you tell it what you want to listen to. But we surfaced the same problem back then, which was that Alexa and these devices were basically intermediaries.

00:17:40 Speaker_01
If you said, hey, Alexa, you know, book me an air travel ticket or, you know, tell me when the next Taylor Swift concert is, it's going to pick on your behalf one place and it's going to go do it.

00:17:53 Speaker_01
So the question is, well, what does that mean for everybody who wants to get their content into the voice assistants? It's actually the same problem.

00:18:00 Speaker_01
We thought it was going to be first in the voice assistants, but I think this problem is coming up now much more so on the LLMs, right?

00:18:07 Speaker_02
Yeah, and I think, you know, the voice assistants, I mean, had their own problems and the adoption didn't come as much as people thought, but people want to be able to see it. And so now really is the time, right?

00:18:17 Speaker_02
You know, we all engage different ways with technology and that's something, again, right, we're figuring out as we go along.

00:18:24 Speaker_02
We didn't really want to have that back and forth conversation with the voice assistants and have it do those multi-step bookings.

00:18:32 Speaker_02
I know that Amazon was definitely pushing for that back a few years ago, and they were trying to do that, but it didn't quite take off. Or you could maybe book movie tickets. and an Uber and dinner reservations.

00:18:44 Speaker_02
You can kind of have that whole date night experience booked through your voice assistant. But now with large language models, really what can we do? And now people have gotten comfortable with it. They've been out for long enough.

00:18:55 Speaker_02
People are getting better with their prompts, understanding the value that it can bring. And so how are we going to continue to move forward? And what is this going to do? You know, it's always hard to predict the future, right?

00:19:08 Speaker_02
We can guess, but we're not always right. Because if we were, then we'd be magic. But I know that I'd say definitely look at some stocks if we can predict the future.

00:19:18 Speaker_01
AI today predicting the future.

00:19:21 Speaker_02
Exactly. But it is really incredible. So if you're thinking about this, drop us a note. If you have been having these discussions, if you maybe have one of these large language models set as your homepage, reach out to us. We'd love to hear

00:19:36 Speaker_02
how our listeners are really engaging with this and thinking about the future of search as well. So you can always drop us a note, info at Cognolitica.com. You can go to our website, find us on LinkedIn. Ron and I are both really active on LinkedIn.

00:19:50 Speaker_02
You can find Kathleen Maltz, Ron Schmelzer, or Cognolitica. And also, if you haven't done so already, subscribe to AI Today. I had

00:19:57 Speaker_02
mentioned earlier in the episode that we have some upcoming topics that we're going to be discussing somewhat related to this, and then other additional topics.

00:20:05 Speaker_01
Let's give them a preview. Let's tell them what we're going to be talking about. We'll be talking a little bit about that AI homepage. There's a reason why. I'm not just like, oh, it's interesting that people are doing it, right?

00:20:15 Speaker_02
Yeah, so we also, you know, we want to bring up this idea of what is an AI-first mindset. We talk about what is an AI-first organization, but what now, we're going to break it down one level further and say, what is an AI mindset for the individual?

00:20:28 Speaker_02
And what does that mean? And how are your habits and your patterns and your behaviors going to be changing? Maybe they have, maybe they haven't, but we're definitely going to be digging into that in a very quickly upcoming episode, so stay subscribed.

00:20:41 Speaker_02
We also have some additional soft skills that we're going to be discussing, as well as some AI use case series. So lots of good episodes coming up.

00:20:51 Speaker_01
Stay tuned. And share. Definitely share. We know some people really love the AI Today podcast. We love that we're part of your listening experience. But there are others who are just getting into AI now.

00:21:03 Speaker_01
We have like 470, I don't know, episodes, something like that, over the past seven years. And you can tell, some of these topics keep coming up. So we know that some people have just gotten into AI.

00:21:14 Speaker_01
They're just talking about things as if they've just discovered. And we're like, folks, we've been talking about these problems and these issues.

00:21:21 Speaker_01
ideas for such a long time that maybe they've listened to our episode, the people who are saying new things, and they're using inspiration from the AI Today podcast. So do yourself a favor, get ahead of the curve and listen.

00:21:34 Speaker_01
And if you haven't listened to our episodes from a while ago, You know, there's nothing wrong with the episodes from 2019 and 2020, 2018. People say, oh, it seems like those episodes couldn't possibly be relevant.

00:21:47 Speaker_01
And the crazy thing is that they're more relevant than ever. So binge, listen to 400 plus episodes in the next, I don't know, a couple of weeks.

00:21:54 Speaker_02
Like this episode and want to hear more? With hundreds of episodes and over 3 million downloads, check out more AI Today podcasts at aitoday.live.

00:22:04 Speaker_02
Make sure to subscribe to AI Today if you haven't already on Apple Podcasts, Spotify, Stitcher, Google, Amazon, or your favorite podcast platform. Want to dive deeper and get resources to drive your AI efforts further?

00:22:17 Speaker_02
We've put together a carefully curated collection of resources and tools. Handcrafted for you, our listeners, to expand your knowledge, dive deeper into the world of AI, and provide you with the essential resources you need.

00:22:29 Speaker_02
Check it out at aitoday.live slash list. This sound recording and its contents are copyright by Cognolitica. All rights reserved. Music by Matsu Gravas. As always, thanks for listening to AI Today, and we'll catch you at the next podcast.