So What? What’s next in generative AI?

Well, hey everyone, happy Thursday. Welcome to SOA, the Marketing Analytics and Insights live show. I'm Katie, joined by Chris and John. And without even planning it, we're all wearing our Trust Insight shirts today. So we are what it John, what did you call us the Borgs?

I don't know what to do. I don't know where the T is on. I can't remember the tagline. It's uh resistance is futile. It's futile.

There we go. Be assimilated in the Trust Insights matrix. Well, that aligns nicely with the fact that we are talking about artificial intelligence today. Uh specifically, what's next in generative AI. And so we're going to talk through governance and processes around training the AI models, uh, results you can expect, and what's likely coming for generative AI models in the next few months.

Um, we've been talking about generative AI this week. So if you want to catch a more technical overview of what generative AI is, you can catch that episode at Trust Insights.ai slash uh TI podcast, where Chris and I break down what it is and what it isn't from a technical standpoint. And then in our newsletter this week at TrustInsights.ai slash newsletter, I give a non-technical overview of what generative AI is. And Chris, you go into more of the technical again in the data diary section. Exactly right.

On today's show, we want to talk about not only generative AI, but what's next? Because artificial intelligence is a technology that does not stand still. It is evolving even as we speak. So by the time this episode comes out, uh it'll probably be out of date anyway. So Chris, where would you like to start with generative AI?

I think it makes sense to talk about to start um how you should be using it depending on your needs, right? So because there's there's a series of pathways that go from hey, let's just try this thing out, to this is going to be a core part of our business. And the use cases and the technologies that go with each stage are different. And right now, when you look around at the landscape, when you listen to people in you know Slack groups and Discord servers and discussion forums and stuff like that, you don't see people thinking strategically about the use of generative AI. Everyone is still kind of in the the party trick mode, like, oh, I can make it do this.

Like, yeah, that that's cool. Um, you know, but you're you're not thinking with any level of strategy. So you know, the the very base level, the the entry level that everybody almost everybody knows and is familiar with is good old fashioned chat GPT, right? Here's the web-based prompt. You go in, you type your stuff, have a good time, um, copy paste, people get this, people understand this.

Um, and for I think most entry level use cases, this is the tool that people will use. And I think this is for non-sensitive, non-restricted data, non-private inf data, um, this is a great tool to use. So I would say this is sort of like the the first stepping stone. So if you want to write someone a limerick for their birthday, uh, then this is a good tool to use. Exactly right.

Now the challenges with uh the chat GPT interface and the the models that are in here, the GPT 3.5 turbo and the GPT-4 if you're a paying customer. Um, these are big models, but they're very general models. Um, these these are models that have a lot of everything in them. And so um they're not as specific probably as some people want them to be. And the process of using this tool still requires a human being, right?

And this is kind of the the big challenge with um what I see a lot of people doing with with chat GPT. They're just coming up with cool stuff to do with it, but it's still a person copying and pasting. And so just to sort of step back for a second, uh, just to sort of give like a quick definition of generative uh artificial intelligence. So, Chris, you know, please do correct me if I'm wrong, but you gave me a really good way to remember it, and it's the acronym FOG. So there's three kinds of artificial intelligence find, organize, and generate.

So you have regression, which is find, and so you actually have to give the model something to look at, uh, organize, which is sort of it makes sense of everything you've given it, it puts it into its own classification system that it understands, and then generate, which is generative AI, which is what we're talking about today, and that's where it creates the thing. And so today we're focusing specifically on generative AI, but generative re relies on uh uh regressive regression and uh classification in order to operate. And so when we're looking at a system like chat GPT, all of those things are happening in the background when you're talking about the large learning model. And so the large learning model, if you say write me a limerick, the first thing it has to do is go into its uh library of sources and say, what the heck is a limerick? Do I know what a limerick is?

And then it finds all of its different references to limericks. And if you say write me a limerick about birthdays, then it starts to pick up limericks and birthdays and organize the information, and then it generates the limerick for you that you've asked. Is that all roughly correct? Roughly correct. Yeah, it doesn't actually have any sources, it doesn't have any in the the underlying model is just a series of probabilities.

It's a it's a library of probabilities, so it's all just math, just a big pile of numbers. Um those numbers have associations. And one of the things, and we were talking about this earlier too. It's kind of a fun example. Um, you know, you can ask uh these language models to explain things, you know, like explain to a fifth grader or my favorite, explain it to your dog.

Uh I did one earlier saying explain marketing attribution to your dog. Um and it did a credible job with it. Um, but there's no actual words in the GPT model, it is just numeric probabilities that then get spit out through the decoding process into words. Um so this is the the baseline, right? This is this is now you can do some classification tasks with this with these models.

For example, you can have it do sentiment scoring, give it um uh a piece of information of text and say score this, uh, give this a sentiment score. Now, whether or not you agree with it is up to debate, you know, about how good do you think it is, although I will say it's substantially better than a lot of older solutions, but you can even do classification with this. But again, we're still stuck with you, the human being, are copying and pasting uh or or typing problems here. That does not scale well. So the next step in our journey for the use of generative AI is talking to these models programmatically.

So this is a piece of code that we wrote. Um that this is actually in the Trust Insights newsletter from a few issues back, um, where we're using the GPT 3.5 turbo model, which is the one that chat GPT uses on the back end, and there's a prompt built right in it. The prompt says you will act as a linguistics expert, blah, blah, blah, blah, blah. Give it a sentiment score from minus 10 to plus 10. And then we feed it um text.

So we feed it URLs from one of our databases that has the text of the different articles, and then it returns a table of sentiment scores. Right. So now instead of you or me, Katie, copying and pasting article after article into chat GPT, we're gonna programmatically have it do this. It's still using the same base model. There's a detailed prompt, but the output now is being programmatically run instead of one at a time by a human.

This is step two. I know I'm ready to spin up R and start writing my own code here. But I think I think that that's, you know, Chris, as you're describing, like, so chat GPT was sort of the step one of like that's the accessible version, but there's limitations to that version in terms of what it can do. And now you're talking about here's step two in terms of taking it beyond this interface, this chat interface. That's right.

We're leaving the the the web interface behind. We're still using the GPT model, the underlying software. We're just now using it programmatically instead of a human being doing it. So for if you have processes in your company, and this is where where governance and and process management really are important. If you have processes, you identify, like, yeah, we do the exact same thing, you know, week after week.

We write the same exact press release or you know, we write the same financials. Imagine if you were a PR firm and you had a client that had to by law had to release a press release stating the results of its financials, right? Is required by regulation FD from the SEC. You would take the raw data, you would feed it into um put it into a table, and then you would give it a press release prompt. And now for your 50 or 100 or 200 clients, just have this thing crank out those releases for you.

You know, it's it's templated, it's exactly the same thing more or less month over month. So just have it do it. And I think that that's as you started, you know, talking about the use cases, those are the use cases uh from a strategic standpoint that companies they're not quite there yet. They're maybe starting to think about it in that case, but there's a long way to go in terms of building it in this way. Exactly.

Um, how does it run as far as API access? So if you have the free account, can you hit the API or do you have to be on the paid version to start making calls? So every account that uses the API has to be have a credit card in the system because you are billed for every usage of the API. Um the billing rate is two tenths of a penny per thousand tokens. So if this thing spits out a thousand word press release, um uh I'm just hypothetical, we would we would have to pay open AI two cents.

Oh, so the and it's since you're paying for every call, it's just completely open access, they don't care how much you burn because exactly made how long do you think it will stay that way in terms of pricing? Oh the prices keep getting cheaper um the previous model GPT three was two cents per thousand tokens this the GPT 3.5 model is two tenths of a cent per thousand tokens now the newest model GPT four that is back to two cents per thousand tokens because it's and and there are different there's differential pricing depending on how lot how much data you're processing because it can process up to 3200 tokens and after I want to say after a certain point it goes to like six cents per like 25,000 tokens or something like that. It's it's still in ridiculously cheap. And do you is that because more and more people are using it so they can keep the prices lower? Yeah because factually once the model is built you just deploy it right yeah they all they don't update these models very often are they the the tur 3.5 turbo model updates about once a month and there have not been any uh declarations about how often GPT 4 has been updated.

And so you know again back to what you were saying about thinking about strategically and integrating it those are the kinds of things that may not be well known in terms of well what is the cost savings if we use AI versus a person well we have to re-edit everything that the AI generates however it's going to generate a hundred things for you to edit at the cost of maybe you know a couple bucks. Exactly exactly so this is step two in the journey. So step one is human being copy pasting into chat GPT. Step two is talking to the GPT model with with uh your your programming language. We're still using the vanilla out-of-the-box GPT 3.5 turbo model, which is very good.

Very, very good, but it's still very generic. So that brings us to sort of the third evolution in our journey towards towards embracing these large language models. And that is OpenAI in particular, and many of these other companies, Hugging Face, Ulithera AI, Carper, etc., allow you to take a model that they have running and you can fine-tune that models to specific purposes. So these are the instructions for you know how you fine-tune one of the GPT three models, uh, Ada Babbage or Curio DaVinci are the four models that are available. And you would basically create training data.

Um, like for example, here's the prompt and here's the completion. You would provide this this data, and then it will open AI will process that data and then give you a special version of its API for you to talk to specifically. So imagine Katie, I took all 50, 60, 70 uh newsletter called opens that you've ever written for the Trust Insights newsletter, and I have the title of those newsletters, and I would uh write the prompt as write a newsletter about you know uh the uses of generative AI. And then in the completion, I would copy and paste your that what you wrote for that issue in into this um file format. And I just have prepare this data.

I would then load this to open AI and say, V-tune the Da Vinci model on this specifically. And what this is going to do is essentially it's gonna learn from your writing style, learn from these prompts and say, okay, now I've got a model that's going to heavily favor the way that Katie Robert writes. And so we would create a uh, you know, GPT-3 uh uh point uh five Katie. And it has its own special model in the open AI ecosystem where it's like now it's talking like much more like you. It's gonna make your uh vanilla ice jokes uh oh, it knows too much about me.

Exactly. And it will it will emphasize much less all the other different writing styles because we're telling it we want you to give extra weight to the types of words and language that Katie specifically uses. This is where you know, if you think about like uh again, uh a marketing agency. Imagine a marketing agency is creating content for a customer. And maybe the customer is like uh our old friends in Citrix Systems.

They have a huge blog, they're like 3,000 blog posts on their blog. You would take all 3,000 blog posts, or maybe the top 10%, you know, the ones that drive the highest traffic, goes into here and saying, okay, now we're gonna train a basically a citrix specific version of GPT 3. And what that does, the benefit of that is that when you now have uh an employee who goes and says, write a blog post about you know uh on some on-premise device, the prompt can be a lot shorter because the model's now been tuned to the specific set of tasks. Like it's it's it's the citrix content engine now. And so you don't have to write four paragraphs like you should sound like this and do this, because that's now built into the model as part of fine-tuning.

That's incredibly exciting and also a little scary. Um, you know, and I don't want to go down the rabbit hole of like the pros and cons because obviously there's a lot of uh risks with doing something like that in terms of misinformation, but um, you know, I'm thinking about, you know, those of us who work really hard on our, you know, creative briefs and standards, and you know, here's what the tone of the company should be. Now you don't have to worry that if you bring on a new writer or a new team member, you have to, you know, double check that they're getting the tone right. Like you're building it into the first draft to say this is the tone. And I mean, that to me is like a huge time saver.

Exactly right. So does it have to create a separate instance of GPT three just for your trained? You do, you have to take the whole thing. And then that JSON file that's there, is that that just run on your machine every time it runs, or is there like a container that has to be set up over on their side that you have to do? So this is all hosted and run by OpenAI.

The whole thing is so you would equal your training data is over there in an instance. So um does this come with your account, or you have to buy another instance that's this is like a different product to make this happen. No, it's the same thing, but the pricing is higher. So instead of two cents per thousand tokens now for uh your own trained models, 12 cents per thousand tokens, right? So it's a six X cost increase.

Yeah, but they do all the hosting. You you don't have to do anything. Like that's that's huge value still. Exactly. So this is sort of step three in the evolution of your your generative AI journey.

Step one is basic proof of concept. You know, can we can we even use it? So you're just sit in there to chat GPT typing. Um, right. Step two is okay, we found some good use cases.

We want to scale using them. So you're now using you know a programming language to programmatically. Step three is like, okay, we want this to be specific to our industry, our customers, et cetera. Let's fine-tune a model, we're gonna run that. Step four in the process uh is when you say, Hey, we've got stuff that we want to do, but we work with sensitive data, healthcare data, finance data, military data, things like that.

We've got information we absolutely cannot under any circumstances send data to a third party, right? You just not allowed, or maybe you're just rightfully a little paranoid about giving giving the crown jewels, or you you say, Hey, you know what? We see the value in large language models so much so that we know we're gonna we're gonna beat the heck out of this thing, and our open AI bill is gonna be like a gazillion dollars because we're gonna be making calls to this thing left, right, and center. The fourth stage in the journey is okay, let's tune and run our own model on our own hardware in-house. And so, this is an example here.

You can see on screen, this is uh a Google Colab notebook. So, this is uh a virtual environment. We can do this, you and you would do this on your machinery, like a good gaming laptop would let you do this. You will download the free open source model from a provider like Ulithora AI, the GPT J6B model, uh six billion parameter model, and then on your own machine with your own data, you will fine-tune that model for its very specific purpose, and you will run that on your own hardware at your company. Um, this is for when you realize that a large language model is part of the secret sauce of your company, and you absolutely positively could possibly cannot let have anyone else have access to it.

Um, very recently, two weeks ago, Bloomberg, the financial services company, announced Bloomberg GPT. What they did was they took the 41 years of proprietary data from their terminal, every stock trade and transaction and inquiry and analyst call and all that stuff, and they fine-tuned a version of this model um to run internally. So now when you use the Bloomberg terminal, you can pop up a little window and say, show me 10 stocks that have had 5% uh Kager over the you know the last five years plus a minimum dividend of nine percent uh or nine nine dollars per share, you know, and and show me those stocks and and it will come up with that analysis for you. Um or show me the Bollinger bands for this stock over this period of time. Can it write Limericks?

Not very well anymore, right? Um it's certainly not gonna be quoting vanilla ice to you, but it is going to do exactly what it's gonna do exactly what they want it to do within that context. And so for the most advanced companies or the companies that have substantial risks if the data goes outside their borders, this is how they're gonna use large language models. And this is sort of the pinnacle of the journey, is you're basically your own AI company now. I'm overwhelmed.

I'm just gonna like I am overwhelmed. Um, you know, because I think I fall into the category of the majority of end users who are really just scratching the surface, and I understand how a system like ChatGPT works. Um, and that's again only a recently I understand it. And there's still, you know, as I was describing my understanding of how these things work, you're saying I was roughly correct. There's still pieces that I don't understand.

So, as you're describing these more sophisticated versions of generative AI, like it's overwhelming for a non-data scientist. And so, I mean, this might be a big loaded question, but I mean, how do we keep up? What is it that we need to understand in terms of what's happening with generative AI and what's next? And that might be a big question. It starts with the first P of the five Ps, which is what is the purpose?

This is this has been true of AI for uh you know decades. Generally speaking, if you're looking using AI, generative or otherwise, just to make things a little bit more efficient, um, you're just probably gonna find a vendor and just you, you know, like an open AI or a Go Charlie or any of these companies, and just use their software and you know, let them deal with the headaches of of all the advanced stuff because all you want to do is just very specific tasks, um, and that and that is great. Um there's you don't have to keep up at that point, you just use the tools um as they're intended to be used. You know, when Microsoft Office rolls out its GPT integration stuff, you can you know say, hey, Microsoft Word, uh, you know, make a uh make a two page summary document of the slide deck, and it will do its thing and you'll you'll enjoy the usage of it. Um if the generative AI is going to be part of the secret sauce of the company, like it's gonna be how your company delivers value, then yeah, you need to dig into the stuff and understand the technology.

So, a real simple example we are um writing code right now. Uh it's actually it's operational, I just haven't um told you yet. Uh you know how much I love surprises, Chris. I know. I and we love giving away surprises on the air in public.

Um what could go wrong to take sets of Google Analytics data, pass it to um the GPT models and say, write an analysis of this data, right? So now instead of us sitting having to sit there and go, okay, you know, what do we see in this client's data? The GPT model will do the first draft, and then we look at the data and we inspect it at the analysis and go, okay, that was a pretty good analysis. So you missed this part here, and so on and so forth. Um that's something that we're gonna protect, right?

We're and eventually we may evolve we may evolve into running our own language model, a fine-tuned model specifically for that purpose because I see it being as part of the secret sauce of what Trust Insights does. And I don't necessarily want to let somebody else you know have access to that. So it in terms of how do you keep up depends on the how much you need to keep up. Makes well, it is before we get back to that question. Uh I will let you know that I haven't even seen what this thing can do yet.

I already have notes for you. I have some thoughts. So we'll we'll come back to that offline. But I mean, but that's a really you know I think that that's not something that non-data scientists or non-technical folks were aware that generative AI could be doing. Um because you know, we think of it as, you know, write me a blog post or write me some social copy, or you know, I actually saw, and this was something that to me was a bit mind-blowing was it can create a spreadsheet if you give it the right parameters.

And I was like, Well, of course it can, but I hadn't even thought that that was a use case for it. And so now if you're saying that one of the use cases is actual data analysis. Like that's again, um, I'm still overwhelmed. You haven't made me less overwhelmed. You haven't made me welmed.

Uh okay. Um, yeah, I mean, the the software can do all sorts of text, right? It is a language model that deals with text and text like objects. So uh let's see if we can make this work here. So what in the world was that?

That was text, right? So that was music can be written out as text. So what I did was I fed the GPT4 model a set of lyrics, I gave it the song genre and said, write the music that accompanies these lyrics. And that's what it generated. Now it generated that as text.

I then had to hand that to a guitarist to actually play it. But that's an example of the kinds of things that these tools can do. If it is text or something like text, meaning is formatted like text, the tools can work on it. What's formatted like text? DNA sequences, right?

RNA sequences. You feed it a genome and say, identify anomalies or develop a candidate vaccine for this novel virus that we've never seen before. Um, there are so many different applications, and one of the most important things that you can do as a even as a beginning user of these tools, is to set up governance and just track what you're doing. So, this is an example. I homeschool, uh we homeschool our kid, and I wrote a prompt to grade papers, right?

So that I can take my kids' reports that they write and grade them. Um it's text, it's just uh you know, processing data. Um another one that's uh I I particularly enjoy is here's a list. Uh you're gonna write a grocery list. Here's the list of meals that I'm planning to cook this week.

What are the probable list of ingredients? Again, we're this is all word association stuff, and it's it extracts it out and it makes me a nice grocery list that I can then go to the store and go, wow, I forgot that I needed half of these things. Um writing cease and desist notices, summarizing. Remember, we talked on a past podcast episode of the four major tasks: summarization, um, generation, uh, extraction, and rewriting. Those are the big things that these models are really good at doing.

Writing code, um, developing answers for market research, doing personality assessment. There's no shortage of text-like things that it can do. So my the my thing I tell you to do that's the most important. Save your prompts, right? And then say them and tune them.

Um and I would encourage you, if you have not already, um, on the Trust Insights website, there is a totally free, no strings attached, uh, download on how to write an effective prompt. It gives you the structure of really good prompt. Go and grab this PDF, pin it up on your office wall. Um, but this this will help you write better prompts. And then once you've got that, you need the governance to store these things so that um A, you can find them when you're trying to remember what you did.

Uh, and B, um, depending on uh the company, you you may want to share them. You know, one of the my favorite ones is uh uh I have I get junk email all the time. And so I have a prompt uh from my assistant Grace Parker Thompson, aka G P T um to write uh go away letters. That's great. And so because you've got that token for that, everyone is different.

So people can't like you know, filter them out or whatever. You've got a unique rejection letter for everyone. Exactly. And you know, it's it's just very, very straightforward stuff. Um but yeah, the models are capable of doing anything that that is text or text like.

So you can put markdown tables of data. That's you know, when we export data out of Google Analytics, it comes in as markdown, it can read it. If you're familiar with the ABC notation of music and text, you can put an ABC notation of entire music scores in uh and have it process them. If you can represent it as text, it can go in and it can come back out. How about HTML too?

That would be another thing. You could do iframes and have all kinds of HTML, CSS, JavaScript. Um I was trying to help solve a problem for a friend yesterday. Um there was this page on this one website, and it was a contest of voting and someone was hacking the site and we're trying to figure out where it was. And I figured out I just fed the back end JavaScript that I didn't read particularly well into chat GPT and said, find the vulnerability uh in this code.

And it turns out that you know they basically did had there was a lot of uh proof and checking up front and on the validation side when the validation email there's no validation whatsoever. So that was how the hackers were were were spamming uh the system. A couple weeks ago, Twitter released its open source um recommendation algorithm, right? And this thing is gigantic. Like there's there's so many thousands of files and stuff in here.

Um I all it's also written in three languages, Scala, Java, and uh Thrift, and which, you know, I don't write any of those languages. What do I do? I find the the keywords in the code that I care about, like you know, retweet or uh something like that. Find that code, then paste it into chat GPT and say, explain this code to me. Like, what does this code do?

What does it mean? Um, and it it did a very good job to the point where I can make recommendations to people now on what to do on Twitter because GPT4 explained it. What all of the, I mean, all of this, again, I keep saying it because I'm just sort of like, I need to process everything you're going over. Um, but I mean, it all reeks of you need processing governance, you need planning. Even if you do nothing else than create a user story of so, you know, let's take the very simple example of you creating a grocery list.

And so, you know, by starting with a user story, so as a you know, busy father who also works full-time, I want to, you know, use this uh generative AI system to help me plan uh a shopping list so that I get in and out of the store really, really fast because I don't have time. And so you're like, okay, so I know what I need to do. What do I need to give the generative AI prompt so that I can get that grocery list? And then that becomes the thing that you start to save over and over again, and then you probably just swap out your meal planning for the week. Um, which sounds really straightforward because it's just you and your family, but to your point, Chris, as you're bringing it into larger organizations, building it into the strategy, you really want to make sure that you have those user stories really tight and concise and clear, so that you can then build that uh repeated code to generate hundreds of first drafts to then farm out to all of your content writers and editors.

Because I remember we used to work with a client who would post like a dozen blog posts a day. Um you know, I don't remember the exact number, but the amount of content was enormous, and wrangling all of the writers and the topics and editing, it was a big undertaking. Um, and so being able to automate some of that using these systems to me was like, it's too bad that didn't exist then, but I hope they're using it now. And here's what I love about user stories. The first two parts of the user story mirror the prompt, right?

As a use who you are, GPT model, I want you to do this thing, right? So that is our part as the humans to make sure it did the thing. But if you look at this product, for example, this is um uh helping uh I needed to help someone write some wedding vows this week. Um, you will act as a secular justice of the peace officiating weddings. You specialize in writing wedding ceremonies, blah, blah, blah, blah, blah.

Your first task is to write some wedding vows for these two women, so and so on and so forth. And it's as a I want to sew that, right? So we were basically taking that user story, kind of uh adapting it, and that is what we're telling the GPT model to do. And so if you if you have written out the user stories and you're super clear about them, guess what? It will generate good results.

If you're like, oh, I just write a limerick about something. Well, well, you're gonna get generic results, right? You're gonna get uh poorly tuned results as opposed to being very clear in your user stories, being very clear in your prompt writing to get very clear results. John, hang it in there. Limerings for everyone.

Limericks for everyone. Now, let's talk in the last few minutes about what's next. What's next? Ooh, I'm scared. I'll scare okay.

What's next? And it exists already, but not in not in broad circulation, is uh the GPT-4 API has is called multimodal capability. Multimodal means it can take in more than one form of input. Right now, today you have to put text in as everything. The new API permits you to load image data into the GPT 4 model and get text back.

So if I were to take a photo of a screenshot right now, I could feed this screenshot and it would be able to interpret. There are three people talking, you know, at Katie Rebere, at John J. Wall at CS Pen. Uh, it might describe our office background stuff, depending on our prompt. Um, these are all things that are embedded in images.

I could take some sheet music, right? I'll take uh an image of some sheet music and load it in and say, interpret this music score and you know, transcribe it and then rescore it from C major to D minor. Um I could take, and this is where for trust insights, we are very, very interested in doing this. Um, I can take could take a screenshot of a report out of your Salesforce instance and then marry that with a screenshot from your Google Analytics instance to say, you know, tell me how well the marketing strategy is working based on the data in these two charts. And so these are all multimodal capabilities.

It's kind of the reverse of image generation, right? So we used to go, you know, a dog on a skateboard where you're a two-tube, ha ha, you know, and it makes a fun picture. Now it's like, okay, here's the picture. Be detailed in explaining it. I would imagine, you know, uh, like one of the projects that we're working on for our own website is making sure that all of the images uh have the alt text.

I would imagine that using something like that to help generate the alt text more uh in a more descriptive way, in a more useful way, and then getting that into your website uh is gonna be a huge time saver, but also you know, a huge benefit to the people who really rely on that information. Yep. Another multimodal capability that you think about this. If you were say HR Block, the company, boy, wouldn't it just save you time to scan a picture of a customer's tax return and feed it to the GPT for model and say, what tax implications did I miss? Right?

What deductions did I not do right and stuff? And we'll go through and successfully audit that and say, yeah, you missed these things. Can you imagine how much time that would save your accounting firm to be able to say, okay, here's your your schedule K1 and there are these seven anomalies. Please have a human accountant, go check it out. It's interesting because the point that keeps coming up in the back of my mind is this is all really amazing stuff.

If you can convince people that AI is not going to take their job, that AI is going to change the way that they do their job. And so that's my first thought as you're describing, like, you know, your accounting firm, your content writers, you know, whoever it is, they're probably like, I don't even want to bring AI in because I'm going to lose my job. And so I feel like there's a whole other conversation and learning curve that needs to happen with all of this. Like for us, we're like, yeah, bring it on. We can do so much more.

But for a lot of companies, there's going to be a lot of insecurity around introducing this, even though it's only going to make what they do better. That is a hundred percent true. And the challenge that um these companies and and people and the thing they they just basically need to understand is this. This is the truth. This is in our keynote that we do.

A person with AI is going to be more valuable than a person without AI, right? If you are if you are skilled at the use of AI, you will be more valuable than the employee without. The company that embraces AI is going to be more valuable and more nimble and more effective and more profitable than the company without it. So, yes, we still need people a hundred percent, right? There's no chart here where it's just a robot and no person.

But if your company is not, if you are if you're the company's like, no, no, no, we're just gonna wait and see if this whole AI thing is a fad. It's like, you know, those companies are like, oh yeah, that internet thing's totally a fad. That's you know, it's going nowhere. Um this is a a major sea change for the way that people do business, and this is going to impact everybody. And one of the big strategic things that people need to be thinking about is how do you train employees to do this?

Because if you don't, a competitors will beat you're up. Uh, and B, there will be so few junior employees because junior employee tasks are the easiest to automate that in 10 years' time you will not have enough of a pool of qualified managers and executives to run your company because everyone will, you know, older will have retired and you're like, well, we got rid of everybody for cost savings, but uh now there's like five people left in the company, and uh I don't you know Bob over there is keeps stealing stuff every time. Well um and if you want help with that training, then you can reach us at TrustInsights.ai slash contact and you will reach John Wall, who is not as far as I know, a robot. He is a real person, but he is the lead statistician of Trust Insights, as we established last week. He is.

I I would say there's there's four ways that uh an agency like Trust Insights can help you. Number one is the basics, right? If you're at level one and you're just trying to figure out what to do with Chat GPT, we can obviously help you with that. Number two, if you want automation um of your prompts, again, you write that code. Number three, if you want to start fine-tuning a model, specifically your business, you know how to do that.

And number four, as you saw, you know, I was playing around over the weekend with building a G uh GPT J6B fine-tune model to run on my own laptop. Um, just because I can. I'm training it to write fan fiction. Um, I want to see if you know, to see if I can do that. Well, think about it.

I don't want to train on customer data and you know, stuff like that. So I I I'm fully with you. It's just I always learned so much on these uh live streams, just in terms of the capabilities and the way in which you're thinking about it. So that to me is like the fascinating part. Um we do have a question.

Yep. So go ahead. No, go ahead. Where would you draw the line on how to use it? I think there would be governance within companies, so they aren't sharing proprietary information such as data and process, correct?

Yeah, it really depends on what kind of service you want to offer to your clients and your customers. And so for us, you know, we want to keep a lot of it proprietary so that we can share the output of it. But there are going to be companies. I mean, you've probably seen the, you know, here's the 100 top most useful prompts to use for chat GPT. There's people who are willing to give it away for free or you know, for a small cost.

So it really comes down to what your specific organization, company consultancy wants to be doing with the customers. Um Chris, John, what would you add to that? What Shane's talking about is at level four, right? Which is what we talk about, building a custom model that you run on yourself that you own so that you know that that proprietary data, that sensitive protective information is not leaving uh your control. This is going to be especially important for any industry with you where you got SPI, but also um regulatory, right?

So if you have um if you have you're trying to operate, say in Italy, which just banned chat GPT because they say it violates GDPR. If you are building and running your own model that you can certify as based on you know certain training data that does not run afoul of GDPR, then you can use that model even though you can't use the GPT family from OpenAI because you can demonstrate in a court of law here's how we are compliant with what went into the model, so that um we can continue to use it. So yeah, you know, Shane, I the it's not that you draw a line, it's which of the four stages that we talked about are are you gonna be at? And if you're talking about sensitive information, you go straight to stage four, which is you know running on your own stuff. John, thoughts.

It's this is definitely something you just need to keep an eye on the opportunities, you know, because this is a area of technology that's emerging, and everybody's going to be using old type thought to figure out what to do with this. You know, it's yeah, create selfies like, well, yeah, that's great, you know, it makes for a neat toy, but there's going to be a lot of things that come out of this that people will change the businesses that they're working in. I mean, we especially when you think about image analysis, there's so much stuff that's going to be available here as far as security applications or classifying weather data, image, you know, all kinds of stuff. There's just a million opportunities here. So play around as much as you can because you know, you might stumble upon something that will completely change a vertical.

I will say this there is going to be a lot of money to be made on customization of models because as companies adopt this more and they get past the right limerick stage, uh, they'll go, huh? Maybe we shouldn't be giving our all of our data to an uh a tech company we don't actually know all that well. Um and maybe we do want to do very specific things really, really well, because a fine-tuned smaller model that runs really well beats the pants off the big models and stuff. So for progressive business leaders who are willing to invest the time and the money to hire the right talent, um, or they're willing to uh to spend the the money to bring in the right partners, um that is that's you can that's where the money is in the next couple of years. I'm looking forward for us individually having those, you know, the John model, the Chris model, the Katie model.

No, and I mean this in all seriousness to help with some of the writing. Um, you know, I'm you know, I write very differently than either of you. And so having just one model to try to mimic all three of us isn't gonna work. So I'm very much looking forward to that. Uh because we all have very distinct voices when we write and when we talk about things.

We do. After this meeting, remind me to show you the neural transfer prompt for chat GPT. It's kind of fun. Um that note. Uh thanks for tuning in today, and we will see you all next week.

Thanks for watching today. Be sure to subscribe to our show wherever you're watching it. For more resources and to learn more, check out the Trust Insights Podcast at TrustInsights.ai slash TI podcast. And our weekly email newsletter at TrustInsights.ai slash newsletter. Got questions about what you saw in today's episode?

Join our free analytics for marketers Slack group at TrustInsights.ai slash analytics for marketers. See you next time.