AI for Inventors: What to Use It For (And What NOT To Do)

Michelle Westfall: Welcome another 52 pitch to product podcast. Today, I have Richard sitting next to me. Richard's one of our customer success managers he spends a great deal of time with our clients after we ⁓ clients. I think today, Richard, we're going to talk about something that comes up a lot lately and that's AI use and how do we incorporate AI or how do our clients possibly incorporate AI to their development of their product into marketing, into anything that to do with AI.

AI is an awesome tool. And just like Richard said, it's a tool. And it's not 100 % guarantee. If it were, then all of us would be out of jobs, apparently.

So again, it's a tool today. And I'll let Richard speak to some of the issues that we have using AI. I know that he's first gonna talk about when you create a logo, for instance. AI is not tapped into the patent and trademark laws and rules.

So we have a lot of clients that come into us with a with a logo or a name that just is not usable, just because they found it on AI, it does not mean that we can use that and not infringe on somebody else's artwork. So, ⁓ you talk that. Yeah, mean, AI is like with ⁓ any tool, can be ⁓ very powerful the right hands, but it can also be quite dangerous as well in terms of how it's being used and how you're using it to create, for example, your logo. your product.

Oftentimes what we'll see is when a client comes in, they'll have a logo, look at my logo's all ready to go, but it's not really in a usable format that ⁓ teams or anyone outside after launched a product with us, was JPEGs, PNG files are gonna be used primarily in a web application. You won't be able to use those to build your brand guidelines, for example. You can't create other variations with the logo. It's going to ⁓ you in a bad situation where it's going to cost you more money for our teams to ⁓ to redraw those logos because they're not in a usable format ⁓ an early stage when it comes out of AI.

it will lead to misconceptions within the logo because with any AI platform that you see out there, all the different AI models that you'll see, they're ⁓ basically quote unquote stealing from other sources. And that might be other logos that are living within well-known companies, for example. And your logo might be derived from a logo that has a lot of IP around it. And they have big attorneys big budgets.

So you've got to be very, very careful because you don't know where that's coming from. And it won't tell you where it's coming from. When you do a lot of text-based ⁓ within AI, some AI tools will actually create Sources, but they'll tell you where it came from so that you can you can go back and check those things as well But with a logo generation, for example, it will not do that. It will pull it from other sources They might even pull it completely verbatim directly from that source and that could get you in a hot water later on down the line So right those are things you got to be really careful about and most of us maybe you agree with this Richard most of us use AI Like most of us use Google when they first came out, you know, you search with with AI you ask questions and it'll know, formulate an answer that's relatively on spot.

That's ⁓ we're using AI or that's what the common person uses AI for today. We're about, you know, type of AI that is not way down the road, but it's just much more boosted or ⁓ on steroids, That's why we're investing the software that ⁓ really talks about what it is that we do our model, it follows our model. So when we're asking questions, it understands who 52Launch is, it understands what our client's product is, and it can really do a deep dive. But again, AI only as good as the information that you provided, I guess is what I'm trying to say.

I kind of stumbled through this because I'm old and I still don't everything there is to know about AI, other than I was told that we need to get on board with it, which I committed to. to getting the software that's gonna help us and help all of our clients, obviously, which is the most important thing. Well, if you were starting, for example, you were a client and you were gonna start investigating some logo concepts using generative AI, if you don't have a good point of view or a good reference point when you're looking for those types of, to generate ideas, if you're very vague, it's gonna produce very vague responses.

And your logo, like say, Dan, you went and searched for logo and we both put in the same prompt. most likely gonna get a very similar result sometimes. And ⁓ was launching a product and you were launching a similar product somewhere else and we were both looking for the logo, our logos might come out very similar to one another because it's driving from the same sources. But having a very distinct outlook and using AI in a way that actually helps inform your own thoughts before having it generate physical images, sometimes that's a better approach when you're using AI.

Yeah, ⁓ think what you're trying to say is that we're still relevant as humans. still to make decisions and ⁓ it from years and years of experience as well. So AI ⁓ is great tool, but it's not the end all be all today. Maybe five years or even less than that.

Things will change ⁓ drastically. again, all of this is going to be to help better the product, the outcome of the product. Research is a good area. Like if you're using AI, I would.

AI questions about market who's likely buy your product. Things like that are really good for using AI initially. But again, it's something that is gonna replace product ⁓ or marketers, So would the best way approach that too, if there's something that intrigues you. Go out in the world and go see it, take photographs of it, do it yourself.

Sometimes you'll get a better outcome in the end because you're not being fed what it thinks it should be rather than letting you drive it. So don't let the AI drive the idea for you. You still need to be very much involved in every aspect of it. Whether that's creating the colors, you can have it give you updates and relative sort of suggestions.

But ultimately, at end of the day, you should be making those decisions. Don't let the AI do it for you. there's one other thing that we also be mindful of when AI is generating a logo, for example. And there's a lot of tools out there that do this.

But the biggest thing is, one is the image that it's deriving from, but also the font. As AI systems get better at generating... putting the words onto the page in proximity to your logo, you don't know where that's coming from. And there's entire industries that are designing the fonts or the typefaces that are used in those applications.

And if there's any likeness, that could open you up to potential problems down the line. not right away, but as your company grows and you start selling more products, it could become an issue where you have to rebrand yourself because that already exists somewhere else or it's actually derived from. someone else who's put a lot of effort and time into producing those items. So you don't know where it's coming from in the end.

Yeah, good point, Richard. area where AI is not very useful today would be in our design for manufacturing ⁓ process, if you guys have listened to our podcast in past, you know that DFM is a very strong acronym. it is... that has to happen for our product to come out successfully also come out Profit ⁓ this margin.

⁓ We want make sure that we're making it the most cost-effective way So without the design for manufacturing stage, it's like throwing a dart against you know a dartboard and hoping that you hit the center Which is very rare But again, that's something AI would not know ⁓ today ⁓ again the information we can provide With our own software that we're developing that's going to speed things up So I guess the end what AI does is it hopefully makes things move faster? Again, there has to be a lot of human elements still in ⁓ in what we do a lot ⁓ actually but again AI Helps things along faster.

And again, we don't know single client that's hired us and says ⁓ take your time ⁓ No, has to at hyper speed. So yeah, I mean it AI doesn't, I like to use this a lot with either with clients or even with, you know, ⁓ designers and being in the design industry for a long time. It doesn't know taste, right? AI doesn't know what what doesn't work.

It only can regurgitate back to you what it thinks ⁓ it be, not with what, like the or the DFM phase where we don't know exactly how this thing should be produced. And AI doesn't know that either. It might give you some... general ideas of what types of manufacturing techniques to use, but that doesn't mean they're the right manufacturing techniques.

And AI can give you, they can kind of put you down the rabbit holes and you start to sort of sensitized to it being produced a certain way when in reality that's not manufacturable at scale in that particular fashion. So you gotta be very, very careful with anything you read online. I would say to bring this all the way back to the beginning, You know, AI is exactly what it is. It's a tool, right?

Until further notice, it's simply a tool that we can all utilize to the best of our ability and to the best of AI's ability. But again, until it changes drastically, it's just a tool. And it's a valuable tool that can be used by, you know, basically everybody. With any new tool.

A healthy amount of skepticism is always warranted. with any tool, but. just because you can make it doesn't mean just like generate it and then just hand it off saying, oh, look, I have all these ideas. no, you need to read through them first.

Make sure that one, they're solid. And two, they actually make sense. the scale can be off, like in a huge way, because it's just taking these random things and putting them together. So yeah, that's the other thing is with AI generated things, the scale is usually way off.

It puts you into context like, no, that product would never live in that environment. that's not that right material. you know, all that, that person has seven fingers on their left hand. Those are the types of things that, you you know, it's a dead giveaway when you see, you know, poorly generated AI images, they stand out.

You any designer, any product designer, any graphic designer, they're gonna be able to pinpoint a fake image. look, if someone comes into us with an AI version of everything they've done through AI and say, yeah, we wanna hire you, but we've got all of this other stuff done, we just want you to make it. I mean, we can discuss that. Like, you don't understand, that doesn't, it doesn't work that way.

We don't even know if we can make this product the way that AI has said they can make it. You know, we're talking about different types of machinery that are used. We would have to, we would have to go through the DFM stage. Otherwise AI might say this costs $5 and it comes out that it costs $5, but it can't be made at all, ever.

So what good did that do you? So we're always starting over from zero because we just said AI is a tool. It's a great tool, but that's all it is right now. It's not going to, I was gonna say it's not gonna make you a coffee, but it probably does actually.

Not good coffee. use AI to formulate your ideas in, ⁓ know, into a into a structure that's easy to understand so that you've you've kind of walked through but always check those ideas, right? So you don't know where it's coming from first and foremost, but make sure that you do check it. Check sources if you are doing it for market research.

Make sure you check those sources as well, because you don't want to go into any new product design having incorrect information that's going to cost you time, money. In the end. Yeah, so we'll circle back with You know definitely come definitely do your research use AI for research market research You know initial design and maybe of your pride to get you excited about what it might look like but in the end come to us we'd to have you come to us ⁓ and schedule an appointment so we can go over it and and hopefully Like you said, it's a good jumping spot, good jumping off spot.

Yeah, guide you in the right direction. Ultimately, we want your ideas to succeed. So we want to make sure we put you on that right path to success. basically, folks, use AI for market research.

Use it for maybe a jumping off point ⁓ on a design. Research what other competitors be out there. How many people are buying a product that you're thinking about bringing to market? What's the demographics look like?

Those are all great things to use AI for. And look, more excited and more educated you are about this information, we'd love to talk to you about that. We'd love to you call an idea consultant here at ⁓ and schedule an appointment where we can discuss it. It's a good spot to jump off of or ⁓ ⁓ yeah, ⁓ we'd have that than to someone that comes in with a napkin design, ⁓ although accept that wholeheartedly, but it's always better to have more information.

The more information, the better off your appointment with us is gonna be. And again, if you happen to know anybody that wants to bring an idea or a product to market, or if yourself, you're gonna bring a product to market, we'd love the opportunity to earn your business. You can go to 52launch.com.

I believe there's a link below that you can click on and reach. out to one of our idea consultants and next time, we'll see you later. Hopefully see you there. Have you already designed your product with AI or have you done research with AI?

We'd love the opportunity to have you come visit us. And what's very interesting here is Richard, he is an AI robot. Richard, introduce yourself. Nice to meet you, Dan.

See, you would have never known if I didn't tell you. ⁓ So we have the first AI cross here at 52 lunch. By the way, it's just worth coming in to see him come in. Click the link below and give us a call out if you want to meet us.

Thank you.