Latest / Pitch to Product / Why AI Can't Give You The Real Cost Of Your Business
Transcript
- Venture Program Intro: I'll challenge you on any of your projects that you're looking at and that you've done AI, have it break down not the BOM, the bill of material is the absolute wrong way to make it. Let me start by first. Remember when you go to AI and it's open source. So you're you're putting it out to the ether. You're putting it out to everyone to be able to hear about it. So you're not really protecting your product. You do want to ask a couple of questions you can get, but don't put everything out there. You have to be protected. Well, I think that that cat's out of the bag. I think it's too late. Everything, everything I've seen, people have already given away all their trade secrets. AI helped them build it. Th there is a discussion out there like Bobby's saying, that's I don't know where it's gonna land. I I think it's gonna land not in our in the favor. I think when you publicly disclose something, you've in you invalidate yourself from actually patenting it. So I would be a little careful. But let's also go with reality. It's a fantastic tool. It helps you design, develop, like the stuff that we see that's coming in helps us visualize the idea in a much more finished format rather than faster. So much faster and and and and and A person that can't draw comes up with sketches now that are phenomenal. Yeah. We have to go back and say, well, now we have to build it. No. You know? So the questions that you need to ask AI that you will not get a correct answer. It does that. It does that incredibly well. Like it shows you what you're asking it for. What you can never really ask it is, hey, I want to design the guts of this product, right? So again, I want this widget to produce something and it's a mechanical and it's a it's a coin operated machine. You know It won't give you the internal components of that 'cause it doesn't know. Or or or or like the guy did with the gave us the the display product, it was eight layers to make this product. Or a product you wanted for twenty dollars and now it's gonna cost you a hundred and fifty dollars because all the components in the circuitry that AI added to this thing that's eight layers deep that we didn't need. It will never have the and look at this is where again our company we're already building those Internal models that will help us take what AI knows as general knowledge and then turn it into specifically applied information. That specifically applied information needs to be connected to our systems, a CAD service, a model. It needs someone to direct it. It needs someone to give it those proper prompts. you know, again, I'll challenge you on any of your projects that you're looking at and that you've done AI, have it break down not the BOM, the bill of material is the absolute wrong way to make it. A factory already has that BOM. It's the process to make the bill of materials into an actual product. There is no cost that can be defined there. That is labor, that is people, that is design, that is the way equipment works. If you are looking at a list of, hey, I bought this from Walmart, this from Kmart, this from like whatever, and I just have the laundry list and it I the item I list is is $14.98. Triple that because those itemized lists don't include and you can probably almost multiply that by 10 because If those items need to be connected to a circuit board, need to have software put in, need to have like the hardware you know, collaborate, th have the proper licensing so that they can communicate, that they can actually deliver all the services or the little functionality. Just to turning a an LED light on and off requires software, so hardware. It requires code, it requires a chip. Like it needs to know if it's dimmable, now it's another step. Like these are things that There is no program yet that AI can do that. It can give you the process, it can give you some of the basic schematics, but it can't give you the shortcuts. So right now you're exactly it. The shortcuts are what you're gonna pay a lot for. AI doesn't know what person makes this piece already that has it there and who's the right person to engineer it for you. Or what we could take off to it to save us money that we don't need. AI doesn't have that. It can't tell you that and it can't give you the budgets. Yeah. And and look it, that's not a knock on AI. Will it learn that? I don't think it will. I th I think it's because it's not specific enough. It's too it's too small. You know, it's a it's a general language model. it's a learning language model, which means someone has to talk to it and give it their knowledge. That's like, you know how long it would take me to get I can't even get all the knowledge out of Sander that he's got for 40 years. In fact, I'm not even sure I want to. No, some of that stuff is like, wow, it's over my head. But that is the Sander knows how to ask it the right questions. And and even then watching him get frustrated that it's not giving him the answers, because we all want it, man. We all want our jobs to get easier, but it's You know, it's like that adage where you know, repairman comes in to fix something and he uses a hammer and charges somebody ten thousand dollars because he he, you know, hit the hit the machine in the right place. And the guy says, Whoa, whoa, I'm not paying you ten grand. I want an itemized list. He said, you know, three dollars for the hammer and nine thousand nine hundred and s ninety-seven dollars for the knowledge. Right. There's your itemized list. That's where that's kind of where the best analogy that I can give somebody with AI. It is knowledge based. We have the knowledge. We are utilizing AI just like you are. I would encourage you to like understand where you hit walls. Walls is does this really cost that this much money? What are the shortcuts that we could use? If I don't have the hundred thousand, you know, this is where again, if I don't have the hundred million dollar budget, how would I do this as a startup? Like this is where you're gonna find you're gonna hit walls by asking the right questions.