Why AI Becoming Cheaper Is a Big Deal β
Let's become an AI prompt engineerπ
2024-07-23
Transcript β
[00:00] if you've watched any videos on the topic of gbt 4 mini they're probably giving you the wrong idea out of curiosity I watched a couple videos on the topic of gbt 4 mini just just kind of see what other people are saying and the ideas on YouTube in holy smokes no one really gets it to be fair I only watched a couple but the ones I did watch I was like oh wait they have no development background they don't actually use the models within their business and they don't actually know the implications of what these models can do therefore in this video I'm not going to give you some bland article and just list some stuff I'm actually going to give you a real world impact a real
[00:30] world use case that affects my AI software business directly of why this is important not that it's just hey there's a new model by open AI so you're telling me Corin all I have to do is watch this video and I'll actually get a real perspective on someone that actually uses these endpoints in application development and actually understand what this means for software development yes let's jump in I didn't mean for this video to start off so intense but you know that old saying curiosity killed the cat oh man this cat got killed I was looking at other videos on the topic of gbt 4 many and I was
[01:00] like oh they're missing the entire point and then I realized oh they don't even use it and then I realized people actually want to know what is the use kit why this is so important therefore let me tell you why and the huge impact this has with software development that we really haven't seen in the past for anything before I dive into a little bit of a history lesson here's some context my name is Corin Brown I'm developing an AI software using open AI as my AI provider it's bump ups.com you can check it out it's made for Content creators end goal is that it's going to be as needed as a good mic is for any content
[01:31] creator got that out of the way but let me tell you why this endpoint's actually a huge deal I've been developing software since I was 11 or 12 but that was in the context of IOS app development When developing apps for IOS app development it was fun it was cool the idea was you know you build your app for the iPhone which was a piece of hardware and people on the store got access to it through the App Store you know this you understand this and apple in that context takes a 30% cut of all the money you make through the fact that you're using their hardware and getting the Expos through their App Store therefore when updates would Ur for iOS
[02:03] developers it was typically in the context of hey you got a new framework or you got a new capability a very specific context of this was in 2017 I believe you got the ability to start doing augmented reality within your apps therefore an update in this context is very much like hey you have a new capability and new functionality that you can do within the actual Hardware of the iPhone in that context would be the AR and the camera therefore this kind of update when it came to software was very much so new feature new feature not the
[02:33] best way to write it okay that's cool but what is your point of bringing that up and what does that mean when it comes to AI providers in this new endpoint gbt 4 Al minum why this is such a big deal and honestly this is such a bubble right now but not in a sense of like stock market going to explode and everything's going to go Capo more in the sense of we're in such a bubble right now that this news is such a big deal that people don't actually understand what this news actually means it's actually insane so like past this when it came to software development obviously we would use other things such as our backend and our front end if you're using a web application so for example our backend could be gcp
[03:05] Google cloud and our front end could be Firebase when we get updates for this kind of software and as a cloud provider this could be in the context of giving us more features so for example a new feature that they provided recently is the ability to talk to Gemini to kind of help you out within your development process Gemini the chat bot from Google notice the trend when it comes to Providers and what they provide for the end consumer or the end developer new features new capabilities new ways of approaching that software what this is
[03:35] doing is entirely different so obviously you have the new features image analysis video analysis voice capabilities stuff of that nature but the bigger thing here and the thing that everyone is sleeping on is when open AI releases a new model and the model is better and it's cheaper this means that if you build your software on top of this model everything you produce becomes better so instead of this just being new feature this is also improved software
[04:07] let me know in the comments has that kind of situation occurred in the past and this is why this is so crazy as an AI provider we've been using open AI but specifically the endpoint of gbt 3.5 turbo for the context of the value we provider and consumer at the time of providing gbt 3.5 turbo we were looking at a cost of around 50 cents per million tokens therefore when we would provide our value as of now this value would be oriented and structured based off a 3.5 model from open AI when I say value by the way I'm talking about time stamps
[04:37] for your video titles for your video descriptions for your video hashtags for your videos Etc with the release of gbt 4 mini though we are getting access to a more advanced model that's going to lead to better outputs at a cheaper cost and here is the best part and the part that makes this even crazier for us to Simply switch our model when it comes to our workflows from 3.5 to gbd 40 mini it simply takes us just replacing the model name within our code once we put the new model in our code of gbt 4 mini our
[05:10] value increases our cost decreases and the overall consumer is getting a better product that's why when this came out it's a huge deal it's an extremely huge deal now we're able to do very intuitive softwares with very very high functioning AI models furthermore you're telling me it only cost 15 cents for a million tokens and you're telling me that long-term we're seeing a trend where this could possibly only cost me 10 cents per a million tokens 5 cents per a million tokens and you're telling me that when they release these new models and all I have to do is simply
[05:41] switch the model's name the end consumer that's within my software gets a better Value Point all I'm hearing is win win and win based off my development experience in the past 10 to 12 years it was always new feature new feature new feature but now it's new feature and improving your underlying software with just a simple model switch to put it blunt and simply if you want to enter the space of software and you're not developing a software that uses artificial intelligence you're missing a huge opportunity here because one thing has become very apparent very apparent
[06:13] once you have the infrastructure build out and the workflow build out adding more and more advanced models at cheaper and cheaper rates is going to be a trend and therefore your end consumer is going to get happier and happier and happier and the overarching argument here is that to give the end consumer a higher quality output with us only doing 5 minutes worth of work within our code that's crazy like that's actually crazy so let's go open AI this will definitely lead to probably other AI providers following the footsteps of open AI providing higher quality models at cheaper rates really causing capitalism
[06:44] to do its work here get some better Ai and I'll see you in the next video These two videos are based off clicks YouTube's algorithm and everything above the board this is my face and I'll see you in the next video