How To Use Google AI Studio with No Code
Let's learn how to use Google AI Studio (gemini 3)📅 2025-01-21
Transcript
[00:00] let me show you how we can start accessing Google AI studio in a no code way to do this we'll use zapier a no code automation software so by the end of this video you'll know how to do everything you want to do within Google AI Studio within zapier and have zero coding knowledge sound good let's jump in welcome back y'all today's video is sponsored by zapier they hit me up they're like Corbin there is a ton of cool stuff we can do within Google AI Studio I'm like yes this is true let's do a video about it my goal of this video is to not only show you how to integrate Google a studio but also give you just a real quick run down 101 tips
[00:31] on how to start prompt engineering when accessing and talking to artificial intelligence through automations therefore let's do it step number one make sure we even have it connected so I'm going to come over to my apps here I'm going to search up Google AI we're going to do Google AI studio now I already have it connected here so I can do test connection you'll see it's good to go for you simply hit add connection once you hit add connection here come to your Google AI Studio dashboard dashboard here all you have to do is go to get API key hit create AP API key simply just create a Google Cloud
[01:02] project and hit create API key just for reference this API key is a pay as you go situation so this isn't like you create the API key and it's like guess what we're charging you $10 a month now no matter what no no no no no it's only going to charge you based off how much you actually use it so don't worry no hidden fees no craziness once you do that simply paste it in there and connect and just to make sure it works you can hit these three ellipses and do test connection test successful let's create a zap I'm going to come over to my zap your zap's folder and if you're like what the heck is going on over here
[01:33] I've done a ton of videos when it comes to Ai and automation so check out this channel as you already know the zap I'm going to create in today's video create new zap is going to be in the description down below for free simply click the link add it to your profile because the purpose of today's video is just to show you prompt engineering we're going to do a trigger here but it's just going to be a scheduler so we're going to do trigger event every day continue time of day 1200 a.m. continue and test trigger I like doing this as the trigger when you're just messing around and zappier for automations and just want to try different Integrations this is a nice little just just do something so I can
[02:04] get into the actions I'm going to rename this zemp to Google AI zap let's jump into the fun here so we're going to type in Google AI Gemini we're going to do an action event of conversation so why are we doing conversation we have two options here conversation or send prompt in reality if you plan on using artificial intelligence in an automation workflow whether that's creating an email draft analyzing a PDF document whatever it may be you want to opt for the action conversation because of the fact that we can optionally store messages as you go or in this context I
[02:35] like to call it the memory key what is the memory key Corbin the memory key is What ensures consistent outputs at scale what do I mean by this if this automation was to run a thousand times we want to make sure the output for those a thousand times are consistent when it comes to the actual AI itself so I'm going to go and choose our model here we'll do 1.5 Pro if this kind of looks confusing like yo there is just what's going on there's too much going on typically opt for the model names that are the simplest so Gemini 1.5 Pro
[03:08] don't do Gemini 1.0 Vision latest you want to go with whatever is simplest so you can kind of proceed that if there's ever a new update it gets pushed to here and you don't deal with deprecated end points let's get to the message now let's get to prompting rule number one with prompting is we need to tell the AI model who they are what they do what the heck's going on so I like to always start with context semicolon what are you trying to do so in this example we'll just assume we're a social media marketing manager that has to create
[03:38] three captions I'm going to say you are a social media caption generator done now what are these steps that this social media caption generator is going to take in order to give us the end value point we're looking for steps to follow the way I want you to think of this specifically this part is Imagine we're not teaching the AI to do the outputs we're teaching an actual employee to do whatever we want therefore break it down into steps like you were training someone to give you those social media captions break it down into steps for the AI model step
[04:10] one look for trending topics now I'm going to go ahead and Link another zap your video I did right up there and in the description down below this shows you how to get live data through website links using a web parser or an RSS feed to get most upto-date information we can load into these chats the reason I bring this up is that a really great block that we could use before this is like an RSS feed of new events or something that's trending so that our data is more up to date with every step that proceeds from here so I'll make sure to leave that in the description down below think
[04:41] of it this way it's like the Chad gbt web search feature but we put it within an automation looking for trending topics step two generate three captions step three add two hashtags to each caption step four add one emoji at the end of each caption break this down if you have to write it on a piece of paper write it on a whiteboard think of you doing the task or training an employee do the task what is the standard operating procedure of approaching the end Value Point so if this is a reply to
[05:13] an email what do you typically go to when you reply to an email well I know business context okay provide that data of your business context when replying so what is step one what is step two for this though here's our next situation step five output as the standard template Corin what is the standard template we're going to make that standard template this is how we want the output to look every single time we run whether it's the 50th time or the 200th time this is what we want so specifically in my context we're dealing with captions caption one caption two
[05:43] caption three caption one two and three are obviously placeholders for the real data to be put into now a last little section here which is not absolutely pertinent but you can add is a format section this is where we put information like hey I only want the caption to be two sentences long each not four sentences along each I only want my email to have five sentences Max and always end with this signature information like this for now though we don't really need this and what's cool is we have a memory key this is what I referenced earlier random string of 32
[06:14] characters this allows consistent outputs now watch this we add our memory key here and if we do continue and test this trigger and we like our output that kind of just locks Us in for further automation flows using that prompt when I say lock in that means that the output we see here for test one is going to look exactly or near the same of test two when it comes to the formatting of it not necessarily the actual underlying data that is outputting for c 1 2 and 3 what I'm describing there is fundamentally one of the most important skills to learn impr prompt engineering
[06:45] is our ability to have consistent outputs at scale this is extremely important as when we do this in software we do this in automation flows that's kind of the name of the game that is why another variable here that could be relevant when you're doing prompt Engineering in this context is the temperature variable what this is is basically telling the AI model how creative you want it to be typically in software we don't want it that creative we want consistency over like unicorns and rainbows okay continue test up and here we go notice how our output has
[07:15] this little square bracket there nice and then notice there's going to be a comma after every single caption caption one comma caption two comma ction three end in addition it followed the steps to a T two hashtags one emoji at the end two hashtags one emoji at the end name of the game of AI models and prompting is consistency that's why we love the memory key and that's why any type of output I do from here on out will do the exact same type of formatting so the next best video for you to watch in
[07:46] order for you to get more context of how do I input like more live data into there where I can search the internet and provide it check out that video in the description down below I'll show you how to search the internet using a no code tool like zamp you're here without further Ado that covers everything I want to do in today's video of Google AI Studio Ai and accessing it see you in the next video Gemini flash 1.5 see random videos that's my face I'll see you in the next video