Sit Back and Let AI Do The Work (Automation Guide)

You ever get an email in your inbox? You're just too lazy to read it. You want somebody to just summarize it for you? Well, I created an automation where I could just tag this in my Gmail with this lazy tag, click apply and within a few minutes I get a quick summary of that email. But not only that, I get the email read by me. The sender is offering a free searchie account and the opportunity to earn money through affiliate marketing. So that email was summarized into a couple sentences and then read to me in my own voice, because who's got time to read emails? I also made another automation where every single time OpenAI makes an announcement on their blog, it automatically summarizes that announcement and notifies me on Slack so that I know about the Latest updates from OpenAI within minutes of them happening. I'm going to show you how I set up both of these automations right now. All right, so I'm really excited to make this video. It's a style of video that I haven't made in quite a while. In this one, I'm going to build out some automations using AI. I love building out automations like this. And in fact, the Future Tools website is almost entirely built using automations like the ones I'm going to show you. I even made a video about how the behind the scenes of Future Tools is run. You could find that video here. And to build these automations, I'm going to use a tool called make.com which happens to be the sponsor of this video. However, I was already using make.com before they ever came along and offered to sponsor a video. I had plans on making another video about how I do various automations behind the scenes in my business. So the timing just kind of aligned well. And Make.com's a perfect sponsor for this one. If you use the link in the description, Make.com is offering a free month of their Pro account along with 10,000 operations. So make sure you use the link in the description to get started with make.com. but as you can see on my screen right now, I actually use make.com quite a bit. If I click up on my organization here, you can actually see, Here's a chart of the times I've used make.com which is pretty much almost every single day. Now, if you're not familiar with a tool like make.com, what it does is it allows you to connect multiple tools together and and then string them together into a sequence. So, for example, here is the flow that runs Future Tools behind the scenes, it connects Google Sheets with a tool called scraping Bee with ChatGPT back to Google Sheets. And then at the end it has a little bit of a sleep timer. And so what you're looking at here, what this automation flow is, whenever I add a new line to a specific Google Sheet, all I have to do is add a URL to that sheet. This scraping B tool then goes and looks at that input URL. See how when I hover over this, Google Sheets sort of pulsates there on the left, it scrapes the content from the input URL that I added to Google Sheets. Once it's done scraping that content, it then goes to ChatGPT, grabs this body content. You can see the Scraping Bee is now pulsating. It grabs the scraped content from the website and says summarize what the above tool does. So it scrapes all of the copy from that tools page and then has Chat GPT summarize it for me. Then it goes into this second Chat GPT here. And in this one you can see if I hover over this, it's pulling the information from our summarized version that Chat GPT already did. And then it gives it this additional prompt. In as few words as possible, describe what the above tool does in one short sentence. Now, if you look at the Future Tools website, if you look at any one tool here, you can see, here's the one short sentence. A platform to build chatbots and automate customer support. An app to simplify problem solving and learning. A tool to prioritize features and initiatives with customer feedback. That's that one short sentence that's generated in this second Chat GPT here. This first Chat GPT instance where it summarizes what the above tool does. I click into Future Tools and let's click right into this movio AI. Here, movio AI is an AI platform for building personalized virtual agents trained on your data, etc. This is the summary of what it found on the sales page. Once I add this URL, it scrapes the website, it writes a summary of it writes a one sentence summary of it, and then puts it back onto the Google sheet into the same initial row where that first URL was. So that's just one example of how a tool like make.com can take a whole bunch of tools and tie them all together and make them all communicate with each other. But I want to build a few more automations. So the first problem I want to solve with a brand new automation is I want something that will quickly summarize emails for me and then display them in a way that's really easy for me to consume. And in this case I want it to send me a notification in Slack with a quick breakdown of what the email says. But I only want it to do it for certain emails. So let's go ahead and start to create an automation and I'll try to explain what this automation is going to do as I go along. So I'm going to click into scenarios here on my left sidebar and I'm going to create a brand new scenario. And every scenario needs to start with a trigger. So in this case the trigger is going to be from Gmail. So I'm going to do a quick search for Gmail. And one thing you'll notice about this tool, Make.com is almost every tool, every platform, everything you can think of seems to be integrated with make. Com. The amount of tools that are connected in here is just mind blowing. So I'm going to go ahead and select Gmail and for the trigger we're going to watch emails. I need to choose an email account and now it's connected to my Gmail. Now here it says to select a folder and I don't want to make it. So every single email that comes across gets summarized and simplified. Only emails that I tell it to summarize and simplify. So I'm going to jump over to Gmail and over under labels on the left side I'm going to create a new label and I'm going to just call this label lazy. I'll create my lazy label here. Now if I jump back over to make over here, I can refresh the folders and now you can see I've got a lazy folder here. So let's go ahead and select that. Now anytime I move an email to the lazy folder, it's going to follow through on this automation that I'm creating right now. Under criteria, we're going to go ahead and set that as all emails mark email messages as read when fetched. We'll leave that as no and let's go ahead and click okay, choose where to start. We'll just select from now on. Now if I right click on this, I can click to run this module only. It will run it once and then I can click here to see what it found. Now it didn't find anything because I haven't marked any emails with that lazy yet. So let's jump back over to our inbox. I've got this email here that's sort of a description of a product. I'm going to go ahead and click on this one and let's go ahead and give it the label of lazy. Click Apply. Now let's come back over here and if I click run this module only. Now when I click up on my magnifying glass, you can see it found the one email that I just moved to the Lazy folder. Now what I want it to do is just break down a summary of that email for me. So I'm going to go ahead and click on this little plus here to add another module and this time I want it to use Chat GPT to summarize the email for me. So let's go ahead and type in OpenAI and you can see it brings up our OpenAI option here. And for this we want to create a completion that's basically having it run a chat for us. So we'll create a completion. I've already got an OpenAI connection here, but if you don't, you can click Add. It will ask you for your API. If you don't know how to find an API key, go to platform.OpenAI.com playground come up, click on your profile, pick at the top, click on View API keys, create a new secret key, and then copy and paste the API key over into the box where it asks you to do that and you'll be all set to run GPT3 here. All right, so now it's asking to select a method. We'll go ahead and leave this on. Create a chat completion. And then under model we have a whole bunch of options. For model, you've got GPT4, GPT 3.5, Turbo, GPT 3.5, turbo, 16k, etc. Now you could use GPT4 and it does do a little bit better than GPT 3.5. However, GPT 3.5 is so much less expensive to use that I highly recommend for something as simple as summarizing emails just using GPT 3.5 Turbo. Now, under Messages we're going to click on Add item. And this is essentially us like prompting a chat GPT. So under role we want it to be the role of a user. So we'll go ahead and select that under Message content. As soon as I click into this message content, you can see it popped up this box and it's pulling in all of the data that it found when it searched my Gmail earlier, including the text content and HTML content of the email. So I'm going to go ahead and select the text content of the email here and then I'm going to go ahead and give this a space and then I'm going to act like I'm just prompting chat GPT along with this email, I could go ahead and give it a prompt. Please summarize this email into a single concise paragraph. Let's go ahead and end it with a colon and I will put the text content into quotations here. I'm going to come back to my first step here. I'm going to choose where to start. I want it to actually pick that same email again. So I'm going to go ahead and select all emails to only put the one email in there right now, I'll click okay and then let's run it once. All right, so it found the email and now theoretically it should have summarized the email. If I click on this magnifying glass, I can see the output here. If I come over here under output and I click on choices, I expand this one, I expand this message. You can see here's the response it gave. This email is a response from 3D Maker Pro to Matt Wolf's inquiry about working with the seal, 3D scanner, et cetera, et cetera. So it's summ summarize this whole email into just one paragraph. Now I want to get that summary into Slack. So let's go ahead and open up Slack here. I'm going to go ahead and create a new channel and I'm just going to call this lazy email. Click Next, I'm going to make it a private channel so only I can see it and I'm just going to leave it so I'm the only one who sees this channel for now. So I've got my lazy email channel here. Let's go ahead and add another module. This time I'm going to search out Slack under messages here we've got create a message. So let's go ahead and create a message. I've already connected my Slack, but if you haven't already, you can click this add button. It'll open up Slack, ask you to authorize it. You authorize it. It's connect, enter a channel ID or name. Let's go ahead and select from a list channel type this is a private channel. And let's select Lazy Email, the channel we just created. Now, under the text, I'm going to go ahead and click under choices here you can see here's all of the information that OpenAI is passing along. So we'll select choices, message and then content. Let's take a peek at advanced settings here and we should be all good to go. Let's go ahead and click okay, once again I'm going to choose where to start just to make sure it pulls in that most recent I will click okay, here's my Slack. I'm going to go ahead and start running this and then we'll keep an eye on my Slack here. So let's go ahead and click Run. Once it's starting to run, let's open up Slack so we can see it as it happens. And there you go. The message just popped right into my Slack. The email is a response from 3D Maker Pro to Matt Wolf thanking him for choosing the product and explaining the basic workings, et cetera, et cetera. So it explains this whole email to me, but it didn't really give me enough context. So let's make sure that in my Slack message here it actually gives me the name of the person that emailed me. So for my text box here, let's go ahead and add a space here and then I can actually pull information from this original Gmail as well. So we have the sender name. So let's go ahead and put name colon and then we'll put the sender's name do another space. We'll put the sender's email, pull this in and then I'll put email summary. So this is how I'd like it to be formatted. Whether it actually comes out with that formatting in Slack or not, I don't know. We'll find out. And then I can actually link straight to where that email is, is if I want to just jump straight to that email in my Google. So let's just put email link and then I'll put this URL right here to the message link we'll click okay, I'm going to choose where to start again. Same reason as before. I just want to make sure it pulls in the same email. We'll click OK and run it once. I'll open up my Slack here and you can see here's the new message that came across name, here's the person's name that sent it, the email that they sent it from, the summary of the email and then the email link. If I click on this, it should take me straight to that message in my Gmail where I can immediately respond if I'd like to. So now I can turn on this automation. So right now it could run every 15 minutes and every 15 minutes it's going to look in my email to see if there's something that got that specific lazy tag to it but it will also use some of my make.com credits every time it does that search. So I don't need it to search that frequently. So I'll probably just set this one to run once every day and then at the end of the day it will just send me a bunch of the Slack messages. I'll just have them come across a at 10 o' clock every day. I do need to change this maximum number of results because if I add a whole bunch throughout the day, it's only going to give me one back at the end of the day. So I can set this as 20. I doubt I'll be putting 20 in there every day if I use less than 20. This is just the maximum number of results. If I only give three emails that tag throughout the day it's only gonna use the 3 and not need the whole 20. But I can send up to 20 emails a day to my lazy tag. We'll click okay and we will turn it on. And now this is going to run for me every day and at the end of the day send me a handful of Slack mess with summaries of all of the emails that I gave that lazy tag. Now that this one's ready, I'm going to go ahead and save it. I actually will rename it real quick to summarize lazy Gmails. Let's jump back to scenarios you can see. Here's the scenario I just created. It shows you all the tools that it's working through and let's create a new scenario. Now one thing you've probably noticed about me is I stay fairly tapped in on all the AI news. I use a tool called Feedly Leo to kind of keep up to date with the news. But I want to be even more timely than that. I want to know the instant OpenAI or Google or one of these big companies makes an announcement as opposed to when I do my sort of daily roundup to check in on the news. And so in order to do that, I'm going to create an automation that also sends details to Slack to me, but it watches RSS feeds, summarizes the content from the RSS feeds, and then sends me a Slack message of just the piece of news that I'm interested in. So let's run through this process one more time. I'm going to set a new trigger here. I'm going to type rss. You can see we have an RSS option option here. I'm going to select Watch RSS feed items and then it asks for the RSS feed to keep an eye on. In this case, I want to go to OpenAI and I want to go to their blog. I want to get updates every time their blog is updated. Luckily they have an RSS feed. If I go to OpenAI.com blog RSS hit enter. You can see here is their RSS feed. It doesn't look pretty, but don't worry about it. We'll go ahead and grab this RSS URL. We'll copy this and we'll paste this in as the URL and we'll click ok. Choose where to start. Let's go ahead and choose for after a specific date so it doesn't automatically pull in everything here and I am recording this on the 25th, so let's just pull in everything since the 18th we'll click okay, run once. Now if I click on this we can see that it found a blog post from September 19th. Let's go back in here and return up to 10 results. I'm going to click okay, let's run it once more and you can see now it found it looks like three different blog posts including OpenAI Red Teaming Network Chat GPT can now see, hear and speak GPT4 Vision System Card all right, so now that I know it's accurately pulling in the latest blog posts from the OpenAI website, I'm going to use a tool called Scraping Bee, which is the same tool I use to scrape content for my future tools website. But this time I'm going to use it to actually scrape the article content so that it can easily summarize it. So if I look in this RSS feed here, they give me a summary of the blog post, but it doesn't pull in the entire blog post. However, it does give me the URL of the blog post so I can work with that. So I'm going to go ahead and add another module. This time I'm going to use this tool Scraping B here and I'm going to use Extract. Now this Scraping Bee tool isn't free. Their lowest plan is $49 a month and since I'm already using it to pull information for future tools, I've already got the 49amonth plan that I'm using. However, they do have a free trial. It's a pretty generous free trial. I don't know exactly how many scrapes you get on the free trial, but I was able to do a ton before I started to actually have to pay. So you can play around with this for free and if you find it valuable enough Then upgrade and pay for a plan. But you can use the free trial to do what I'm doing. So jumping back over to make.com click into the URL and you can see here's all of the data that our RSS feed pulled in. So we have the URL here of the actual blog post that the RSS feed found. So let's go ahead and use that as our URL under extract rules. Now, there is a little bit of a code that you have to paste here. I have it in a notepad here. We want it to grab the title of the article and we want it to grab the content of the article. So I'm going to go ahead and copy this here and I'm going to paste it into our extraction rules. And if you want to go ahead and pause the screen and see what I did here, you need to enter this code to get it to properly scrape the body content and the title content from the article. Now I'm going to go ahead and click ok, so I'm going to go ahead and right click on my RSS here. I'm going to click choose where to start so that we can do another test run. Let's go ahead and select from the RSS feed and let's just go ahead and select this most recent chat GPT notification. Let's click ok. And I'm going to run it once just to make sure everything's working so far. All right, so it looks like it worked. Let's go ahead and take a peek. If I open Operation one here, let's click on our our body output down here. Open up our content and you can see here's everything that it scraped off the website. It scraped everything. So it's got menus, everything that's in the footer. It just pulled it all in. But that's okay because we're going to send it to gpt3 and have gpt3 clean it all up for us. So let's go ahead and add another operation here. Let's go ahead and add OpenAI Create a completion just like last time under model once again, I'm going to use GPT 3.5 Turbo. And then under messages I'm going to click Add item the role I'm going to select as user and then message content. We're going to go ahead and select the content that was just scraped using Scraping bee. So we'll select that content here and then I'm going to give it the prompt of summarize the above article into a concise paragraph, then explain it like I'm five. That way if it's over my head, I get that quick explanation as well. I'm going to go ahead and click ok, do another test run using the most recent post here. Click run once. Now OpenAI is trying to summarize it for me now. If I click on my magnify glass up here, I can see what output it gave me. So if I click on choices one, open up our message here, you can see we have a much shorter paragraph. OpenAI is introducing new voice and image capabilities in chat, GPT, their AI powered assistant, et cetera. All right, so now I want that to end up in Slack. So as soon as OpenAI puts out a new blog post, I know about it. So let's open up our Slack here. I'm going to go ahead and create a another channel inside of Slack and this one I'm going to call News Updates. I'm going to go ahead and just make that one private as well. And I want to call it News Updates so I can actually create this automation for multiple RSS feeds and then stay in the loop with not just OpenAI but also Google, Microsoft, Adobe, all of those companies as well. Now let's go ahead and add a final output here and let's have it send it to Slack for us. I'll select Slack, create a message. Once again, this is a private channel and we've got our News Updates channel. Now for our text, let's start by pulling in the original title all the way back from the RSS feed. So I'll go ahead and put title. I'll put a title with a colon in front of it so that I know that that's the title. Let's give it a space and this time let's bring in our summary. So let's go ahead and type summary, give it a space here. And then under our choices messages we'll bring in our content here and then I'll give it a couple more lines and this time I'll put source and for our source I will link to our original blog post. That way I can quickly open to that blog post. When I'm ready, I'll go ahead and click. Okay, and now our full automation should be done. So let's ahead and select an RSS item that we want it to start with. Let's go ahead and use GPT4 Vision System card and let's let it run once running OpenAI now. And now it's sending it to Slack. So if I open up my slack here. Check this out. News Updates title chat GPT can now see, hear and speak OpenAI is introducing new voice and image capabilities in their Chat GPT AI system. This means that users will now be able to have voice conversations with ChatGPT and show it images to display discuss. The Voice feature allows users to talk to Chat GPT and have it respond back, while the Image feature allows users to show images to ChatGPT for analysis and discussion. This these new capabilities make ChatGPT more versatile and intuitive to use, so it pulled that in and then it also pulled in this GPT4 Vision System card. OpenAI has introduced a new system called GPT4V that allows its artificial intelligence model to analyze images image inputs from users. This addition of image analysis to language models like GPT4 is seen as important, et cetera. And then down here it actually gave me explanation for a 5 year old OpenAI made a new thing called GPT4V that can look at pictures and understand them. It's like a smart helper that can think about both words and pictures. They worked hard to make sure it's safe and won't cause any problem. This is important because it can help people do new things and have fun experiences with computers. Here, here. And of course like I asked it to, it links straight to the source so I can click over here and get to the blog post on OpenAI. Now I do want this information as soon as I can possibly have it. So right now it's going to search the RSS feed as every 15 minutes to see if there's new information posted. However, it's unlikely that companies like OpenAI and Microsoft and some of these companies are going to make huge announcements in the middle of the night. Here in the United States it's also unlikely that they're going to make huge announcements on Saturdays and Sundays. So I can have it check at regular intervals, but let's tidy it up a little bit. Now I don't need to know within 15 minutes of the news announcement. I would say once per hour is probably fine. If I know about a news update within an hour then I'm probably still going to be pretty dang on top it of of it. So I will have it check at regular intervals of every 60 minutes. And then down here we have the option for advanced scheduling, let's say from 7am to 9pm and then for days of the week let's just do Monday, Tuesday, Wednesday, Thursday, Friday. We'll exclude the weekend so it's not checking every hour on the weekends as well. And even if something pops up on a weekend or outside of this window, when it does check during this window, it will still find it. I'm just less likely to be at my computer outside of those time frames anyway, so I wouldn't say see it regardless. In fact, the time two I'm actually going to pop this down to 8pm so it'll check about 13 times every day Monday through Friday for any new news and I will get it within an hour of it coming out. We can go ahead and click okay and turn this on. And just like that I have two new automations. One where I can get a quick summary of any email that comes across my inbox by just adding the lazy tag to it. It's going to look at the email, it's going to summarize it, and then it's going to send me a Slack notification with just what I need to know about the email so I don't have to read the whole thing. And I also created an automation where every time OpenAI makes a new blog post, it's going to summarize that blog post for me and then notify me in Slack and I will stay as up to date as I possibly can with all of the latest news that comes out of a company like OpenAI. I can now go clone this exact process here and do it for other blogs that I want to stay in the loop on. Now if I want to get extra nerdy and extra lazy, let's go back to our Summarize Lazy Gmails automation that we created here and let's add another automation into the mix. This time I want it to actually read it for me because I'm lazy and wants to read it. Just read it for me. Let me add a module in here and this time let's add 11 labs. This is a popular text to speech AI. Let's do create a speech synthesis. I'm going to put it right here in the middle after Chat GPT but before Slack and we will connect all of this up. So now it's going to read the Gmail, it's going to summarize it with OpenAI. Let's actually get it to create an audio file for me. So once again you're going to need to connect your 11 labs count. I've already connected it. You're going to click add. It will ask you for an API key. You can find it on the 11 Labs website. Paste that in there. You're good to go. Now under voice they have all sorts of amazing pre Made voices inside of 11 labs. But I did train my own voice into 11 labs, so might as well use my voice for the model. I'm just going to use the 11 English V1 that seems to work well. And then the text that I actually want it to speak out, we're going to pull that from our OpenAI output. So the summary that it created for us, I'm going to click on choices messages is content, and then that is the content that I want it to actually speak to me with. Now let's click ok. And now in this flow, it will create an audio version. Let's go ahead and run this through its process to make sure that things are working so far. If we click on our little magnifying glass here, you can see that it took our input here and it created a file here. Now if I actually go to my 11 Labs account and I click in my history, you can see here's the actual audio that it created. The email is a response from 3D Maker Pro to Matt Wolf. So now I've got that audio version there that I can use, but I want that audio version to pop into Slack for me when the message comes across. So in order to do that, I actually need to store the audio file somewhere that it just grabs. So I'm going to go ahead and unlink these two here and let's go ahead and add a module in between. And let's go ahead and save the files to Google Drive. So we'll connect Google Drive here in the middle. We'll upload a file. I created a folder called temp for temporary. Let's go ahead and use that folder. And then the file, it's going be this 11 labs, create a speech synthesis and we'll click okay, and now in Slack we want it to pull in that file. So I'm going to open up Slack here, come down to our text. I'm going to put audio link with a colon here. And then under our Google Drive we have all of these options, but I want to select web view link, go ahead and click okay, and we should be able to run our full automation here now to run it once. Hopefully all of these pieces work together. And I just got a Slack notification here. Let's take a peek. We've got our name, we've got our email, we've got our email summary, our audio link, and the link directly to that email. If I click on our audio link, the email is a response from 3D Maker Pro to Matt Wolf thanking him for choosing their product and offering assistance. And there you go. There's the summary of that email being read to me by me. So there you have it. There's a couple fun automations that you can build using tools like make.com 11labs and GPT 3.5. Do you like videos where I show some of these fun automations? Because I love making videos like like this and I have so many different more automations I can test and play with and show off and explain to you. If you guys like this, I'll make more videos like this one. And if you want to keep nerding out with me and find out about all the coolest latest AI tools that I'm coming across, check out FutureTools IE where I actually share all the latest cool AI tools that I come across as well as all the latest AI news. And I give you one simple place to find all of my updated YouTube videos as well. So you don't have to hunt YouTube for them. They're all in one place place. You can find it all over at FutureTools IO. I also highly recommend joining the newsletter where each week I send you two emails. One email breaks down all the news you need to know for the week. The other email breaks down all the coolest tools that I've come across for the week. They're both short, simple fluff free emails. I am very anti fluff so check out those emails. I really think you'll appreciate them and get the nice TLDR for the week of both tools and news. You can find more info over at FutureTools IE and clicking on Join the Free Newsletter. So thank you so much for tuning into this video. If you want to stay in the loop with AI news, AI tutorials, some fun AI challenges, maybe the occasional vlog and you just want to nerd out with me about AI and tech, maybe subscribe to this channel and give this video a thumbs up. It'll make sure more show up like this in your YouTube feed. Once again, I really really appreciate you and thanks to make.com for sponsoring this video. I'm a huge fan of that tool myself. I use it all the time. So they are the perfect sponsor for this video. If you use the link in the description description you'll get a free month of their pro plan as well as 10,000 operations to use so you can test out the types of things I made in this video for yourself. So thanks again to them and thanks again to you for watching this video. I'll see you in the next one. Bye bye.