Speaker 1: You can now make academic presentations using AI. This is how I would do it. The first thing I would do is go over to gamma.app. That is the app we're gonna be using to create our AI presentations. Not affiliated, not sponsored. I think this is a really great one, though. This is the easiest thing you can do. The first thing is if you are working on a paper, like a peer-reviewed paper to submit to a journal, which you should always be doing, by the way, you can just put this in as a file and it will generate its own presentation based on the draft of the paper you're currently writing. And that's really easy. You go here to import a file. This is the one I'm working on at the moment. This is a paper from my last postdoc before I was fired. And here we are. This is what we were working on. And you can see that it's not finished. It's got lots of tags. It's got tables. It's got lots of graphs. So here we can see what gamma.ai will do with this. So we'll upload a file. We'll click here and we'll upload the seed layer article. That's really easy. We open it here and then you can see we get into loads of options. Those options will not really dictate what's in there, but it will sort of dictate the quality of the presentation and give you lots of options as well. So here we are. This is the seed layer article that I want to transform into a presentation. And I click continue. And then they start configuring and working with your file. So here we've got the settings, which is what you should start with first. So you've got generate, condense and preserve. Here we want to condense because we've got a lot of content that we need to condense. And you can see it's got all of this information here. Now I have upgraded to pro. So I don't know why this one's down here. It was only $10 for a month. And here I've got card by card. So you can see here it's saying, well this card is based on this stuff. And then as we scroll down, blah, blah, blah, blah, blah, blah, blah, blah, blah, blah, one card total. Well, that's not what we want. So we'll go here to free form. And now we go max text per card. Now this is best for a presentation. We've got brief, medium or detailed. So we can go to brief because I want it to be super brief and then I can talk about it using my own words as opposed to just reading from a slide, which is really boring for everyone involved. And then we've got output, we've got AI images. We're not really going to be too concerned about these because it's going to produce images where it thinks it should produce images, but they're a bit rubbish. We're going to work with that a little bit later. And then we've also got image model and where we want to grab it from. Not important for us, I don't think. And then all we have to do is pick a theme. We go through, choose a theme that we like. I like this one. And then we'll say generate. And you will see it generate the slides live in front of your meaty eyes. And this is what it comes up with. So it's saying here, this is the title slide and look at this rubbish photo. No, that's rubbish. It's not very scientific, but that's all right. We're not using this for the photos. We're using it for the content. So here, metallic adhesive layers. So we've got that one. We've got optimizing silver thickness. And then we've got seed layer effects on reflectance. Then we've got morphology and adhesion. So it has done a pretty good job at separating out the things that we would want to talk about. And so then we've got adhesion mechanisms. We've got optimal seed layer and conclusion. So look, this is a good structure, but it's not done a couple of things. So the first thing it's not really done is put graphs in from our actual paper. So this is where we would need to grab this one, copy it across. We can put it into this slideshow by using the interface on the web app. So we can go here and just say, add an image, upload an image, and then put an image in. Or what I like to do is download it. So we can go up here and we can say export. And then we can export it to PowerPoint. Brilliant. And then I can work with that a little bit later. This is a really sort of like quick shortcut of getting design and getting the content and getting the story structure in one place. And then after that, you can work on getting the data properly presented in an outside program like PowerPoint. But this is not my favorite way to do it at all. This is my favorite way to work with this app. I don't think Gamma AI is very good at coming up with the story structure, a really robust sort of narrative that you can talk about during a presentation. So I prefer to use ChatGPT first. The first thing I do is head over to ChatGPT 4. This is the one that I paid for with my hard-earned YouTube money. And here I am. I'm saying I am creating a presentation for my PhD to attend a conference, give it as much context as it needs to do a good job, and need some help creating a structure. Can you help with this text? That was a bit rubbish prompt, wasn't it? But it did all right. Can you help with this text from a recent paper I wrote? Perfect. Okay, so we've got pathways of high throughput, low cost, indium three, transparent electrodes. And all I did was take this entire paper and all I did was go Select All, go Control A, there we are. And I just copied and pasted it all across into here. So I didn't actually like format it. I didn't sort of like say, I want the abstract, I want the introduction. I just sort of like mashed it all in and said, do your best. And it did do its best. This is why I like using this as a first port of call for creating the narrative structure and blah, blah, blah, blah, blah, there it is or there, there it is. Blah, blah, blah, blah, blah, blah, blah, blah, blah, blah. Come on now. Come on, what a long paper. Okay, we're at the references. And there we go. Okay, and there it says, for creating a structured presentation for your PhD conference based on your paper, you can organize into these slides, title slide, introduction, objective. And I was like, yeah, okay, this is very, very sort of like high level. Give me the actual content. So here I said, create an outline of slides for content with 10 slides. Good, another great prompt, Andy, well done. Certainly, here's more detailed outline. So here we go. This is what I was actually interested in, title slide, presentation, introduction, research objective. So all of these are actually, look, you can see it's got actual details in it. And this is what I wanted to use to put into Gamma AI. So I copied and pasted all of this across. So there's this option where you can go paste in text. So I went here, pasted in the text, and I went through that exact same process. So then this is what happened. I went here and this was the output. And you can see it is a much more detailed, I think a much more well-rounded story than the one that was a little bit kind of just like a little bit of a scatter shot of the content that I wanted. Here I've got a nice flow, a much better flow in my opinion. So we've got the need for transparent electrodes, which is like the background. We've got the research objective and you can see it's got bullet points and then it's got materials and methods. You can see it actually has a go. It's sort of like saying how I created these transparent electrodes, which I like. But obviously here, I would just put the schematic of the process that I used and that I generated throughout my research. So then we've got electrode fabrication process, which is great. Once again, I would just use the images that I've got, but this is a good start. Key results, so figure of merit, advantages, performance testing. This is with OPV devices once again. Good bullet points, but I can flesh this out much, much more with actual detail and data and then conclusions and future directions. So I think overall, this is a really great start. It's much better than just giving it the Word document. We've allowed ChatGPT to sort of like work with the narrative a little bit more. Lovely. And now this is what I did. I went and I said, give me this in a PowerPoint presentation and here is the PowerPoint presentation. Now, you can see that this is the PowerPoint presentation and there's plenty of space for doing other stuff and get rid of made with Gamma. No, I can't go. But you can see that over here, yeah, I've got all of this space. So I can play about with these slides and importantly, I can take the information that I want. So from this one, I can say, you know what? I want this schematic. So I can go in, grab that schematic if I wanted, capture it and then in PowerPoint, where would I put it? Just down here, materials and methods. I can say, actually, I want that in there and I can't do that because I need to save it first. Enable saving. There we are. Now we're talking. This is the stuff made of dreams. Can I get rid of made with Gamma now? I can. Yes, you don't even need to pay for the full subscription if you want to get rid of that made with Gamma. Okay, where were we going? Device electrode fabrication process. This is where I want it. There we are. So I can put in my thing and I can play about with the design. I can maybe even have this at the top. Oh no, this is going to take way longer than I thought it was going to. Okay, I can have that one at the top and then I can do that one underneath and perfect. We're starting to work with the data we've got. Sometimes you don't have the full story to work with. When you do a presentation, your supervisor comes to you and says, you're going to this conference and you're like, oh my God, do I have enough data? Do I not have enough data? Trust me, you've always got data that you can present and a story you can tell, but you can use AI to make it super obvious for you and this is how you do that. Let's say you've been working for a few months. On your PhD, you've got a few graphs. They're a bit scattered. They're a bit weird, but you need to do a presentation. This is what I would do. First of all, I would put in whatever I had as an image file to chat GPT for, the one you have to pay for. I would upload these. Here you can see I've got four figures. This one is AFM data. This is IV curves. This is wavelengths of certain materials after they've been heated. These are just random things I've put in there and I've asked this. I said, I'm attending a conference and I have to give a presentation. Here is the data I have collected. Can you help me craft a compelling story from these figures? Then it does such a great job. This is why I love chat GPT vision. It just knows what these are about based on the text in the images, which is just fantastic. Here it says, it looks like you have a rich set of experimental data. Thank you very much, chat GPT, saying all the right words. It says, these materials are commonly used in the production of OPV cells, which is correct. Your data includes atomic force microscopy, current density voltage characteristic, UV viz, and x-ray diffraction patterns. Here's a story that we can use. Here we go. We've got introduction to materials, objective of the study, experimental methods. This is what I would use to generate the first draft of my presentation. And if you can see down here, I said, can you give me the content for 10 slides based on the structure above? This is what I wanted. I wanted the actual text that I would put in here. So it says, certainly, here is a detailed breakdown for a 10-slide presentation. And all I did was take this. I copied and pasted all of that in for 10 slides. 10 slides, one slide per minute. So that's what I was working on for a 10-minute sort of like quick presentation that quite often PhD students give because you end up with like five minutes for questions. Too many people put too much information in their presentations. They talk for too long. They don't have time for questions. Make sure you're on the little bit of a shorter side. Trust me, it's much better to have time at the end than to rush through your slides because it just panics everyone in the crowd. So all I did was take, no, back to this, by the way. I would take all of this. We'll copy it across. So then I would go to new, paste in text. Then I would paste in the text here and I would say presentation and I'd say continue. And then we'd wait. We'd allow the AI gods to do what they had to do. Oh, that was quick. And then we've got this, generate. Okay, no, I want to condense or I want to preserve. So reformat while keeping your original content. Let's try that. No, we'll always try our best to preserve your content. We can't guarantee an exact match. Excellent, thanks very much. Then I've got presentation format. I don't want car size to be fluid. I want this, exporting to PowerPoint or Google Slides. And then I'm going to say continue and I want, let's say it's 10 cards, there we are. And we're just going to, oh, first of all, we're going to do the design stuff. Okay, here we go. What do I want? Oh, it's professional. Let's have a look at professional. Okay, body, that one. That's pretty professional. I like that one. Boom, now we wait. We enter the AI god realm. Oh, here it is. Spitting it out in front of me. It's doing such a good job. There we are. Slide one, already done. Slide two, already done. Slide three, slide four. And this is how it works. So I'm very impressed with how it can just sort of like mash stuff together to make a presentation. And then what I would do, like you saw before, I would take this and I would then take it to PowerPoint and I would work on it in there. This is how I would use AI to create academic presentations. Super easy. You should be thinking about using something like this too because it takes away all of those first steps that are really annoying and hold you up. You can take this, work it in with it in PowerPoint and start working on the important things, which is generating the story, putting the graphs in, making sure it's obvious to your audience exactly what you're talking about. That is what you need to do. If you like this video and you're doing a presentation soon, check out this one where I talk about how to deliver a killer presentation in academic circles. Make sure you're remembered for all the right reasons. Go check it out.