Speaker 1: ChatGPT now has the feature where you can look at custom GPTs. Check out this. Discover and create custom versions of ChatGPT that combine instructions, extra knowledge and any combination of skills. So there are loads of them. Check out all this. There's featured ones, all trails, consensus, books, culture and all of these ones. But I'm interested in the academic apps. So in this video, we're going to have a look at all of the academic GPTs and whether or not they're useful for you. I went into this with a skeptical mind and I'm still not convinced of the power of these custom GPTs. But here are the best ones that I found. The first one is consensus. It says it is your academic research assistant. You can search 200 million academic papers from consensus, get science based answers and draft content with accurate citations. Now this is based on consensus, this website here. Now the one thing that really interests me about the GPT as opposed to the website is that if you go to where I was before, which was the plans, you can see that here you only get a certain number of GPT-4 summaries. Whereas here you get unlimited GPT-4 summaries for $6.99. But in the chat, if you go to about, you can see that it's powered by ChatGPT-4. So does that mean we get loads of free searches? I don't know. It's starting to get so confusing. There are loads of things you can do. And I really like looking at these examples down here because what it gives you is the ability to kind of get a snapshot of what each one of these tools can do. So I went and I asked it, draft a blog on science-backed benefits of mindfulness. That was just what it recommended. And so it actually talked to the app first, which was interesting. And then it brought back this really basic blog. I actually quite like asking ChatGPT to create like a blog post or a summary because it helps you sort of like just get a snapshot of the most relevant information quickly. So you can see it's done a pretty good job. It's got different references here from 2018, from 2015. So overall, this is a pretty good thing to be using if you want to sort of search semantically for papers to reference in your literature review, for example. But it may come with the added benefit that this always works with GPT-4 and you don't have to pay for it. I'm not sure. Let me know in the comments what you think. The next one I've not seen before, but it's Scholarly Assistant, your go-to for detailed academic advice with specific citations. So if you want to find resources, if you want to kind of find that extra bit of information to bolster up your literature review or your introduction to a paper or your thesis, this is what you can do here. You can find resources. You can suggest improvements for a paper. You can ask it for information about something. I went ahead and I said, I need information about new materials used for nanocomposite transparent electrodes. One thing I noticed when it was searching, it said, you know, new from 2024, that's good. So here it gave me three different recent research. Now I was always skeptical on ChatGPT's ability to get recent research. It always like looks further in the past than I want. So let's have a look. Nanocomposite electrodes for solid oxide cells. I'm going to click on this. And where is this from? This is from 2023. Good. That is what we want. We want recent stuff. So this chat, this GPT could be really useful for finding recent papers, which I like. Now let's check the next one. This is from 2019. So not as recent as I would want. And obviously I can go in and say, find me papers from 2023, from 2022 that do this. And I think it would probably be able to do that a little bit better than just saying because depending on the field, recent means different things. But overall, yeah, this is a great way of finding different papers. And it can do other things, like I said, like find resources, suggest improvements. Go try it. And let's look at the next one. Next one is Paper Interpreter. This one says you enter the URL or upload the PDF of an academic paper and then it explains it in a comprehensible manner. Let's take a look. So you can actually ask, how can I use it? And then here is a link to a paper. So I clicked that link to a paper and here it just said what the title was and it gave me a summary, which is what I really like about this. A short abstract, a background, methods and results, discussions, limitations, possible applications. So it gave me everything I would want to know as a snapshot from a paper. I've seen something like this with Paper Espresso from, I think it was Paper Brain, something like that. But this is a really great way of getting that snapshot. And it works really, really well. So all you have to do is put in a URL. But I was thinking to myself, can you put in a URL from Sci-Hub, which is definitely something you shouldn't use to get free access to loads and loads of papers. Now, I wanted to see if I could get this. This is one of my papers. I've already got access to it. I'm not doing anything illegal. And I put it in and I said, find me this one. It said, thank you for your request, however, I'm unable to access or retrieve content from websites like Sci-Hub due to copyright and legal considerations. Boo. So if you're definitely not using Sci-Hub, like you definitely shouldn't, you should definitely not use this one as well, which is Anna's Archive. I found this recently and what it does is it definitely shouldn't be used to get different papers from a range of different places you definitely shouldn't get papers from. So if you go in here to the Sci Database, which is in beta mode, all you have to do is click here, put in the DOI and then push open. Then it will go away and it will find that DOI, which you definitely shouldn't do to get the paper for free without having to pay for it and give all your money to those horrible, really nice journal publishers. So on the side here, we've got the DOI at the top and then we've also got where we can download it. We also can see it in Sci-Hub and we can also see it in doi.org. And in the past, this one popped up. Actually, it's still trying. So the paper should turn up here if it can get it. If not, we can just click to Sci-Hub and then it definitely goes to a place where you shouldn't go to get it, which is Sci-Hub, which is where you definitely never should end up because it is really bad, isn't it? So bad. I'm feeling bad being on it right now. Oh, I feel so bad. There it is. It's popped up. Boom. Another one you should know about is Scholar AI. So this is Scholar AI, AI scientists to generate new hypotheses, analyze text figures and tables, blah, blah, blah. One thing I really like about this is it seems to be able to do a load of different things. It goes away and finds papers based on the URL or the DOI you give it, which is brilliant. So I wanted to test this thing in particular. What is the experimental setup in this study? And then I gave it a link to an archive paper and it's great. So first of all, you needed it to sign into the Scholar AI plugin, which was really easy to do. And then it went away and it did this. So it looked specifically at the experimental setup. One thing I like about this is it really seems to know what it's doing. And so here it tells me exactly what it's been up to, all of the different components of the methodology. And then it went and got a figure, which is just incredible to me. I don't know whether it was like it went to get the paper and then cut out the figure or it grabbed the figure from somewhere else. But overall, being able to not only just have a complete rundown of the methodology and a figure, which it sort of like knew was super related to what I was asking, it's actually incredibly powerful. So I'm watching Scholar AI at the moment and it's GPT because I think this could be one of the best ones to use in the future. The next one is Scholar GPT. They all start to sound the same in the end, don't they? All right. Enhanced research with 200 million resources and built-in critical reading skills. Now they're really leaning into this like critical reading skill thing. So I was really interested in that. So down here you can see samples of the sort of stuff it can do. The one thing I really liked about this is that you could type LS to list my built-in critical reading skills. So if you click that, it'll actually go away and say, okay, well, these are the things you can actually do. I wish all of the GPTs had this function where it was just like, let me know what I can do with you. And then you can just sort of like see all of the stuff that it does. Anyway, I use this one to upload a PDF and use critical skills to read it. So it went away. It was in here. So I uploaded my paper. It got the right paper, which is great, which is something that a GPT didn't do. We'll see that in a minute. But ultimately, it just gives me this really nice sort of like breakdown. It's not as detailed as the other one that I've shown you previously. So I don't like it as much. I don't think it's as powerful, but it is there and it does sort of like, you know, give you relatively good information, but it doesn't go into the nitty gritty, the stuff that I would be really interested in. So there's lots of room for improvement for this one. The next one you should know about is Academic Assistant Pro. Professional academic assistant with a professional touch. Hmm. That's not professional at all. All right, so can you help me with an outline of my paper on ecology? I need to polish a thesis abstract. Could you rewrite this paragraph to me be more concise? I was interested in all of sort of like that stuff. So I asked it, I need to polish a thesis abstract. It said, absolutely. Ah, they're always so polite. That's one thing I love about GPTs. I'll be delighted to assist with polishing your thesis abstract. So I actually put in my thesis abstract from my actual thesis, my PhD thesis, which is not here anymore. Here it is, it's over there under a load of stuff. Oh, what a nightmare my desk has become. Okay, here it is. This was from my thesis and I put in my abstract. Oh, this is so boring. It's at the beginning. There we are, there it is. I put in that abstract, which was a couple of pages long. That was worth it, wasn't it? So I put in my abstract here and down here, it says here are different versions. Great, version one. And then it's got the abstract, version two, abstract. So I like that it's given me three versions. I think three versions probably could be a little bit overwhelming. Because at one point I was like, well, what one is better than the other one? I don't know. But the fact you've got three to choose from is maybe a good thing. A little bit overwhelming for me personally. But then it said you can go on to write the next steps. Brilliant, thank you, I'll do that. And yeah, there it is, Academic Assistant Pro. I think it is quite good if you want to do writing and polishing and reviewing. And that's up next. The last one I think you should know about is Academic Research Reviewer, where they provide a comprehensive review for your research paper or thesis. This is brilliant if you're finished and you just want that little check. But does it live up to the hype? I don't think so. Here, you can review my manuscript, you can assess the data analysis, feedback on my literature review, please. And then rate my manuscript, thesis, or draft on a scale of one to 100. I actually put in one of my papers, which was the accurate thickness measurement of graphene nanosheets. And unfortunately, it just went to get another paper. Here it is, it says, a qualitative investigation of the occupational stress in the hospitality tourism caused by, in Bahrain, caused by COVID-19. That is not anything to do at all with that paper. So a massive fail. Don't use this one at the moment. There's also a network error. So yeah, overall, I don't think they sort of like have an issue with, you know, that's not their fault. But overall, it didn't even get the right paper. I uploaded it. So Academic Research Reviewer, ooh, big fail at the moment. So there we have it. There's everything you need to know about ChatGPT's recent release of public GPTs and the specific academic ones that I think you should be looking out for. Some of them are good. Some of them are not so great. Some of them are just plain rubbish. Let me know in the comments if you've found a particularly good one. I'd love to know about it. And if you like this video, next one you should go check out is this, where we talk about the epic things that ChatGPT Vision can do for researchers. I was blown away by its power and how better it is than anything else that I've tested vision-wise for academia and research. Go check it out. It's a great watch. I'll see you next time. Bye.