A Free AI Connector to Find the Closest Research Papers (Full Transcript)

[00:00:00] Speaker 1: What if you could type your research question in plain English and just a few seconds later have the five closest published papers in the world to your idea, with links, with citations, and be completely free? Ladies and gentlemen, in this video, I want to show you something I've been quietly working on in the background. If you check out one of my latest videos, I've been curating, working with AI tools to help bring you the best AI powered workflows you can use ethically as a researcher, the same ones I, as a professor, have published, reviewed papers for a living for over 20 years, use myself and use with my researchers. But what I showed you involved paid proprietary tools to search literature, and they're great. They help solve the hallucination problem that you all know about. People have scanned recent papers and found a huge quantity of hallucinated made up references. By integrating connectors with your AI use, you give your AI tools vision into real peer reviewed literature. But the problem and the catch is that they cost you money. So in the spirit of open access, which I'm highly committed to on this channel, I'm not getting kickbacks from any of the tools that I'm recommending today because I made them and they're free. I've created a tool that you can now do this and I'm going to give it to you for free and I'm going to show you how to use it. It's power and you can check the description below for details on how you can set this up for yourself. So this session is going to be interactive. I'd encourage you to follow along and take some of the same steps that I take. So let me show you how it works and then I'll show you how you can set it up for yourself. So many of you are going to an AI like Karen Claude and just saying search for literature on something. Let's say here I'm going to use an example, electronic vehicle battery usage in developing countries. Okay. And I'm going to specify here, the difference is I'm going to specify to use our tool. You guys, when you go, you load this up, you can just say search the literature, it's going to know to use the tool. So using fast track, open literature connector. Okay. Forgive any, uh, infelicities of spelling that I might have talking and typing for some reason they're on different neural circuits and it doesn't work as well. You can use any of your models. I'm using fable because I've got it, but this will also work well with any of the different models. I won't get into this while it's pulling up where you can see here, it has invoked our literature connector tool. That was the little icon that pulled up and you see what it's doing in the background is it has realized that I didn't use the standard terminology. This is something that happens a lot with novices when they're searching literature. This is not the terminology that's commonly being used, and this is kind of the power of using AI. It can call and search and cultivate your search to get the right stuff that you want. And look, it did actually three sweeps of calling the literature. This tool is tapping into over 250 million papers that are free and open access. So bear in mind, it's not the full set of literature. I've got a more robust engine inside our mentorship communities, and that's just because we have to pay for each search, which is why others will charge you. But this is already fantastic as open access and free, and it's going to massively improve the vision that your AI has into research. So three sweeps done, and here's what I found. So look here, and what I love about this, I've set it up so it takes you straight to the paper. So if I click here, open link, it's often going to be worried that because it's coming from Claude, you might have to do a human verification text. It takes you straight to the paper, which I love. And the core of the topic, it's come right here and found some of the key papers, which is really, really nice. I really love this. The other thing that I've got is high citation papers. So it does exactly what I like to do in Google Scholar is take me straight to what are the core papers in the field and how are the citations evolving. So I can see, look at this, this is a young, fast growing niche, not a saturated one. I have intentionally set up this search algorithm to not just do something basic to go pull article, but to help you see what is hot and what is emerging. So it's giving priorities to papers that are growing in citations. You might have an old paper that's just died. It has a lot of sites, but it's not getting cited anymore. So this gives it a dynamic look at how the field is emerging. Like I said, the best way is to show you how this tool is going to work. And it is splitting into different conversations, stuff on adoptions, stuff on second life, third life, recycling, other things. This will help you map the field. We can go further with this, but I hope you can already see this is giving you a bunch of real literature. And it's going to tell you when stuff is gray, but still potentially useful. We often want to look for something called a nearest neighbor paper. So this is set up to run our duplication test. It's something that I always want our researchers to do to find the paper closest to theirs. I see so many researchers who don't know what that paper is. And that's really critical. It's an anchor for all of your research project for calibrating your gap and your value add in the importance of what you're doing. So let's say I want to run a duplication test here. This is another tool. And I'm going to show you another tool that I built into it of mapping your topic debate that you can have. I hope you already like what these results are coming out with. So let's go here and say run a duplication test on a potential systematic review, looking at the determinants of electronic vehicle usage in developing countries. Again, my spelling's all off, but the model is going to fix that. And then I'm going to have it map the topic debate, which does something very cool, because what this can do is it can look at citations and figure out who is citing who and find common ancestors. So it can find reference points in the field, which sometimes can be a clue where a good debate is. Sometimes they're all citing that paper because they're debating. It's a debate because they're citing it because that's been settled. And that's the new reference point in the field, but you still want to know that. Okay, now this is started. This engages with some of our skills, but it's running duplication test here. You can see this going here with the open access tool. I've got some descriptions about how to use our tools. The one I just pulled up is available for free for you in the first video that I've cross-referenced here. So here you go. It is running here. It's looking for systematic reviews to compare against, and it's going to run for a second. Let me just come back to the output here. This is going to focus your reading list. Again, what's different about other search tools is that I have optimized this to get you things that you're going to need as a researcher in your research journey. This is built by real researcher for researchers. So let's go. See how it's also, I really like that it's saying that the citations are rising fast for its age. This tells you that this just got that hotness. I love, I love this. When I compare to Google Scholar, let's say I did this in Google Scholar, determinants, electric vehicle, battery usage, you see it does pretty good. It's going to probably find some similar papers, but you see it didn't go through and adjust that you might be using some non-standard terms, and it's going to get some papers, but it's not giving you that same kind of synthesis of what you need, where oftentimes you're doing initial searches and trying to figure out your originality. Now it's also really great for forensically finding a paper when you need to do that. But let's come down here and see what it comes up with. Here we go. In developing countries. And I didn't specify, I just realized here, I didn't specify developing countries, but that's okay. It found one in South Asia. Look here. I found one in drivers and barriers. It's a preprint, so it's not published yet. So let's get curating here. Here's a systematic review of this global scope, and it's telling you, you need to be able to add over and above these papers. Guys, this is one of the common failure modes I see, especially with junior researchers and the duplication test. It's saying adjust before proceeding and specify what your value add is going to be. Some of this power again is because it's tapping into our topic validator skill, but it is running the duplication test. Guys, it's the worst when I see people just duplicating a paper and they don't realize it and they only find out later when they're getting lots of desk rejects or can't publish it. So the other thing I want to show you is the, okay, now map the topic debate on, on this potential topic. And this is the one that's going to go and look through the co-citation patterns. Here we go. Map topic debate for common ancestors. I'm next going to show you how to set this up. So we'll just run this in for a second here and hope I'm not making you dizzy going up and down. But the first time I came back thin, what's great about this is it's tapping the full power of Claude because now it has greater vision into the literature. It's getting the citation counts so we can do more with the real literature than it could do before. I'm jealous of you guys. When I was just starting out, I wish I had these tools. This is a great use of AI because sure, if I had the ability to assemble all this data, run through it myself, I could do this, but a machine can do it so much faster. So here we go. Shared root is getting us the common ancestors and here we go. We can see that where the papers are clustering around these ancestors. So this is also going to focus your reading. These are, it's going to tell you the citations, the big things to read here. And this is nice. This is, this is really nice. These are some anchor papers in your field. Here's another shared ancestor. A lot of them appear to be using structural equation modeling, that's SEM, and saying have a look, look at this and where the debate is, is live. So how can you set this up? I think you're going to have a lot of fun playing with this. Feasibility test is next. It's spotting that. All you need to do is follow the steps that I'm going to show you here. So what you're going to do is go into customize. I asked a bunch of researchers using AI if they had even been hanging out on this panel and 90% hadn't. Go into connectors. Now with Claude, you get one connector for free. If you're going to pick one, absolutely. I recommend ours just because of the power that it has. You're going to go into add custom connector and call it FastTrack Literature Open. You can call it whatever you want, but this is what I recommend. And just type this in. I'm going to put the link for you below, literatureresearchfasttrack.com forward slash m c p. You're going to add it here and you're going to see the different search tools it has, can check gap saturation. It's going to get to the journal profile, which we use in our systems to help you pick the right journal for you. Avoid sham journals, figure out which ones you don't have to pay for that are the best in your field. Mapping the topic debate and run duplication test and search and getting papers is what we just used. I recommend putting on always allow for these, otherwise it's just going to prompt you when you use these tools and that's it. When you go for search literature, you're going to get our icon here and it's going to search using our tool. Like I said, we've got a more powerful one even inside our membership communities. And I wish I could make that free. I'm still working on it. It's just because it limits the gates of how many requests can be made per second. And if somebody is making the request, then the lane is blocked and it can't be used. So we have to limit access to that. It's completely free for our members, but this open access version is fantastic. It's going to massively up your AI research workflows immediately. You can use it today. I would love to hear from you in the comments, how you're using this yourselves. And if you're loving it as much as I am, because it gives you exactly the things you need as a researcher in your research journey. Use it also for forensic searches when you know exactly what you want. For example, if you know you want a site to reinforce something in your paper. For example, I need a citation on how much the prevalence of autism has risen in the past two decades. It's really great. It's going to go like a laser beam right to what you want in the field. And it's 100% free guys. I'm committed very much to open access here and getting you the support that I wish I would have had when I'm in your shoes at this stage of your journey. Stay tuned. I'm going to keep curating and even developing open access AI tools for you to use. And I listen and respond to all your requests and comments below. So if you comment below, it is me personally responding to you. Look forward to seeing you guys in the next video.