Speaker 1: I think research has just got a lot easier because there are now AI tools that can read multiple PDFs. It is like a second brain for your research. You are able to get together all of your references and use one prompt to look at all of them. Let's look at the apps. The first one is called PowerDrill and it's all about bridging your data and AI and this is a very very new app and it's only really in beta mode at the moment so go check it out. You can get started for free but ultimately you build an AI knowledge base and then you can ask it questions about your data. You've got the ability to go past what ChatGPT can do because you're giving it your data and the one thing I love about this is that it's completely code free. I've been trialing it and I think it's so easy. There are questions around where your data is going, what servers it's going to so I'll show you at the end of this video or the next app a better way of doing it but ultimately this is such an easy start into testing AI for all of your data that you've got at the moment. So once you sign in you get these apps. Really we're interested in this one, Simple Chat. So you click on it and then you get an opportunity to ask anything but before we delve into that there's something that's very important and that's uploading your data. So if you go here to datasets you click on this one and you can see I've already created a dataset. You can create a new one by clicking up here but you've got Andy dataset and you can see I can just click here and then I can add data sources. I've already put in four of my papers and you can put in as many papers as you want. The annoying thing is it's one at a time at the moment or you can put a file and the one thing I love is that it's got PDF. You can put Excel or documents, you can put PowerPoint, Word, anything you want. We've got four data sources already in there so let's go chat with them. You can also go on settings and just change what's going on. The power is in this. We go to Simple Chat and let's go to a new session and we can say okay well I want to ask anything about and this is the most important thing is we need to change it up here from ChatGPT to Andy Research. Now that's my database that we're going to ask it questions about. So down here let's just see what it thinks of the research that I've done. So down here all I've said is based on the data set is there any obvious next steps for this research. We click go and let's see what it comes up with. So based on the data set provided the research could potentially involve further investigation into the final location distribution of SDS in the device. So this is a multi-layered approach to polyfluorine water-based organic photovoltaics and it's given me the source. I like that but you can ask it so much more. So I've just said give me a summary of my, let's spell that properly, research. This is so precise to my research that this is I think the next step in AI. It's just incredible. It gives you the references where it's actually pulled out the information and it's a way to make sure that you don't miss anything about a large set of files. You upload them, you put them in, you ask questions and remember this is a conversation so go backwards and forwards with the chat bot and I'm sure that you'll be able to uncover so much about your research. The one thing I love about powerdrill.ai is you can upload a range of different files and it is a really great start and a little bit of a taste of what AI will really be able to provide researchers in the future. I love it but this isn't the best way to do this. It's a little bit more complicated but here is the best way to do this at the moment. Now this next app is just incredible. We're talking about Quiver. Quiver is the second brain in the cloud designed to easily store and retrieve unstructured information. That is information that you give it. PDFs, word documents, text, URLs and everything you want to think about in your second brain you can upload here and the reasons I love it better than PowerDrill is because first of all it's open source so you know exactly what the code's doing. You've also got a secure app. Your data, your control, always I love it and you can store any kind of data. Quiver can handle almost any types of data you throw at it. Text, images, code snippets and everything. So you just upload your data and there are two ways that you can use this. The first way is to use it online. So you click here on their website, get started and the first thing you do is upload your information. I've already done that and if you click up here on explore you can see that I've already added two of my papers. Once we've uploaded some files we simply click chat. We say that we want to ask our default brain and the way we ask is we say at and then we say default brain and then we can ask an academic researcher for example by using a specific prompt we can ask it what is a gap that I can fill in this data and then it's going to weigh, it's going to think about it, it's going to ask my data set that question. Based on the provided data a gap that can be filled is missing value for the RAM and spectra and images collected blah. The data mentions the collection parameters and resolution but does not provide actual values for these parameters. The best way to run this is not using their website but rather creating your own local instance and it's a little bit tricky but I did it. It took me about two hours to kind of troubleshoot all of the stuff that went wrong with it. You go to their github page and you just go through the getting started. So you need to install docker on your computer, you need to sign up for super base which is how your data gets kind of chunked up and stored so that you can sort of ask questions about it later and then you go through these steps one at a time. Now I consider myself pretty sort of like good at using the command line and installing stuff. I've done a little bit of coding and in the past it was a little bit tricky to work out what was going wrong and how to do it but with the appropriate kind of like googling and troubleshooting I managed to do it. So if you're interested in me running a full video on how to actually install this let me know in the comments. This is where we end up. You can see this is running on localhost 3000 and I click on Quiver and now it's asking me to upload knowledge. This is exactly what it is in the web app but this time it's running on my computer and it's running on my data set that I have uploaded to super base. That's what it's called. I've uploaded to super base. This is it. You can see here that I've uploaded a load of my papers. Here they are. So after you've uploaded all your data you can go up here to brain and then you can see you've got some settings about this brain. It's name, it's description, you've got an open AI key which I'll delete as soon as this video gets published. You can choose the model that you're using, all of the settings and you can set a Quiver personality here. I've said that this brain should act like an academic researcher and I've used their basic research prompt. So here it's all saved and then simply all I need to do is start asking Andy's research brain questions. Exactly the same I'm going to say can you provide a gap in the literature that I may be able to study and then we're going to say chat. That's going to go query my data set and it's come back and said even though I don't have real-time access to the current state of the literature, based on the references I provided it seems that I could have a look at new materials, improved fabrication techniques, interface studies, device integration, durability studies. Durability studies are so important, it was actually the next step that I was looking at for my devices. You can see that this is only as powerful as first of all the data that you give it and also the amount of conversation you have with the chatbot. I mean no doubt that both PowerDrill and Quiver have given us a little bit of an insight into where this AI movement is going and I think it's incredibly powerful for researchers. The great thing about this is you can create new brains as well so if I've got something here and I want to talk about my YouTube channel and I can give it an open AI key blah blah blah all of that sort of stuff. I can select a particular personality that I want it to have and I can also set it as the default brain but I'll push create. I'll click on YouTube channel and then you can see my brain is empty because I've not uploaded anything to this particular brain. So I need to go to upload, I upload my files and we start all over again. So if you've got lots of different areas in your research field that you can sort of like segment and put into different brains I think that could be a really powerful way of thinking about the different areas of your research and querying specific data sets. So there we have it there's everything you need to know about how you can use the new AI tools for querying a load of different file formats creating a second brain for your research. Let me know in the comments if you've tried them what you think of them and as always I'd love to know what you think and also remember there are more ways that you can engage with me. The first way is to sign up to my newsletter head over to andrewstapeton.com.au forward slash newsletter the link is in the description and when you sign up you'll get five emails over about two weeks everything from the tools I've used, the podcast I've been on, how to write the perfect abstract, the perfect daily schedule and more. It's exclusive content available for free so go sign up now and also go check out academiainsider.com that's my project where I've got my ebooks, I've got my resource pack, I've got the blog, I've got the forum and everything's over there to make sure that research and academia works for you. Alright then I'll see you in the next video.