 I was talking to my father last night and I told him what topic I was going to speak on artificial intelligence and his idea for the title of this talk was artificial intelligence is better than none. I really thought about including that but I decided to go with these hashtags especially AI Lib just because I think you know we're all as a group we're smarter than than any any individual person and I know there's probably things that you know that can be shared on that hashtag that could be useful for other people so we could make this a combined presentation all of us are doing it together so I'm I'm Chris Erdman I work at the NCSE Libraries I just joined about a year or so ago I'm chief strategist for research collaboration and so I have an interest in in artificial intelligence and how we were we're working with our research community to start I I so it's it's not intimidating at all to talk about artificial intelligence is such a short amount of time but I have the easy task of doing a little bit of background for all of you and one of the things I I found I I come across things on Quora often but this quote from Monica Anderson but to make computers easily do all that which comes easy to humans as a definition for for artificial intelligence I thought it was pretty simple explanation there's online you can go to this what is AI for people in a hurry I thought it was pretty like con style way of demonstrating all the different aspects of AI which is really an umbrella for all the different things that you must must have heard some of these grab bag of terms but you know really send centers this explanation of AI around people but really AI has been around for a while it's nothing new if you talk to CS people they'll kind of roll their eyes I think they'll why are you interested in it now and I think the reason why we're so interested in it now is that the high performance computing power is readily available it's it's there's there's more and more data to train these these are these AI systems there's democratization of AI skills so you see a lot of training online even on YouTube for learning AI and in a day or whatever you know whatever they're pitching and then proliferation of AI products and solutions these are a lot of the platforms that are being used now to conduct AI and you know just again going back to that yet let's use an example to explain AI I think a really good one that recently came out was this quick draw which looked for patterns in how people draw doodles and and it was it was done on the Google platform and they were able to just collect this massive data set to train their models on with system you probably have heard of TensorFlow but you know they Google was able to leverage all this data and able to come up with a predictive algorithm to to to really understand what people were about to draw before they even finished it another example that was in the news recently is Kepler the telescope space telescope will have an announcement on Thursday that they found they potentially found something in their data set using machine learning so really it's being used more and more in the library context we so at NC State and I'm sure every one of the other libraries out there we're not unique in the sense that we're turning to our collections and thinking about how can we implement some of these it you know AI approaches so this is I'm thankful for these for this group that put out this really a blog post on how they use different tools out there like clarified AI to some of the Google and and Microsoft computer vision services but as you can see this is a great example of hey the algorithm got that person that individual right but he's not holding a laptop and that's not a pizza in that in that one in that one abstract painting and so that that really emphasizes the fact that as we experiment with these AI solutions the underlying data is trained and it's trained on other data sources so I really wanted to focus my talk on our spaces in AI this is something we're particularly proud of at NC State we have amazing spaces and one of the examples that pops up recently is this project called digital life decoded you may have seen it advertised on Lita it'll be a presentation tomorrow a webinar but really this is a number of our fellows that got together and decided that we needed to really educate our students on how their data was being used you know how how machines were sort of interpreting what kind of person they are and so here's some of the examples that they used especially electronic frontier privacy tools the other thing too is I think that was a great example of of engaging with our community and and really working with them to to understand the dangers of how our data is being used and here's a plot project that ran in New York and in now London funded by Mozilla where people can actually go to stations where they can see how their data is being used I think they have physical representations to digital one of the physical ones is they have volumes of passwords that have been stolen online printed in volumes and so it and there's a there's a detox station for all the people that go through this experience saying the same thing I I I need to do better about privacy of my data and you know just so I think as libraries we can really we really can do things in this space we have an amazing maker space at NC State and one of things that's that's happening there is that some of the students are doing projects with some some of these AI services like Watson but recently we've also come across this do-it-yourself artificial intelligence kit and some of the ideas coming out of there are amazing sort of a voice recognition system for working with our bookbot or you know capturing surveys and transcribing them into you know so that people don't have to input in anymore if that you have hand written ones so we're we're going to experiment with this I think it's a great project there are more kits to come that Google is sharing we have a research computing series that we started not too long ago and and some of the things that are coming up in there that some of the researchers are interested in in understanding how they can implement their AI approaches on on our clusters with the GP GPU computing that we have on campus this is some of the topics that are coming up in that vein but one of things I talked to a computer science our computer science department at NCA NCSU and they often tell me that that that they're short on data ethics like that their whole shelf is filled with programming books and and all different you know all all different sort of approaches but then there's nothing really on their shelves as far as data ethics and they they admit that they that this is an area that they need help with and my colleague Bonnie Terahina at data and society just showed me a post today about how Google needs this help as well it's some something that CS people come out of school and don't really get exposure to and and you know you've seen many of these stories it's just one of how that underlying data has inherent problems you know there's examples of this data being taken and that in some ways being presented to certain communities and and those communities not being served up the same things that that other communities will are served up and so this is a place where we really need to help our community be a bridge and then this is from my this the university I graduated from University of Washington a great program that's that's teaching students now about calling bullshit about looking at the underlying data and looking at the models use and and and critically thinking about what is being presented. We also have a program called peer scholars where we work together with our graduate students and they come to us with ideas and one one of the ideas recently was to create a deep learning machine learning modules on on data camp and so just recently I don't think they've been uploaded yet but we the student plans to share those with the libraries on data camp which is a service being used that you used more and more for data scientists and people in AI and I mentioned this in our previous session about data science in libraries that this report from more more Sloan data science environments that really makes the point that their experience of working with librarians and using their spaces that they make the case that librarians are definitely needed in this space that we are excellent for for bridging some of these big questions. So these are some newsletters that I I track I you know I'm I I'm trying to track this as much as all of you but I really appreciate these new the newsletters especially from Laura Noran and I another one too that is called humane AI which really talks about that the social side of AI is done by Roya Paxa and I encourage you to go there and and potentially sign up for those lists serves there I find them immensely valuable they're a they're a read but the great thing about these newsletters is they paraphrase what's happening in the post so you can skim through it and get a general sense and there's a new initiative called partnership AI where it's some of the big companies thinking about AI and they also have at least a Twitter account that you can follow on and keep up with what the big companies are doing as far as sharing what they're doing in AI space and this is this is someone I know personally Lynn Charney she she taught and data scientists training for librarians she's very creative she's she's big in the open Viz community the visualization community and she thinks creatively about how AI is used and and I appreciate following her for those kind of ideas like using Minecraft to teach in this case and then just recently there was this conference nips 2017 or nips 17 I showed this picture because just look at the rock star stadium now for AI and this is this is what is there's there's a lot of people in this space now one of the impressive things that happened I mean there's more work to be done but in this latest round they had a black in AI group and then they had women in AI there was more prominent and I think that's a good sign but that's a great resource to track and see what they're talking about as well and get get a better sense of what's happening in this space so thanks move on to all these people help with all these different services so thanks to them too yeah thank you Chris we're gonna allow plenty of time for for questions it's not full but it's okay my name is Karim Bugida I'm the dean of libraries of the University of Rhode Island I wanted to talk about AI for a very long time and I will tell you later why and I was glad that Chris Erdman and and North Pickering agreed to join me today because it's a very important topic and everybody's talking about it so it makes sense that you know within our field library information science and now even data science that we have to pay attention to it so now I always show this because people confuse Rhode Island with Long Island we're not part of New York okay so we're the smallest state actually there is a funny meme saying if everything is big in Texas does it mean everything is smart in Rhode Island so it's not smart we're thinking big too so this is why we're talking AI so within the last two years good things are happening at the library this is a little bit of context so I presented on last time CNI and New Mexico our initiatives around big data and data science which is under the umbrella of the library and also previously a year ago presented our also the maker space and we have two new things happen URI spark which is the entrepreneurship and innovation within the university it's within the library data spark data spark is a new data analytics program that we acquired it was part of another organization it's a it's linkage with Rhode Island and beyond there are 10 people in it and the newest newest one is get a new grant around design thinking this happened just last week so this is why we're talking about AI and here I don't want to define it at all that same thing when one Chris presented data science it's a whole spectrum and it's normal that it's work in progress and there's something what we call the AI effect is that when a technology is mature enough it's no longer called AI in fact like OCR optical character cognition is no longer AI a spiritual cognition and actually one of the leaders actually now came from the spiritual cognition and signal processing field so whatever it means you know it's important you to pay attention to it the same thing what is data science and that's back to your presentation in our earlier and by the way NSF is funding what is color shaping data science and actually Columbia and Brown got a piece of it which means that it's really really work in progress and that's fine because we as in the library archives and museum field we're not good in handling ambiguity so we want things to be totally defined and structured and and we have to get away from that like thinking okay it's okay it's a new bandwagon you think about it what scope it and be agile on it so this is why it's okay because sometimes we didn't this field you have everyone is talking about it and you say okay is this really AI or not it's okay but you know it's important so time magazine specific edition New York or two and it's around robotics and the impact on robotics on humans popular science the economist this is toes out last week this is why why we're thinking about AI not just in the library but the campus and and nation why it's because of what's going on in this field the student literally there's another slide I'll show it later our student literally before entering they have a wish list of what kind of classes they want and I was a top 10 request so this is why we we've seen the library as kind of a hub and intermediary say not a neutral space not at all because we're not neutral at all but look at this the top one top jobs nationwide it's huge data scientists so this is this is a regular job customer success big data developers too so we need to prepare the new generation that the reason why I want a I in the library 2002 do you remember this movie it was in New York public library and this is the holographic librarian and basically he was a reference library and and later in the movie the whole library disappeared what's left is only this guy so he he kind of was answering any question so this was kind of you know the impetus to to think about it I mean technically was not ready back then this is a nude it was the trailer was out last week why I'm showing you this because literally in the trailer it's all about the library in Cincinnati the movie it's happening there's a new movie by Estefan and it says the library is the last best scene of democracy the police because the part of the story is that it's occupied by homeless and then the librarians were stuck between the homeless and the administration what's the role of the library here the same thing so it's important for us because we have a social mission so whether AI or not we are part of the democracy of the social system and values back and Cliff Lynch mentioned it earlier and many of you that our values are the same AI now is more on like the technology side and Chris mentioned does the ethics side it's it's missing and we need our voice to represent the communities around that so I mentioned that already this is our neighbor Brown University just basically in November 7 that they're struggling to keep hype with with the training and and curriculum literally like the CS classes used to be like 16 17 now it's more more than 100 so there is a huge demand and the market we have we heard it from Rhode Island employers they want folks with AI so what you should do about it was the role of the library we're not just like waiting to be in a passive mode so we have to do something about it so a little bit of history do you remember this I mean 2014 the reason I'm showing this because when they ask at the head of the library Westport Connecticut why are you having those small robots in the library they said it's the same one early 80s we had computers on the library because the library was the only ones so it's part of that movement and do you remember also it was in the news February 2016 you the robot librarian so and actually there was a session at computer libraries in last March where they were talking about this too so this is specifically they designed in in UK a robot that will tell you where to find where to find things so that's that's a very interesting and then I'm citing here Chris and Chris is here hey I'm citing you because this is really really important like AI can magnify existing system of inequality and racism sexism homophobia and the like this is why natural is I would say we're the voice of the voiceless and the voice of reason because the US specifically is a very capitalist society so AI is going to happen whether you like it or not and it's happening already so you have to have kind of control and part of it is also as university we need to train also the student you have a voice you have a role even if you work in companies where the motive is basically making money but you have a voice in making kind of some embed some values in it this was just last Friday I was pleased the one I saw it so if I don't know whether actually Amazon is aware of this or just one programmer decide to make it but when you ask is about black lives matter so Amazon Alexa Niko they have response and feminist response that's interesting and actually the whole I would say racist and Nazi are freaking out about it so and they need to figure out because it's not neutral at all and I'm glad there's already something that shows you this so there was something also Ruth you presented at this event to around AI means now it's it's it's no longer really the cat in age it's almost is going to be mainstream at some point so now us you are I so we start to process in March start talking to internally with with some faculties some librarians and also somebody mistresses or how but if we apply for a grant to get to get AI live in the library and it was good good response like people were really ready for it so we applied in June and we got it recently so we got the grant to have an AI lab in the library and I believe this is the first in the US to have literally a dedicated lab lab inside the library so this is the team so it's it's really diverse as you can see so we have Kunal from Department of Electrical Computer and Biomedical Engineering we have Sheryl Foster from Department of Philosophies because of the importance of ethics John Peckham computer science and also she has another role within the library as in charge of data science and the big data collaborative Harrison Decker data librarian for with us your libraries Angel Faria is the manager of the maker space and myself so this is the the team that it's the same thing this is an old ethical dilemma that we need to we need to face and especially now with the driverless cars this is basic basic rendering of of of the lab so we have we repurposed two rooms in the library and it's initially around I would say demystifying the concept and it's not a it's not research oriented because the Champlain Foundation is very educational so they wanted to really hands-on and prepare undergrad and grad for the new world so so we have a bunch of toys robots and and so forth maybe we're gonna get some drones too and we have few stations those are stationed with GPUs and kind of space where you can have a conversation like around whether it's seminar or one-to-one consultation and this is a basically the the idea we come up with so we have zones so the first zone is you have workstation and if you can if you notice here it's a premium space you can see it when you enter the library literally you can see it and it's heavily heavily used by so I want them to be kind of exposed to it be intrigued and ask so those are the two rooms next to it so and we have also this kind of you know the the level for beginner medium and advanced and we have the first zone workstations and second zone hands-on project zone three the half of collaborative thinking and also this is also for kind of debriefing and also for for for staff so that's basically it's not big big wrong I mean it's a 600 square feet but we prefer this room because of the location so it's very important those are for now I mean we want to have NVIDIA initially but NVIDIA has been changing the spec and they said we're not ready to release it yet so we may change our mind whether we want to go with lambda or not and the rest is is very familiar and the professor Kunal is also the expert in term of IOT is why we have to add the internet of things in it too those are the computer the robots and that's it