DBTB INT Greg Lindahl

so hi my name is Greg grundle an engineer at the Internet Archive being from the internet archive it's a lot of people know us and use our collections extensively and so the the biggest fun for me is meeting people who I haven't met before but they're fans of the archive long-term users of the archive and so I love being able to quiz them on how they use the archive because we have this big push to make all of our collections a lot more accessible than they are today we've been very good historically at gathering bytes up but not so good at delivering them to people who want to look at them and so the ability to talk to to end users especially technical ones who can comment inter professionally on our UX and UI issues is worth its weight in gold to me so my background my previous startup was a web-scale search engine and the Internet Archive has you know that and a lot more so not only do we have 12 petabytes compressed of web archive going back to 1996 which is a really beautiful data set and it's really fun working on making that accessible but we also have a large collection of video and TV and for me the exciting most exciting thing other collection is our book collection so we hope over the next five years maybe to scan as many as 10 million books and building exploratory tools for that book collection it's like the web only much higher quality and so even doing really simple things like taking the sentences from books and finding all the sentences that mention a year and extracting the entities that go with those years just it's a fascinating data set and and has you know incredibly huge value as an exploratory tool in a way that no library has ever often offered to the public before and so know that data exists at the Internet Archive and that's and I'm just fascinated with all the different things I could do there so a lot of people who come at data science have a pure computer science background and so they they may or may not have had a good grounding in statistics which is super important to understand in data science if you don't understand statistics you're doomed to make foolish recommendations and I've been lucky that the data scientists I've had the pleasure of working with all head very firm grounding and stats and that's one way and which having a hard science background can be very useful so really the people I encouraged to be data scientist so that folks who have a grounding in a hard science or economics or something that's that's very numbers driven and then trying that you learn to use the various tools that are available today as scientists now which are much more powerful and don't require a CS degree to understand so that when I am running around hiring data scientist I tend to focus on people who are like me which is probably not the best thing in the world but but they seem to work out really well so so don't think that you can't be a data scientist because you don't know a ton about computer science the more important thing is to have mastery of some corner of science that uses statistics to the point where you really understand statistics that's in my mind is the most important thing for a data scientist you