Transcript: Lucas Kuczera Interview

foreign I'm the founder and organizer of SF scholar Meetup here we are at launch Darkly invisible Auckland and the shore of Lake Merritt for our first post pandemic Meetup has been a few years and we're super happy to have with us Lucas koshera a founder and CEO of scholar c a company which helps clients develop code in beautiful programming languages and we have a talk from scholar C in the program on duct tape the utility set for skull productivity and we have several other talks and we are very thankful that Lucas sponsored this meet up so we can have this before recording thank you Lucas uh for helping absolutely and I didn't know your local until the recent list so it's great to have you in the community and yeah tell us a little bit about scholarship about yourself how I remember so an amazing picture you had this office in Gran school like yeah like you were sitting like your climb like was super awesome office why did you live in beautiful office right like I was in this office all day long like with a big screen so what brought you over and how tell us about your path right right so um so story goes back to 2000 I think 10 right now I don't remember exactly I went to some unconference and um I saw the Scana for the first time and I see I didn't get anything about it because the guy who was presenting was really bad I did so um then I I bought the the Scala book by Bill Banner's other scheme and I discovered that this is really cool language that fixes a lot of issues that I have that I had back then with Java and I immediately fell in love so I um I set out set up on this path to promote the you know start I started to use it in like day-to-day work at least as much as I could and it was really hard back then because there was not too many projects using scar so yeah I decided Well maybe at some point it's going to be really good to set up a company that's going to help you know promote the language and also back then in 2014 when we started it was really hard to find any Scala engineers and so initially um and and before that I was working for a startup from uh Silicon Valley so I learned how to work remotely and also I learned the startup World more and um yeah so that was a a like a background for a launching scholar C and my idea was to promote the language and since I had experience working remotely and also I knew it's impossible to have like Scala team in one place especially in Poland back then I decided to start remote first so we we did remote work before it was the thing and uh yeah that that's how it started and then suddenly uh this Scala started to grow in popularity really rapidly and the company grew on really really quickly so I had to quickly transform from you know being a software engineer my plan was to be okay we're gonna have I don't know maybe a bunch of scale engineers and we're gonna you know hack some scalar codes and that's gonna be it and the company grew really really quickly and I had to switch more into managing the company instead of doing you know software engineering which was you know a bummer but um I took their own and um um in 2018 still um I had this idea that well let's try leave somewhere abroad and California seemed like a good place for various different reasons for me and my family so I decided yeah let's let's try it and I didn't want to rush so I think I took some time and then my Visa was approved in February 2020 and you know you know the story so everything was closed and done yeah so I decided to still pursue this idea so I moved I moved to Bay Area two years ago and that's why I also wanted to restart the the meetups I hope that you know the covet is gonna end and everything is going to go back to normal which still is not the case but it's getting there we are getting there and uh yeah I'd love to promote the language and and Technologies around Scala and organize meetups and conferences and also as a company we help other companies you know get up to speed with uh Scala expertise so if you need any help for you know your projects then we are here to to help sounds awesome and it's a great story I remember you came to scale by the bay yeah 2018 your team was very effective meeting everybody dropping up business right so like I saw clearly this is the you know this is a go-getter company like you make things happen and yes you get like a nice hoodie from you guys this was awesome so thank you for that it brought us a very nice t-shirt specifically for the smith top so yeah I think that's the first time so you know this is a historic moment right like we're redoing it uh so I totally you know agree with you on the beauty of scholar that was brought me into Scala but I also learned at the same time as scholar all right a lot of other function programming languages so I learned cross kill closure or camel I knew I like a little bit and and basically it's called so I thought you know when I was finished my PhD this is like the last time in my life I'll be able to learn languages for my own fun which was true it's because like you said like as you progress with your career responsibility to grow so so in my mind right scholar is a representative of functional programming family of languages and even earlier I noticed that in a lot of before the crypto boom there were not many high school jobs much more scholar jobs right and so a lot of cross Killers I call them moranic or crypto crypto crypto hostels programmers but they were exchanging type definitions in high school which is much shorter right and so I think that kind of clearly you see cross-pollination between the hospital and Scala and uh so I wonder what's your take right so like we have a bunch of Geeks Who basically love programming languages scholarships happen to be the one which pays a lot of money compared to others which is employed by Twitter confluent data breaks right so uh there is a bunch of scholar inside although a lot of stuff outside this python now like by spark right but I learned doing this Meetup that a bunch of companies use cover for instance launched Darkly I didn't know they use skull I know them yeah I didn't know that too so good good to promote you know through the meter because we can learn that there are more users of of Scala and other languages if we you know if you organize let's say Haskell Meetup then we're gonna learn there are more people out there and there is a lot of talk about that Scala is losing popularity which I wouldn't say that's the case I would say there is less like growth the Grove has slowed there was a big time for Scala a couple years ago and now it's more like mature and and slowing down but also uh interesting approach to like Haskell guys uh hiding behind Scala I totally agree so I would say like Scala is this uh this uh maybe not the Holy Grail but some like a meal ground between object oriented and functional programming and uh it plays a very important role because you have this uh like hardcore uh functional programming languages like uh Haskell and uh closure and all come and and those are very hard to learn and start and then you have object oriented which is uh like the basic bread and butter for for programming nowadays like you know Java um.net and and other popular languages python so um so then there was this huge popularity of functional programming and Scala was helping to onboard people into functional programming and now a lot of this stuff is being taken from Haskell and Scala and being backboarded to other object-oriented programming languages so as a whole I would say the the whole industry is moving forward and learning we're learning from you know from each other and uh and like the classic object-oriented programming are nowadays are having all those functional features as well because it helps to write better software you know less box and it's easier to reason about functional code instead of object oriented so yeah there is a lot of good stuff is going this way and also we have you know AI everyone's talking about AI nowadays and uh we have big influx of people trying to do AI without even thinking like what is it and you know what are you gonna do I just want to do AI because he plays a lot so I would say it hurts like the industry uh a bit because it's driven by uh like not uh not the drive for I don't know for technical or like you know I don't have like this internal drive because I want to be a software engineer it's more like money driven because all right it pays a lot so maybe I'll uh I'll become a software engineer uh so so this path is a little bit uh sketchy I would say but still uh like the software engineering is is growing like crazy we have a slight downturn uh right now because uh the market went really crazy uh with the salaries and you know every company is buying out Engineers from each other so right now with this slow slow down I think it's getting back to like a healthy state which is which is uh really important and asking about the programming languages itself I just came back from Seattle from scalar days so I have a fresh intake on what's going on in specifically in Scala so they they move forward with the language itself now we are you know in this big move from 2.13 to our Scala 3 which will promise you know forward and backward compatibility something that is uh holding scalabag was holding uh in in Scala too and there is a lot of good stuff happening so so Scala is thriving um yeah there is adoption uh is going on maybe a little bit slower than it used to before uh and uh yeah I I don't know exactly what's going on with for example Haskell or closure I'm not so close to those languages but yeah definitely there is a lot of movement forward and uh a lot of like r d going on in in uh in those less popular languages because that's um possible you know in case of java Java has to be very slow because you know if you break something in Java like the whole world breaks down that's right uh so there is all the banks stop working yeah exactly so you know there is this uh like uh very hard balance of uh Innovation versus stability and definitely the functional languages are more on the Innovative side and uh with the AI I think right now python is dominating but I don't think that's the best tool for uh for that job so at this point I don't see any like a specific tool I mean from the programming languages that would be the best suited for developing AI Solutions machine learning but I think that something else than python definitely is going to appear that's going to be very performant and type safe for uh for programming AI so that I would say that's my two cents on what's going on right now you know on that you know I have couple kind of you know uh thoughts about what you said so first of all like for the Department languages you know I totally agree with you like I was really hoping that something will come up that will be secure and fast for AI so there was a switch for tensorflow project which for a while looked like you know it's going to be it so Chris laudner creator for LV I went to Google right and they did this and then basically I was so happy I did actual Swiftor tensorflow Meetup and Jeremy Holbert with the speakers you know scale by the bay in uh Creator fast AI who teaches deep learning for social Engineers he was super enthusiastic okay actually this course on Swift you know and then after like Chris went away the group was picked up like by Apple compiler team and others you know I felt sad but just like last month Rod Chris announced a new language called Mojo from modular.ai which basically would have like a drop in runtime for python just speeds things up magically like three times but and then it's a typed compiled language so and he showed this demo the moment you wanted to eat one like thing as a vector it's up by 10 then it's a matrix now it speaks up by a hundred now as you vectorize it and this is up by a thousand right and sold out three types he claims that will be totally compatible with python so I think Mojo is a huge thing which will be presented by modular at scale by the bay in November so hope you guys track it and come you know scale by the way is back and this is where Lucas was and this is also coming in Oakland actually very close on the lake of yeah like Merit right very very nice venue I love this place yeah so it's really an exit and we are next to the original Oakland uh skill by the way which we did at Kaiser so like we're actually in this beautiful place uh where we used to do it and we'll do it again so that's on on programming languages uh also when you mentioned you know popularity of scholar declining I kind of felt the same looking at all the drama in in on Twitter right and so you would like look at it and you would think that but when I was doing this and then I'll scale by the bay a lot of our volunteers came back and I learned from one of them who was a volunteer it will be a volunteer again Eric and great guy he works at carvana everybody in America knows carvana you know the used car company you buy online that has like sometimes they have a you know a vending machine for cars so they use scour right I didn't know that so I want to learn what they're doing so I think the real fact is if you talk to a lot of people in Industry who haven't heard about launch Darkly we haven't seen them on social media arguing with people obviously they're in business right like they're not there for for that and so what I find is that a lot of work is being done with scholar by people who really don't care evolved you know arguing about politics and so and yeah yeah and also sorry for interrupting but also that was one of the arguments during the scholar days so they they see uh the down the number of the artifacts Scala artifacts downloads and there is no like downturn it's uh there is a steady growth the growth is not so dynamic as it was before so looking at the data yeah you can see that there is still uh adoption is going on so that's really good and uh yeah about the AI and and modular I think that's the way to go something like um you know uh compile into python so we're gonna have something that uh like Scala is being compiled actually to Java byte code so this is something that's being very performant and compiled probably maybe into python or python something else I know that's I think that's the way to go because right now the like the libraries and the ecosystem in Python for AI and machine learning is so huge that it's going to be really hard to to you know just get out of that and and do something new so very cool you know just like a little aside that's like my one of my hobby projects right so there is a new social called Blue Sky and tons of top Geeks are there like it's like jobetta the creator of kubernetes is live coding the protocol like you see you know you know breakfast spot like you know one of the critters will go like being there because it's written and so I did little science projects so I invited them brother and Martin andersky to Blue Sky I also when he talked about object functional obviously the big other languages of sharp so I'm very happy that I brought down sign to Blue Sky I really want to bring rishiki and you know we talked to spga if he will come so we invited him so I have invites for the founders program languages right I work with kind of the developer team and they like web developers they like programming languages so if you anybody who is watching knows and can invite Rich KKK goodov and Rossum and like you know all the other creators of programming languages typescript you know let me know connect me I want to give them an advice because we should have good gigs in that place in that places as opposed to Twitter is kind of spam freeze and white only the early Community is full of developers and it's like it feels like early days people are hacking people are talking about stuff it's very free of drama and like it's really nice so you know if you are tired of drama go to Blue Sky and have some advice hit me up bring some founders of programming languages that's that would be really really cool I have something I want to run by you I think I might have shared with you but I get more and more convinced of this when I go to La Land events here's kind of what I see happen in Ai and why python is not necessarily there like only thing in town basically pie torch was used to train a whole bunch of models people spend hundreds millions of dollars thank you you know open AI thank investors now we have these machines right so I treat them as an oracle it is a black box right it's like Oracle of Delphi people will come people would pray to it and say magic words and will spit back answers right and it knows so so what's happening like if you look at llm programming prompt engineering all of these Frameworks is basically string apis that's all it is it's basically it's an API scale you send strings to it it speeds all strings right and now like you can see the structured uh input it will maybe spit out XML back at you maybe it will be broken and made up XML but generally strings and who knows strings best first of all erlang was built to exchange billions of short strings this is exactly the case for hidden an API and so who knows I think we know I think our community is a thing for breakfast right with no Futures like I I claim that you know this community with akka and Futures and everything knows I think programming at scale better than any other community right like no matter which other languages have this libraries archive was exercised at planetary scale right otherwise Euro you know penalty or something like huge things yeah Disney is using uh Disney for streaming and uh yeah the streaming also the AKA popularized reactive programming and streams and this stuff so there is a lot of innovation that is happening in Scala that is being on adopted by other uh you know other Golan or python or whatnot I mean like the approaches to to architecture how to you know solve those big scale issues a lot of those things happen in Scala and I think that was a big hype before and uh and people think oh no Scala is gone and that's completely not the case kind of is thriving and and being used every day in you know a lot of different companies of course there are always cases where someone says oh no no we are getting out of scallop because XYZ or whatever or there is this huge drama in the community which actually doesn't affect like majority I would say 99 of Scala users are not affected by the dramas that are going on between let's say the the faces of Scala and uh yeah so there is a lot of innovation going on and it's going to be keep it will keep going so definitely it's worth to uh jumping in so one of the reasons I want to organize meetups and conferences and speak about scar is that the you know the way to um grow uh The Scholar communities to get new Engineers joining especially young Engineers so one of our speakers today Alexander his in his 20s we picked him up from the University from the uh we we do Scala camps during the summer we we pick a bunch of students we teach them on on some real projects and you know we decided to take a couple of those guys and and keep them teaching for the next usually like two years maybe when they become Junior engineers and they can start working on Commercial projects so it's a very long and very slow uh process but you know I'm really happy even if you bring uh let's say one or two persons graduates or even students because they start usually they're under a third year uh then I'm really happy that we can grow the community because that's my goal to promote the technology that I love and and I want to you know keep it going that's awesome I really like that approach Red Rockets versus the magic you invest in people you understand learning it's not going anywhere right like what I found I wonder if you saw the same right like people are talking about what's the charge GPT effect on programming right and some people say like this is going to take their jobs it's going to disappear so one nice uh Insight I got from my friend who is at Google and she's a senior engineer specifically like to speak than to say that judgpt will get rid of programmers it's like to say that magic will get rid of magicians yeah like it's not right like it's it's it's it's when you like see that that makes no sense but but I also found a lot of people in our communism like Roland tweeted right like he didn't care about a pilot for a long time and then he tried it he found that he completes his project and transvestor right so what I found that and so the same thing I've seen for many senior people instead of seeing it as a threat and like ringing their heads and like lamenting the demise of software engineering is a job they just like now I can do my thing faster five times faster three times faster so there they know exactly what they're doing right so people who know what they're doing now can get you know to do it much faster people who don't know what they're doing we will be replaced by child gbt right and so this is kind of like so what I think will happen in my mind right A lot of people you just mentioned who shouldn't be doing it because it's not their calling right they will be probably replaced by shot GPT and do something else but people who really do it well and care about it will just see exactly what they need to do with this what what do you think yeah that's uh that's a really good question and everyone is asking that and I was thinking about it also well obviously and uh so when I started my uh software engineering Adventure that was a very long time ago um it was more it wasn't as popular and it wasn't as uh as well paid as it is nowadays so I I was more like Enthusiast and there was a much more Enthusiast in enthusiasts in in in general in Tech and then you know the the big.com happens and like the everything like exploded and right now we had this thing again during the covet like the tech exploded us again um in popularity and and in salaries as well in the money that was being poured in which actually made things even worse so yeah totally I mean charge GPT is a nice tool and it's not useful for everything uh it's you know for example if you ask it something you don't you don't get any sources so for my from my perspective you know I can't ask it about something because I don't know where it's coming from like where does this chair take its uh information right where can you at least give me some sources that they could verify if they are legit or not so that's a that's a big thing that people might get by us if you know they don't verify but for programming language programming jobs definitely it's a it's a useful tool it helps there are one of the companies Source graph they developed AI tool for helping coding yeah you can easily just type that I'm looking for something and you know give your uh like text description and it's going to give you what you're looking for and the structure search so they have this new tool and uh and definitive charge GPT or any other uh text AI can uh help with fixing bugs in our software then awesome I'm gonna take it you know all day you know it's a lot of times software Engineers spend their uh work doing you know just looking for bugs it's a waste of time and money we don't want to do backs we we want to just get rid of them and if you know if AI can do it it's uh it's cool and the risk of losing jobs well you know it's always um it's always the case like we've um robots in you know factories and and other places where we can automate and so yeah definitely we're gonna have other jobs so actually charge gbt is creating new jobs like a chat prompt uh engineer that's being called right now yeah so we have a new category of jobs you know there's someone gonna be have to program the charge GPT to you know give the inputs and outputs so um yeah I I'm not afraid that we're gonna lose our jobs to charge GPT I rather see it as a tool that's going to help you know be more creative and you know do our job faster and maybe get rid of mundane tasks like you know back finding and stuff like that because uh why should I spend my time doing less creative work of you know back hunting while I could spend it on thinking about like a new creative solution for our product so yeah I'm looking forward to to the new era it was awesome you know I want to do a shout out because I was looking right like I want to program against GPT API I don't want to do it in Python can I do it in sculpture at Google and I find this you know open AI scholar client of course it's written by an Eastern European company in Slovakia so as a shout out to Peter bonda if you're out there I found this right he implemented the whole client everything in a dramatic Scala including scholar three right he found bugs in the python many many bucks right when he was doing this and then he proceeded to do baby IGI and Scholar he found a whole bunch of boxes right so now you can do that and again like there's absolutely no reason to do this in Python because you're just sending strings you're receiving strings is stored in history of strings you know partial strings like you're gonna don't scholar beautifully you have structures you have objects right so I'm super excited I think it's a Renaissance of programming now we can program the strings like programmers no strings we know something even don't know anything but with no strings right so so I'm really excited plus of course and Eastern European guys and girls and everybody have really good education so this is coming out we have a lot of Polish programmers doing scholar Slovakia and everybody right so like I I think the result of learning like you said the red leaders are strong schools of learning people have this knowledge they pass it along so and we're back in the physical community and I think a lot of drama happened during pandemic because people were just United right like they imagine things like now where we're back I think it will bring back the community I think the drama will decrease I think it's the kind of friendship will increase so thank you Lucas for bringing us and we're looking forward to the talk yeah thank you Alexi for having me