Musings of AI & Space Enthusiast: Part I
Hi! This is a relatively rare post for me. It’s more than a year since I’ve returned to active blogging, but it’s probably the first time I’ve ever started blogging about things that are more close to me and that I usually posted on Facebook. Why Facebook?
Well, my first blog was born on the now long dead Microsoft blog platform called TheSpoke.net, in Spring 2005, more than 15 years ago. Turns out I’ve got multiple blogs started and abandoned over the years. Some were killed because the platform died (TheSpoke.net, GotDotNet.ru) others because I’ve left the ship (two blogs on blogs.msdn.com, two startup WordPress blogs) or simply left to vanish (LiveJournal, Blogspot). Over the years Facebook became my primary place for blogging. Although its “Notes” feature was relatively bad it had an access to an audience of people I’ve found and added over the years making my notes directly accessible by those whose opinion I cared about. More than a year ago however Facebook killed their Notes feature though promising that the contents won’t vanish. In a way it’s true though finding my own notes published on Facebook is now a painful process. And in a way I’m hopeless. And so last year I’ve started to selectively bringing here at least some of my blog posts from all these places, with the hope this to be perhaps my new long-term blog. And now more than a year later I want to talk about things that I would earlier talk in my Facebook notes but now will have to resort to talk about them here. Like a dear friend of mine, Chris Karr, I’m leaving Facebook in a way and using a different medium to share my thoughts. Ironically that medium is Medium, and so… Let’s begin!
This post is the direct consequence of me starting reading a book “Education for Educated” by Anatoly Levenchuk. A dear friend of mine, Kirill Pimenov, recommended me to read it to help me more systematically look at my life and things I do. A bit earlier this Summer I’ve tried to visit a psychologist; while preparations and the very first talk helped me to clarify whatever the hell I want from psychologist, this didn’t turn into a successful story. Delays on her side, then I’ve got into a crazy situation where I’ve prepared for a call at 1pm while actually it was at 12pm, then delay on her side, then I had problem with time on mine… Long story short we didn’t make that second call and perhaps I don’t actually need it. Instead this book — ”Education for Educated” — seems to be what my soul had been looking for. So what am I looking for?
The answer to this question goes back to my 30th birthday. Back then my grandpa was still alive, lying in hospital, and he reminded me that I’m old enough to begin my own family. While I’ve always wanted one, I also wanted to become a product leader in the area of intelligent systems. The name changed over the years — adaptive, context-aware, AI— but my passion for them remained the same. And although my Master’s thesis (formally it’s a “Specialist” degree in Computer Science but Russian “Specialist” degree in STEM usually equals to Master’s so) was focused on the context-aware computing systems, back in 2008 I was quite far from building these systems. Now, I’m 35 y.o. guy, and my job is to make sure our Conversational AI lab has a great portfolio of products. I’ve changed my life starting with a crazy dream back in 2005. Success, isn’t it?
Except that this is not exactly what I wanted in 2005. 16 years ago my understanding of the future was like a lot of fog (like fog of war in the strategy games like RTS or Civilization-like games) ahead of me. I felt the direction but I had no firm idea whom I want to be. Well I had some ideas. Back in 2005 I was passionate about Microsoft and its way of building products. Its at the time coolest AI-infused project — WinFS — was back on track in August of 2005 (almost a gift for my 19th birthday), and I knew that there were some magical people — “Program Managers” — who were as I thought driving things. And so of course I wanted to become one.
Except that I was in Russia and they all were in Redmond, WA, USA. And the only two countries I’ve visited by 2005 were Turkey (in 1998) and Ukraine (in 2004). I literally had no idea how to become an actual Program Manager at Microsoft. Remember it was 2005 and the only way I could learn about Program Management was a few blog posts on blogs.msdn.com, a piece by legendary Joel Spolsky, and job descriptions at careers.microsoft.com. And I had no firm idea what I want to do. I just wanted to be a part of something, and manage the direction of that something, and I also wanted that something to be related to intelligent systems. The mere idea that computer is so powerful and yet we yield so little from it was making me mad; I thought that we should be already at the verge of intelligent robots and other things. If you watched “Her” then this is what I was expecting from computers 16 years ago. In 2004 I’ve got my Windows XP Pro with Speech Recognition, and spent a day or so training it so that it could recognize my Russian accent and still understand me. It was a miracle (as it was to buy a separate web cam, separate microphone, and speakers, and getting everything up and running). But I wanted more. I just didn’t know how to proceed. The only thing that mattered was that the experimental school (I’d call it “business school” but it was more like a pre-MBA of sorts focused at public sector) ignited in me a huge belief in myself. I knew I could achieve anything, if I’d be smart enough, and work hard enough. That’s all I had 16 years ago. Oh, and participation in a few Microsoft Beta programs.
One guy, Alexander Popov, who’s now my best friend, was a Microsoft Student Partner nominated by someone from Microsoft Russia I believe for reaching finals in Microsoft Imagine Cup 2004 and later winning Microsoft Imagine Cup 2005. Al is a fascinating person, super smart, super intelligent, kind, a bit sarcastic, and also very detail-oriented. Like other MSPs, Al had his own blog on TheSpoke.net. He noticed me among other bloggers, and, after a few exchanges, we became friends of sorts. He also noticed my passion for WinFS, and he recommended me as a candidate for the expanding Microsoft Student Partner Program. Thankfully, I’ve got this chance, and became a 25th MSP in Russia in the Fall of 2005. I didn’t plan this; it happened to me. Of course, the fact I was so passionate about WinFS was the key, but it was also Al’s attention to me and my blog posts that sparked our friendship and led him to nominating me to the program.
Now, this is not exactly the story of my entire career; to those who know me, they are aware of it; for those of you who do not yet know me, well, you’ll hear this story at some point, anyways. The purpose of this blog post is to show something interesting; a new perspective that may also trigger something in you like this “Education for Educated” triggered something in me.
The thing is, although it took me 15 years to get a job at a Conversational AI lab, it’s not like my entire life was organized around this idea. Hell, as far as I understand, there was no such term, “Conversational AI”. Back in 2005, a few universities across US participated in the DARPA’s CALO project, and Siri was still in its infacy. It’s very easy to look at my entire career now and say that it was strategically leading me to this very position, but it would be a Survivor’s Mistake. I was very lucky to get into this position, I admit that. And while lots of events in my life, and countless decisions led me to it, I wasn’t planning for this. Truth to be said I wasn’t planning any of this.
The only thing of all I’ve planned in mid 2000s that I’ve managed to build was a country house in Moscow suburbs. This is the only big dream of mine I’ve made happen, more or less in the way I wanted it, including landscaping! And this only happened because of tenacity, hard work, and the fact that I’ve spent more than a year planning this enterprise, designing different variations of our house and destroying them.
Somehow this project made me believe I’m good at strategic planning. However this is not exactly true. I’ve made lots of mistakes, and some of them led me to the need of making a cosmetic renovation on the first floor (which is still due) as well as re-building a new brand new cellar on our land long after destroying the existing one. It’s just some of the decisions were good enough to withstand the decade and more. The jury is still out on the whole thing but I’m more or less optimistic.
Judging by many major renovation projects led by our neighbors I can say that our project is indeed a relative success. But it’s not a sign of me being good at strategic thinking. It’s a sign of me being not too dumb at planning a house after spending one year of planning; I’ve got a good book, “Building Your Dream House”, by William P. Spence, bought a few CAD apps that helped me to make several 3D models of the future house, got a lot of feedback from the future inhabitants, incorporated it into the plan, and that’s about it. Nothing fancy. Oh, and I didn’t try to save money on everything; we’ve got good vinyl siding from Canada, for example. Nonetheless, some of the equipment didn’t pass the check of the time; we’ve had to replace our gas boiler a couple of years ago, remodel WC on the first floor so that our gas boiler would get its separate boiler room etc. Lots of mistakes done over the years. I still think it’s a relatively successful project though. My granda and grandma lived in it till 2016 and 2013, correspondingly. My mom and our cat live in it, too. Several of my friends visited me here including even some of my friends from US and far parts of Russia.
Still, this “American Dream” was success partially because of my “strategic” planning, partially because I’ve been following the book and instructions, partially of course because there were spare resources, and also partially because of luck.
This is a rather important observation, at least for me. Why? The thing is, this is a really big project (ask anyone who had ever build their own house, or at least renovated their flat — it’s super easy to make super costly mistakes), and although I was 23 y.o. guy when I’ve started the process of demolishing and re-building almost from scratch, I’ve made it. Why? Well, it was hardly a something new. There were tools and methods, there were lots of success stories, and we had some spare resources, so it was quite a possibility. Why not, in the end? It’s not a rocket science, just a house, after all. Anyone can build it. Yeah, you need to think a lot, you need resources, tools, methods, materials etc., this process might become really lengthy but in general it’s relatively easy. While it was hard for me to imagine I’ll get this done it was clearly that eventually it’ll be done unless something big will change our plans. There was no big uncertainty in front of me. There was no “fog” of war. The only uncertainty I had was the bank loan I’ve got in 2009, for 0.5M Rubles that I needed to accelerate the construction. I’ve got it for 5 years, and I wasn’t 100% sure I’d have a full-time paying job for 5 consequent years (and this was right on spot; I’ve paid this loan in 2 years, and started my own venture in 2012). Even though back in 2009 I was working for Microsoft Russia, had growing team, and in general my career looked rather bright.
So the thing is, and it goes back to “Education for Educated”, you can succeed in big dreams when it’s quite clear how you can make them happen (build a house, write a big story etc.). Sure, you need planning, tools, methods, resources etc., and you need to monitor things.
But if you have a big dream like I had in 2005, well, and there’s a huge fog of war ahead of you (meaning you have no idea what exactly you want and/or how to get there) like I had in 2005, well, then your chances of getting whatever you dreamed of are very slim. While I can easily say that my very vague perspective of what I wanted in 2005 strongly correlates with my current position at DeepPavlov.ai, this is not what I ever wanted for myself. It partially fits into one of the future visions circa 2008–2009 when my advisor, Al and I wanted to build a Context Awareness Envisioning Labs at Microsoft Russia.
But resemblance is small. I actually wanted to be a PM in Redmond, WA, working in the (now defunct) PC|3 Division (which introduced Windows Sensor & Location Platform in Windows 7), overseeing this work from Corp while the Labs would be located in Moscow, Russia, with my friends overseeing them there. While my (unfinished) PhD thesis was “Modeling HCI of Personal Digital Assistants based on Context-Aware Computing Systems”, I didn’t think of focusing the proposed lab on building AI Assistants yet. This whole idea was a continuation of my Master’s Thesis which was focused on the research for “Opportunity to Build a Context-Aware System in the Everchanging Information Environment”, grounded in the relevant research and development projects at Microsoft. We wanted to combine forces of MSR, Windows 7’s PC|3 Division, and some other groups to make Microsoft Russia’s MSR Lab solely focused on Context Aware Computing Envisioning. Why Envisioning? Well, the idea was that Microsoft Business Division (MBD, it was largely formed from Office product groups) already had Office Labs Envisioning Team that was envisioning the future of productivity for Microsoft, and we could use the same methods but focus on adaptive systems, and do that in Russia. Why Russia? Well, my friends and I were from Russia and at some point we had some sort of official backing, so that was thing number one. Number two? Well, Microsoft Russia was actively growing at the time and it felt almost natural to do something crazy cool in there.
Anyways, whatever I wished for in 2009, it’s not what I’m doing today at DeepPavlov.ai. And while I’m super happy for my current role I’ve never planned to get it. At different times in my life I wanted different things, and tried a lot of them. Building and leading Context Awareness Initiative across Microsoft? Check. Doing cool stuff with Windows product group? Check. Showcasing own work to a couple of Microsoft CVPs? Check. Starting own venture? Check. I could go on and on, but the thing is, I’ve experimented a lot since 2005. Made lots of choices, mostly wrong ones. Yet their sum led me to where I am at the moment.
The question is what’s the lesson?
The lesson is, you can’t efficiently predict the future. You may foresee some trends, and you might be incredibly lucky to make right connections and moves, but if you have a fog of war (or fog of innovation), you can’t foresee the future and your big dream will most possibly fail to materialize. And the biggest question is, will you be able to change yourself a lot enough times to adapt and win.
Having big dreams and reaching for them is super romantic, I get it. I was a guy like that. Yet my very own story tells me that if you have a big fog of war/innovation ahead of you, your best friend is evolution which doesn’t have long-term goals except survival but is a great instrument to adapt and win.
While I was super unhappy 5 years ago that many of dreams didn’t come true, 5 years later I’m grateful for what I have. My old dreams helped me a lot to change, to adapt enough to be where I am today. And I have no clue where I’ll be in the next 5 years. I’ll keep planning things, but I’ll do my best to adapt more often, and feel less bad about the fact that my big dreams don’t become true. The thing is, life’s too unpredictable for big dreams born 5 or 10 years ago stay relevant for so long.
Of course this is just one of the lessons I’ve learned from the “Education for Educated”, but for me this is the most important lesson. I am no longer bound to my 15 year-long dreams. And I will not judge myself by these dreams today in my 35.
Oh, and if you want to buy the book, feel free to do it here or here (Amazon).
So long, and thanks for all the fish!
