Summiting 2022: A Mountain Climb through the Ugly and the Bad

Steven Kolawole
17 min readJan 19, 2023

--

I initially talked myself out of writing a reflective article for several good reasons. For one, I am (finally) wrapping up my undergraduate degree in less than three weeks, and I am struggling to work on my thesis and prepare for my finals. Secondly, I thought that whatever I wrote would be incomplete since many of the updates are still a work in progress, and I won't know for sure until late spring. Moreover, I had a rough 2022. Most readers look forward to wins and achievements, and I wasn't sure I wanted to rain on the parade by writing a solemn article. But I also know people who have expressed interest in reading about my 2022, and I am still a big fan of documentation, so here we are!

I try to partition it into several sections where I talk about the failures, wins, close people wins, and critical takeaways sprinkled throughout the different areas. In this article, I will laud other people who have impacted me in 2022 without much thought to structure. Try not to feel bored while I do that.

The Ugly: Facing Failures Head On

I go for ambitious goals, usually believing that I can get them done — that's how I operate. More often than not, I'd fail and just dust myself and pick up the next thing on my to-do, but this year was different — mainly because I was becoming quite confident in my stocks and my ability to pull off challenging goals. That belief was my identity, and everything else about me was centered around that identity. So as the failings piled upon each other, they affected me more and more. My mental state was in shambles, and by the third big failure, I had lost most of my swashbuckling spirit and avoided trying out any other endeavor lest I cement that "I am a failure" thought in my psyche.

PhD goals going down as "first attempts."

In my year 2021's mountain article, I talked about how I want to go for PhD, the overwhelming support I had toward this endeavor, and how half of my 2022 "goals" are centered around being in a PhD program. I honestly thought I was going to get in, considering the positive receptions I had from all my reviewers — and by mid-January, I had had 4 PhD interviews with four professors (2 discussions were by cold-emailing the professors before I submitted my applications while the other two were after my applications were already in the schools' systems). Two professors connected me with their students; chatting with them was fun and dreamy. It was almost unbelievable at first when I wasn't getting any offers. I was monitoring other students' updates via GradCafe (a site you know you probably shouldn't visit as a grad school applicant, but you can't stop yourself from checking) as they'd announce their acceptances/rejections/interviews over there. I wasn't getting any acceptances, and I was getting antsy. When students on GradCafe confirmed that they had received offers and visitation invites from my supposedly safest school (a professor I didn't mention in my application but who took an interest in me; he interviewed me and connected me to one of his students) and I hadn't received any email from them, that was what broke me. It was a relatively rough period for me. I thank Vivian for being my coziest shoulder to lean on during this period.

By mid-February, I realized quickly that I wasn't going to get any invite for some reason — I couldn't pinpoint why. I was convinced I had what it takes to succeed, and I threw myself into the application process with all my might. But as the rejections piled up, I began questioning my abilities and sense of self. I felt like a failure, and my confidence and motivation were shaken. After I had crawled back up to get over it, I started working toward my "backup" plan.

AI Research Residency at Google Accra

I am thankful that Thao Nguyen, one of my favorite friends cum mentors during the PhD application phase, recommended that I apply to residency programs as backup options. As of then, I didn't see any reason to. I thought I was significant enough to enter most schools I applied for. After she mentioned it more than once, I eventually applied for OpenAI, Meta, and Google Africa's residency programs. I made the cut for the Google Accra program and got feedback from the recruiter about three weeks after I moved on from the PhD heartbreak.

I was going to have two interviews — a research interview and a technical interview. Naturally, I was leet-coding and eating data structures and algorithms for breakfast and dinner. I was scheduled to have the interviews by mid-April, and I started having calls with folks who have had similar experiences. After completing her undergrad at Stanford, Thao was an AI resident at Google Brain before resuming PhD at UW, so she was naturally my first point of call. Annabelle, my reading partner turned friend at Cambridge, also recommended a friend of a friend who went through the same experience. Leila Marie Hampton, my MIT mentor cum friend, recommended Julius Adebayo — another MIT PhD researcher who had passed through the Google residency program. So I had the calls and practiced with a study plan I put together from those helpful calls.

Unfortunately, my interview date clashed with my (hectic) preparations to travel to Germany in April. I chatted with my recruiter to ask if I could reschedule, and she affirmed while reminding me that hiring is on a rolling basis. I rescheduled it for late April, which coincided with the Easter holidays. I finally had my interviews in early May, and they mostly went well. In the technical interview, I solved the question well-using recursion, which I later upgraded to dynamic programming. I got the time complexity of my solution right, but I wasn't sure about my space complexity answer. We started with statistical questions for the research interview, which I struggled with, but I answered the case study designs and explained the requested terminologies well.

I was a bit surprised when I jumped on a call with my recruiter about three weeks later, and I was told that Google couldn't move forward with the application. According to her, I didn't perform well on the coding interview but did great on the research interview. It was a bit surprising because I thought it would have been vice versa. I realized that the "hiring on a rolling basis" issue she raised earlier might have been a contributing factor, and I probably didn't perform exceedingly well to get in when the open slots got fewer. Around this time, the academic union strike action was in its 3rd month, and it was evident that I couldn't graduate in June/July again — which means that I wouldn't have been able to resume PhD or the residency program if I had gotten in either — but this did very little to soften the blow.

Google again!

Around the time I finished my sign-to-speech work and was moving on to resource-efficient NLP, Rosanne submitted an internal application for me to work as a student researcher at Google Brain. At first, I could have been more enthusiastic about the prospects of this work. First, it is Google Brain. Secondly, it would be at one of the San Francisco offices if it works. The part of me that still feels impostor syndrome thought that a US student might have a higher chance of getting in than I do. Finally, everybody else that I had heard of working as a student researcher at Google Brain is either an MSc and PhD student — except for this super brilliant researcher lady that was one of the founding members of Cohere. I felt I had a low shot as an undergrad, especially one from Africa.

To cut a long story short, the application was moved forward in few weeks after I was rejected from the research residency role. I submitted an official application and was scheduled for an interview a few weeks later. I'm thankful to Rosanne for connecting me with several friends, Johan Obando and Hattie Zhou, who were kind enough to walk me through their interview experiences. The interview was tough — heavily theoretical — but I was pretty well prepared, so it went well. A week or two later, I got confirmation that I had passed the interviews and was in the final phase of the hiring process, the project/team search phase. Rosanne and I had a meeting about finding the right team to join. Interestingly, I didn't want to join the Translate team since I was getting more theoretical with my research interests. We were going to find a new group to join.

A few weeks later, I was in Tunisia for Deep Learning Indaba and was opportune to make the acquaintance of Sebastian Ruder. We talked about many topics ranging from our general research interests to my research work at ML Collective and how I was about to join Google. He suggested that I let my primary mentor, Rosanne, know he'd love to work with me at Google on efficient NLP. I got back to Rosanne via email immediately after returning to Nigeria. But Google had announced a hiring freeze the week before due to the global recession, so all pending offers had to be withdrawn!

The feelings of Inadequacy

I haven't felt as unmotivated in my career trajectory as I did this year. I wasn't even feeling — I was numbed. I felt bereft of my ambitions, my usual confidence, and my appetite for work. I couldn't work during this period — I couldn't even "force" myself to. If I had felt that I had another option aside from tech (or machine learning, to be precise), I probably would have considered it. I couldn't put together a string of decent work in several months. My primary research partner during this period, Ender, actually considered quitting on me at some point, and we had to have a very frank conversation about how we were feeling, our other commitments, and how they were affecting our collaboration. I was working engineering on the ASAlytics project then, and I kept failing to deliver countless times, and my team was so tired of me. We had maybe a dozen frank conversations with me, but I couldn't. If it were a "standard" tech company I was an employee at, I probably would have resigned or gotten fired.

I couldn't place a finger on what had happened to me. I considered visiting a therapist at many points but talked myself out. Through frequent (free-form and gratitude) journalling, I postulated that my most substantial theory is that I am usually gung-ho and very self-confident about my abilities and the wins I want. Hence, when the strings of failures that I thought I was equipped to land concerning arguably my most important career goals (at that point) started accumulating, it affected my subconscious in a way I wasn't expecting. My entire belief system was shaken (or collapsed) as a result.

The slightly better stories

The academic union strike action: Healing while there is no pressure

One big fortunate thing is that all these happened during the 8-month ASUU strike, which means that I had enough time to sulk, pick and drop some vices, and go maudlin on myself — without any significant effect on my career trajectory. When I returned from my Europe trip, I thought I was getting better and picking up the slack slowly. This won't be the last time I'd experience a mid-life crisis. But I am glad this big one happened during this period. One of the most notable quotes from Rosanne that stuck with me is how we must go through feelings of Inadequacy to become adequate. Going through this experience during a slightly inconsequential period was a boon to me.

ASAlytics: Building with my friends

I think how I seem to struggle to work in a proper 9–5 setting is well documented— remember how I quit my last AI Engineer job in December 2021. It might have been one of my biggest innate reasons to venture into research — the chances of working solely on my interests and the flexibility that comes with the work. ASAlytics was a boon in this direction. From writing the proposal that got us the $115k grant from Algorand to building, the flexibility in working on something that is my conception, at my convenience, was too good to be true. The icing on the cake is that I worked with my closest friends, Ernest, Busayor, Temi, and my sister, Precious.

Aside from building together, we got closely knitted during this period. A regular occurrence was having an hour of discussion of family matters after our weekly standups, where we'd discuss personal issues, gossip, and drag any of us (usually me — I am a soft scapegoat) that has shown some weakness. We also hang out at every opportunity we get. When Precious got her Shopify offer to relocate to Canada, we added her celebration party to our 1-week hangout in Ibadan. When Temi got his job offer to relocate to Europe, we threw another hangout in Abeokuta. It fell around when I got into the project matching stage of my hiring process with Google, so we extended Temi's hangout by another day to celebrate my "win" prematurely.

I am quite grateful to have this family in 2022. It is my most significant positive highlight of 2022, and this bond teaches me a lot about how much other stuff aside from work matters.

Being a big boy financially. To some extent.

I am not so materialistic; hence it is a bit difficult to gauge how much financially comfortable I am— I am still using the same old HP laptop that I was given while being part of DevCareer's cohort in 2020. Don’t judge me; I just haven't seen any solid reason for an upgrade till now. I am also not so hyper about making huge money — I like to say that I am more knowledge- & prestige-driven than income-driven. Hence, I want to maintain that a reasonable estimate of how financially comfortable I am would be how well my family is, considering that I am not shy about splurging on my family. For example, I didn't realize that I wasn't a financially struggling person again until 2021 when Grace broke the screen of her Macbook, and we got a new laptop for her in about 2 or 3 days. Likewise, I didn't realize how financially OK I’d gotten in 2022 until Precious was restricted from boarding her non-refundable Lufthansa flight (worth around 2M naira) to Canada because she didn't possess a US transit visa. We got a new flight ticket in less than 24 hours. Aside from my travels and buying my dad a car and a landed property in a highbrow part of our state's capital, there is no accurate metric for how financially comfortable I'd gotten.

I'd always believed in investing in my career capital (borrowed from this book) crazily, and my financial needs would be taken care of automatically. This can be amusing for someone who came from this in 2020, but I am grateful for going through the contrasting experiences.

Traveling, socializing, and healing

This is the list of countries I was in this year. Sorry, it is sideways; I didn’t want to spoil the symmetry of my article :)

Traveling solo and in a group were significant parts of my 2022. On solo traveling, getting lost while traveling from Bologna to my Florence hotel, hiking through Florence in the nights, taking a road trip across Switzerland, getting lost at that very big Berlin, and missing my return flight at Berlin Bradenburg airport are highlights. On group traveling, Busayor and I becoming the lives of the party at Deep Learning Indaba's opening party night, and a midnight hike with some of our Tunisian friends were the high points. Across all, I enjoyed all the performing arts that I encountered, and some actually struck nerves within me.

After Busayor and I had hijacked the DJ’s playlist and changed it to AfroBeats, we opened the dancefloor, then came up the stage sometime later to play backups to our random Ed Sheeran. :D

Aside from my well-documented travels outside Nigeria, my trips to Nasarawa, where I got the AI Champion award from the Nigeria Computer Society, and to Abuja 2x were undoubtedly fun. A particular experience that stuck with me was when I decided to take a night road trip from Abuja back to Abeokuta (I didn't tell anyone except Abraham :) ). We were somewhere at a filling station in between Abuja and Kogi, and there was fuel scarcity, hence a very long queue and crowd. I spent an hour there sampling all the local foods I could find!

Aside from the unique experiences, traveling offered me solitude away from my problems. I may say that getting back into the groove was faster due to the time spent away traveling and reflecting.

Still speaking, rain or shine

One of my bids in late 2019 was to cure my 100% introversion and social awkwardness by taking up community work via teaching, mentoring, and public speaking. Since I wrapped up my tenure as the Community Lead for the Google Developer Students Club in my school earlier in the year, I had become a 'godfather' and less active in community mentorship. For speaking, it's different. Since I gave my first talk at PyCon Africa in 2020, I haven't looked back. Even through my rough phases, a regular part of my life is speaking about my work, my interests, and the tools I use. Giving talks outside Nigeria was undoubtedly a highlight, but I also gave my first high-profile local talk in Nigeria at DataFest Africa, and that was lovely. I also granted several interviews on my sign-to-speech work, including with the AIHub organization and the U of Toronto News.

A snip of most of my high-profile talks on my CV. I haven’t added DataFest Africa ‘22, Build Stuff Lithuania ‘22, and my several talks at GDG Abeokuta, including the last DevFest.

A slightly amusing fact is that I could have spoken more. I had to turn down speaking at All Things Open, Beer City Code, and the SciPy Conference in the US around 2022’s fall because of the complicated situation of getting a US B1 visa in time. I also canceled going to Albania to speak at OSCAL 2022 in the summer since I wanted to be back in Nigeria early. I had lots of visa issues. I like to tell my friends that bearing a Nigerian passport or (more generally, being from the Global South) means that you are probably already 1–0 down if you're competing globally or striving to go global.

Maintaining my workout

Another constant part of me was sticking to my workout routine three times a week. Having that regular part of me in a somewhat chaotic period gave me a sense of calm and relaxed enjoyment. I mainly focused on strength and bodybuilding, and I bulked up by about 25 pounds in the year while at it. I don’t look so slim anymore! :)

The Moments to Celebrate

Some positive happenings in the year got to me very strongly, and I discuss them below.

My support network >>>>>>>>

Rosanne doing Rosanne’s doings when I told her about my PhD rejections ❤

My sisters, my ASAlytics family, Moyo (or Mardiyyah, if that's what you know her as), and Rosanne are perhaps the most supportive people on this planet. Throughout all the rejections and being down, they constantly supported me and provided much-needed solace. Dr. Babatunde Olorisade, Vivian in the earlier part of the year, Nayan, Gloria, Victor, Abraham, Annabelle, her partner, and Oreva in the later part of the year were also remarkable for me.

Me calming down after Moyo talked sense into me through one of my senseless rants.

One day, an organization invited me to speak on my research. I felt everything was going the opposite way for me, and I didn't deserve to be a guest speaker alongside arguably more accomplished ML practitioners on the speakers' list. I was going to reject the invite with lots of bitterness. I'm grateful to Moyo for speaking sense and calm into me. And for the sudden surge to stop my pity party and wallowing in my strings of failure, I am grateful to Oreva for doing that part unknowingly.

Understanding myself and my wants

The most fortunate part of going through a lot in 2022 was the fact that it made me self-reflect over and over again. A large part of my belief system was rendered useless, and I had to build from scratch (in part) again. Because I was now more aware of my failings, I became more humble, maybe a little wistful. Naturally, I became more philosophical and queried my former accepted life meanings. The fact that I am irreligious means that I couldn't be satisfied with putting everything down to a divine being's omniscient tendencies, so I had to find my truth most time.

A very career-focused example is this; I found myself evaluating why I wanted to pursue a PhD, and I asked myself hard questions. For example, am I doing research because I want to go for a PhD, or am I applying for PhD so I can learn and do research? If I remove research from the PhD experience, would I still go for the PhD? And if I could do research without getting a PhD, would I still do it? Asking myself these questions made me realize how much I had strayed from my goals and how misplaced my motivations were. By failing to get into a program, I had enough time to do research calmly and determine if I do enjoy doing research.

Reflecting on the year, I am reminded of the importance of resilience and perseverance. Life is not always easy, and we will all face our own set of challenges and failures. But through them, we find meaning in life. And through these struggles, we grow and learn and ultimately become stronger.

Prioritizing genuine relationships over career connections

Because I got disillusioned with working, I naturally sought solace in my social relationships — I learned that this term is known as "enantiodromia." I realized that I was making more efforts to have genuine social relationships with folks instead of being hyper about my career most of the time. It might not seem like a lot, but my kind of person probably wouldn't have a relationship where I couldn’t see how it empirically adds value to me. But discarding career thoughts while connecting with people means I could find more meaningful and less stressful connections when I click.

My people winning left, right, and center

I am grateful for Precious's successful transition from her medical degree in Nigeria to starting her Computer Science degree at Carleton University in Canada while also interning with Shopify at the same time. I can't blow her horns as much as she does herself. Read her article here.

I am grateful for Temi's successful relocation to Europe via his new software engineer role. It means a lot to me since I remember how we'd walk home together (we used to live beside each other) from the 720degree hub in 2020. I was doing my DevCareer program then while he was in his first front-end development role. We'd talk about our different fields and talk about future goals. I remember telling him I wanted to work with Google as an ML Engineer, and he'd also exchange his plans for the next few years. Guy finally did it!

I am grateful for Mary's and Opeyemi's PhD acceptances. While they are close buddies themselves, it's funny that I became friends with them independently — Mary at the DSN AI Bootcamp in 2019 and Opeyemi when she came for her youth service program in Abeokuta in 2020. Mary resuming at UCSB, and Ope resuming at Edinburgh significantly validated my dreams. Since they did it, I could as well.

I am grateful to all my close friends, relatives, and mentees for their big wins, for not giving up, and for learning from their failings. I am also grateful for the ones that made the word "friendship" carry another meaning. I can't name everybody here, even though I am tempted to try to list names. I like to believe those important to me know that I rate them very highly.

What's Next

Quoting Rosanne’s anecdote on living in the moment below;

"[Days] lost their names. Only the words' yesterday' and 'today' had any meaning for me." — Albert Camus in The Stranger.

How 2022 went was different from my goals and expectations, except for the traveling part. In 2023, I am focusing more on my process and system than anything else. I don't have any real long-term goals. I want to live in the moment and focus on what I can control. My research, my exercise, my readings, my people, and maintaining my work-life balance are my big deals this year. Asides from this, I'll be 25 in 2023, and it’s been crossing my mind these days to calm down and maybe maintain a steady love life. I finish 2022 with a healthy amount of strength and calm. I wish to continue that trend in 2023.

--

--

Steven Kolawole
Steven Kolawole

Written by Steven Kolawole

Machine Learning (Engineering & Research). CS Graduate. ML PhD Student.

Responses (4)