thoughtwisps One commit at a time

Hello and welcome to thoughtwisps! This is a personal collection of notes and thoughts on software engineering, machine learning and the technology industry and community. For my professional website, please see race-conditions. Thank you for visiting!

the benefit of the technical doubt

Who gets the benefit of the technical doubt in the tech industry and who has to go out of their way to prove they belong?

Editor’s note: Finished an published on the 30th of May, 2019.

Even Sisyphus - the mythical man condemned to push a rock up a slope - gets to the top eventually. But just as he’s reached the top of the mountain, the rock rolls back down and he has to start all over again. Or something of that sort. The precise details of this story have left me.

The reason my neural circuits sometimes conjure this image if that it strongly correlates with the feeling one sometimes gets being an underindexed (to borrow a term from Cate Huston) person in the tech industry.

Not a long time ago, I sat in the office of a professor at my university and talked about the tech industry. You have to be good, really good to make it, she said. I knew she was good and yet she’d chosen not to enter the industry. I also knew I wasn’t good. And I knew she knew I wasn’t. In another context, this might have sounded like a peptalk and perhaps it was intended as such, but now with hindsight and with my history of success or lack of thereof in many of the subjects I attempted, she was perhaps trying to give me a fair warning to not raise my expectations of success in this field too high.

The gift and curse of youth is the boundless naivete, which makes us so completely utterly convinced that ‘we know better’. The rules apply to everyone else, but not to me. I will definitely be the exception to the rule. It only takes a few good, hard disappointments to realise that while exceptions do happen, most of us probably won’t be in circumstances exceptional enough to be exceptional.

So I took my professor’s advice as a rallying cry. As long as I tried hard enough to excel in tech, I too could succeed. The zeal with which I immersed myself in everything from late night coding to meetups and giving conference talks quickly dissipated when I realised that just as I thought I’d accumulated enough of tech street cred to not have my basic technical work questioned, I’d have to start all over again. Just like that guy rolling the rock up a hill.

A few memorable moments have remained. Sitting in a meeting with the QA engineering team and talking about mocking, I get interrupted by another member of the team so that we can all make sure “I am absolutely sure I know the difference between mocks and stubs”. While geeking out about tech with the people in my graduate engineering program, one of them stops talking mid-sentence, turns to look my way and, with a concerned look on his face, says “You do know what XOR means, right?” What he means: I mean, you don’t look like you would know what XOR means.

It’s ok when it’s small things like this. They’re paper cuts, but in the big picture one can probably live with it. When it becomes real, is when my peers are promoted to managers and tech leads and retain the inkling of the same biases that made them ask those questions in the first place. It means that they will be customers of the tech company I work at and they won’t want me on site, leading the project, because I might not be technical enough. They will be the manager who sits in the meetings of the promotions committee, looking over my profile, and they will say to themselves, ‘yes, but is she really technical?’. They will be the tech lead who will choose to give the challenging engineering project to the early career engineer just because he looks like ‘he knows what he’s doing’, he has the benefit of the technical doubt.

While working at Big Co, an internal user was asking for help in my team’s technical chat. I offerred a solution and the user came back asking the exact same question, but directing it specifically to male colleague, who gave the user the exact same answer I had posted. At the same Big Co, I went downstairs to talk to an engineer from an upstream team who had introduced a bug into our codebase and thus caused a production outage. He refused to listen until my male colleague came with me.

You can have thick skin and learn not to take these things personally, even though they clearly are about nothing else than what’s in my underwear and yet you can acknowledge that pushing this rock of “technical chops” up the damn hill is a Sisyphean task. Everytime you think you’ve finally proved yourself, along comes yet another hill, the rock rolls back down and the process starts all over again.

I’m not good at tech, so it’s no surprise I’m still pushing the damn rock up the hill and will continue doing it until the end, but there are underindexed folks who are really good and they are still stuck doing the same damn routine just because they never get the “benefit of the technical doubt”. They never get assigned on the stretch assignments or the important customer meetings, because it’s important that “they prove they have the technial skills” before they can be given these challenges.

the tableflip club

We are told to lean in to get a seat at the table. But who built the table? Who decided how many seats there are around it? Who designed the structures that support it?

A few days ago, the AI Now institute published its report on the diversity crisis in AI - opens a PDF file. Among a wealth of material, it contained an interesting passage about the work of Sarah Banet-Weiser,

As she [Banet-Weiser] describes it, “the inclusion of women becomes the solution for all gender problems, not just those of exclusion or absence. It is, of course, important to have bodies at the table, but their mere presence doesn’t necessarily challenge the structure that supports, and builds, the table in the first place”

The phrasing - the extended metaphor of the seat at the table, which the gospel of Lean In tells us to desire - remained with me. It echoed back to 2015, when someone started a tumblr called table flip dot club. “Women are leaving your tech company because you don’t deserve to keep us around.”, the manifesto begins. It goes on to describe what is essentially a big fuck-you to the politics of the Lean In movement- it’s a table flip movement. We’re done with bargaining for the smallest corner seat at the table, we’re flipping over the table and starting over.

It was a co-incidence that this same turn of phrase found its way into the AI Now Institute’s report in the form of Sarah Banet-Weiser’s words. In all of this discussion about getting more diverse voices at the table to build large data hoovering systems, we’re not stopping to ask who built the table and why? Who is keeping it propped up and for what purpose?

When we make an argument for diversity at that table, we should also be talking about the table, who built and why should we seek their approval to join the table.

This is what I am hoping the next generation of movements for inclusion in the technology industry will achieve: not just helping diverse folks get seats at the table, but rebuilding the table completely from scratch.

the dispossessed

Yesterday, on the train back home, I finished the last few pages of Ursula K Le Guin’s the Dispossessed (1974).

I’ll try not to spoil too much (if game-of-thrones-spoilergates are anything to judge by, we are living in a world, where our stories, just like our plastic, have become very much single use) by recounting an encounter that happens between Shevek, a resident of the planet Anarres and Keng, an ambassador from Earth to Urras, the sister planet of Anarres.

Shevek’s society is built on a form of non-authoritarian socialism:, while the society he visits on Urras, in the land of A-Io, is a tightly regulated capitalist society. At the end of the novel, through a series of unfortunate circumstances, Shevek finds himself asking the ambassador from Earth for help and in the conversation describes the world on Urras as a kind of hell. She says,

“Now, you man from a world I cannot even imagine, you who see my Paradise as Hell, will you ask what *my* world must be like?”

He was silent, watching her, his light eyes steady.

“My world, my Earth, is a ruin. A planet spoiled by the human species. We multiplied and gobbled and fought until there was nothing left, and then we died. We controlled neither appetite nor violence; we did not adapt. We destroyed ourselves. But we destroyed the world first. There are no forests left on my Earth. The air is grey, the sky is grey, it is always hot. It is habitable, it is still habitable - but not as this world is. This is a living world, a harmony. Mine is a discord. You Odonians chose a desert; we Terrans made a desert….We survive there, as you do. people are tough! There are nearly half billion of us now. Once there were nine billion. You can see the old cities still everywhere. The noes and bricks go to dust, but the little pieces of plastic never do -they never adapt either. We failed as a spceis, as a social species. We are here now, dealing as equals with other human societies on other worlds, only because of the charity if the Hainish. They came; they brought us help.”

I came to think of this phrase over and over again today as I thought about the findings of the upcoming report on the state of biodiversity in the world. It seems, that in 1974, we were already aware of the facts and we chose to do nothing, or not nearly enough to avoid the climate crisis we are in now. I remember being in school and making colourful cardboard posters about endangered species and climate change, copying those hard words from workbooks, and then standing in front of the class explaining what was what. A few years later, I started studying chemistry and Gratzel cells - the long weekends in the laboratory were driven by the need to make a way for us as a society to keep going as we were - consuming, flying, traveling more and more - without destroying the planet. Perhaps, I always had an inkling of doubt about our ability to make a sacrifice to keep our planet livable.

Interestingly enough, in Le Guin’s juxtaposed world of Anarres and Urras, it is unregulated consumption, not capitalism on its own, that is as the root driver of planetary destruction. When Shevek first arrives in A-Io, his guests take him around their country for sightseeing.

He was driven into the country in hired cars, splendid machines of bizarre elegance. There were not many of them on the roads; the hire was expensive, and few people owned a car privately, as they were heavily taxed. All such luxuries which if freely allowed to the public would tend to drain irreplaceable natural resources or to foul the environment with waste products, were strictly controlled by regulation and taxation. His guides dwelt on this with some pride. A-Io had led the world for centuries, they said, in ecological control and the husbanding of natural resources. The excesses of the Ninth Millenium were ancient history, their only lasting effect being the shortage of certain metals, which fortunately could be imported from the Moon.

But for our world, ever increasing consumption has become closely linked with out increasing obsession with growth under capitalism. Companies are supposed to post better and better quarterly results, for better and better share prices and dividents. Startups are supposed to grow at eye watering speeds. The adoption of social media technologies must keep increasing at all costs.

In the same class where we worked on coloured posters about endangered species and climate change, were classmates who read Ayn Rand and believed that climate change was a hoax designed to make them consume less. In fact, the first time I openly heard someone saying that climate change was a hoax, I was standing in the home of a former classmate. The door to the refuse bin was ajar and from there peaked the strap of a luxury handbag. I’d never seen anyone so casually dispose of a luxury good, just like I’d never heard anyone doubt the facts that we read in or science books.

Here we are, in 2019, and things have not changed for the better in the past 10 years. I know that there are limits to individual action: eating less meat, flying less, travelling with a smaller footprint, but I refuse to accept that therefore, these measures are pointless. If we do this together, we can reverse course before we make the livable Earth into something that only exists in the tales and myths of our future selves.

no ducks in a row

a long farewell to the tech industry

I suppose one could say that I’m losing hope. In fact, I’ve always been losing hope, one small papercut at a time. Yet, after the publication of the black hole image, the rate of loss of hope has accelerated dramatically. After news broke out of another ‘diversity manifesto’ incident - this time at Microsoft, it accelerated a bit more. Sometimes, I have to wonder, why I keep coming back to a place that clearly believes I can never be as good as my colleagues because of the way my body looks. In an industry that always makes a point to talk about “the future”, “progress” and that dreadful word that in practice means the complete opposite than what says in the dictionary, “meritocracy”, we get quite hungup on someone’s external characteristics that have no bearing on how they will do their job. I thought I could help us fix that, but it looks like no one truly wants to fix the problem unless it’s a quick brushup for an upcoming photo-op.

This change may seem sudden, but in reality I’ve always nursed small kindlings of discontent ever since my first job out of university, which by all accounts, was as close to a dumpsterfire as one can get. It’s always fun to watch as my teammates rate the appearance of a woman job candidate from 1-10, scavenge her facebook account for pics. Another woman candidate’s coding test was passed around after her coding interview so that it could be torn apart, mocked and laughed at. “For the lolz, you know”. Furthermore, you know it’s a complete dumpster fire when the manager of the said team observes the proceedings with a smirk and then laughs it off with a “I guess I should make you stop”. Yeah, I guess, my dude.

The point when those kindlings blew up into a fullblown fire was the day the image of the black hole was published. News outlets had stories of Dr Katie Bouman, the MIT researcher, who had written the algorithm to generate the image of the black hole from observations made by a global team. Within hours, internet message boards were full of messages discrediting her contribution. The supporting evidence was found when someone dug up the Github repo with the code and calculated the percentage of her contributions (in terms of lines of code written). Every person in software engineering knows that lines of code is a shit metric to measure someone’s contribution, but alas, the internet bros were at it again.

Shortly after this debate flared up, Mekka Okereke, engineering director at Google who I follow on twitter, wrote a thread about it. Dr Katie Bouman “survived” this challenge, to quote Mekka, because she had all of her ducks in a row - enough ducks and good enough ducks to satisfy the nasty commetators looking to discredit her - a PhD from MIT and a professorship at CalTech. That is how good, how amazingly excellent you have to be, to survive in this industry.

Amongst all of this, I can’t help but think, that I have finally become one of those ‘mediocre/terrible women developers’ that Amy Nguyen wrote about, the awkward type that won’t fit on any kind of diversity poster (which I suppose was my only purpose to begin with) and in a post-black-hole-gate world, I wonder if I should just call it quits now, before someone realises that I don’t really have my ducks in a row, I have barely any. I never studied CS in university. I don’t have a degree from a well-known place. I don’t have stints of work at the Big 4 (or is it 5 these days?). All I have is this, the work I’ve done, which is ok, the stuff I’ve written which is kind of ok and the talks I’ve given (also kind of ok).

Perhaps all of this sounds too gloomy and self-defeating, but I honestly see no way forward for me in this career. In fact, I can even clearly pinpoint the day when I realised that my progress in this industry would always be slow. As a recent graduate, I participated in a graduate program at a company in London. There were five of us in the group, I think, and I was the only women. Three or four other women (out of a department of 60 or 70 engineers) worked there.

During my graduate program, I had expressed, to my manager, a strong interest to join the team working on our backend component to learn more about the project, but no one on that team seemed terribly enthusiastic about the prospect of me joining. I asked if the senior engineer, who I shall call X, would be available to pair program for a little bit to show me around the codebase. This request was met with a smile and a “X doesn’t pair”. There. That’s it. He just doesn’t pair, especially not with you.

I wasn’t defeated by this. I had previously navigated large codebase without anyone guiding me, so I was confident that with enough overtime, I would be able to navigate this one as well. What really killed it for me, was a conversation that took place in the kitchen during snack time. Another senior engineer from the backend team was talking to one of my colleagues from the graduate program. ‘We’re really looking forward to you joining the team’, he said with enthusiasm and I remember how my fellow graduate’s eyes lit up. I felt jealous, because throughout my program, it felt that no one else, except for perhaps the support team, was enthusiastic when my rotation (we all did those in the graduate program) with the team began. More bad things happened at this workplace, but the stories are too long to put in this post and too sad for a Sunday morning.

Similar patterns reappeared at the places I worked after this. The constant need to prove yourself before acceptance into the team. Clients who refused to accept technical explanations from me until one of my male colleagues corroborated them. It became embarassing to always have to take one of them with me when I went to speak to other teams or users, but I swallowed my pride (and tears), because this was the best for the project. I’m not sure I can keep proving myself again and again at great cost to other parts of life and I’m quite confident that the end result will be the same. Nothing will change and I’ll be relegated to writing and re-writing that script no one else wants to touch over and over again.

If you made this far in the text, thank you! If you’ve successfuly transitioned to another career path after tech, I’d love to hear from you. Please feel free to drop me a note on camillamon[at]gmail.com.

relativity

It seems that at some point, back in 2011, I had a mistaken idea that I was capable of poetry. Today, I came across an old USB stick and in a fit of curiosity decided to perform some digital archaeology. Herein is one of the recovered artifacts.

relativity


white feather morning, you
kissed me with your dead lips 
inhale, exhale, 
I, in a thousand little pieces, 
dissolve. 

Time, 
wait for me. 
the membranes
of memories, like silken strings
tangled at my feet. 

You open like a flower at midnight
stains of life
leek out from your black petals. 
 
I peel myself from you, ripe peach skin
drip, dripping, fluid falling flowing
quiet honey nectar, I am liquid. 

Whispers, kisses, lies
time smothered our goodbyes
like an ink-blotted thumb 
grease fingerprint 

your light is falling in the footsteps of twilight, 
on black velvet darkness
a finger left a trail on still water 

the mind falls within itself, unfolds quivers and disappears, 
are you older or younger, flying at the speed of light
yet tied to me,
I don’t know.