Published on: 2020-09-13
Code as pointless as tea
I'm a fan of tea and cinnamon, but I dare say my enjoyment of these things wouldn't be motivation enough to go on a murderous crusade across the globe and fund the world's biggest empire.
Granted, it wasn't only tea and cinnamon that motivated the British, Dutch, French, Belgians, Spanish, and Germans to conquer the world. It was also silver, gold, exotic wood, peppercorn, silk, and coffee.
But on the whole, it always seems odd to me that there was this 400-year gap in whig history, where technology and social cohesion was progressing at an astounding rate, with its almost sole driver being to fund the colonialist armies that supplied spices to the aristocrats.
This is only made more surprising by the fact that this period seems like a sucky one to live in.
Don't get me wrong, the middle ages weren't all fun and dandy, but on the whole, I can see worst fates than being a peasant doing that time. Life expectancy was up into your 60s once you made it past 4, wars were seldom and not bloody and your diet would have made a lot of modern health nuts jealous. Add to that a relative lack of plagues, reasonably lose sexual taboos, and having a strong community that's there for you both in good and bad times.
Compare that to the first and second industrial revolutions. Smog is covering half of Europe, children as young as 10 are working 14-hour shifts in factories, the countryside is overpopulated. The cities spawn so many plagues that it's regarded as a common fact of life. None of the advances of medicine are upon us, clean water has become a commodity, and food for the "working classes" is so vile a modern-day human would be forgiven for preferring starvation.
Now consider the fact that Europe was "the oppressor" in this period. People here were faring "better" than the ones being massacred and enslaved by our guns and germs.
Yet during all this time, somewhere in the background, not heavily incentivized but developed out of curiosity and sometimes necessity, modern physics, chemistry, biology, mechanical engineering, and medicine are born.
This is outright infuriating, it's not a story of progress, it's a story of luck. Of luck that the pointless slaughter in search of spices and shiny rocks somehow resulted in a few visionaries being able to put the scaffoldings for a better future.
This is kind of how I feel about the modern world and the way we are using hardware and software.
Improving hardware means we can run software faster and gather data at ever-increasing rates and resolution. Improving software gives us more complex algorithms to extract insights out of that data and solve coordination problems between people.
And by golly, solving coordination problems and gathering insights about the world seem like a good idea since we have some fairly serious problems in those areas.
I don't want to draw attention to any particular political problem a given sociopolitical tribe thinks is of great importance at the moment, but rather to the broader context.
Every day it feels like more pieces are falling in place and we are getting closer and closer to solving problems we would have thought to be the realm of actual Gods when the 20th century was in its infancy.
There are real opening through which we can tackle the problem of aging, essentially solving most diseases and potentially defeating death itself.
Visionaries among us are cracking the problem of traveling over long distances and surviving in harsh environments. So much so that a Mars colony now seems like an overdue reality rather than the realm of science fiction.
We have machines that can automate the most physically demanding and repetitive tasks away from us. Be them driving, mining, working in warehouses, working a checkout counter, or building homes. It feels like all modern humans could be philosophers, poets, mathematicians, musicians, outright hedonists. Living a life full of creating meaning and exploring the joy-inducing limits of consciousness.
Most if not all parasitic and bacterial diseases, together with the hunger that's plaguing the less developed countries, seem like a matter of funding to eradicate. Some extra mosquito nets and better insecticides would be enough to rid the world of a disease that is killing hundreds of thousands and making hundreds of millions more suffer terribly each year.
On the other hand, we face an ever-looming specter of death and misery. Or rather, we face several. From the damage being often pointlessly inflicted upon our soil and atmosphere in order to cut costs. To the ever-growing population bubble in places that will be hardest hit by environmental calamities, to the literal arsenal of humanity-eradicating weapons that China, Russia, India, Pakistan, the UK, France, and the USA are pointing at each other. A few dozen clicks or 2 unlucky software bugs away from triggering Armageddon.
Ok, maybe I am taking a bit too much poetic license here.
Maybe the scary things aren't so scary. Maybe the utopian vision is not so close.
Still, one gets a feeling we could be doing better. Maybe not all of us, but those of us working on software certainly could.
Have you had a look at Unicron startups lately?
Other than SpaceX, the US and Europe have Strip, JUUL, Palantir, Airbnb, Epic Games, Instacart, Wish, UiPath, Robinhood, Coinbase, DoorDash, WeWork... etc.
These are not companies working on humanity-advancing problems, these are marketing-driven monsters trying to create monopolies around phone banking, vaping, home-delivery, crypto trading (which I find particularly disturbing since the whole value of crypto comes from decentralization), ad data collection, and short-term property renting.
And when one looks at the tech giants, companies like Google, Apples, Facebook, and Twitter, the situation becomes even more dire.
A sea of talented people working on the most redundant of problems, how to keep a monkey glued to a dopamine-inducing screen in order to better imprint advertising material into its brain.
From this sea of pointless rehashing of crab-bucket-economy ideas a few interesting things arose. After all, Google is funding Deep Mind and creating JAX, and google maps. Facebook is funding the development of PyTorch. And apple is... certainly doing something useful, it initially put a lot of the money behind the original development efforts towards clang & llvm.
But any and all advancements that will bring us closer to a utopian future are incidental, happy accidents not core goals. It is indeed a happy coincidence that the same techniques used to auto-label photos and separate them into albums lead to miraculous advances in molding in-vitro protein folding. But make no mistake, 99% of the progress that we have right now are happy coincidences.
Neural networks are praised as an innovation of modern ML/computer-science that has near-universal applications and much-untapped potential. But autograd and the kind of tensor-graph based computation being used have been "a thing" since the 50s.
It isn't researchers that made neural networks viable and useful, it's GPU companies. And the fact that Nvidia stumbled upon this market is in itself a fortunate accident. The purpose of GPUs was to generate higher framerates and more detailed graphics for people playing World of Warcraft as part of their doritos&monutain-dew fueled run on the hedonic treadmill. It is a stroke of luck that this same hardware proved adequate enough for solving problems that might revolutionize the way we do science, math, and intellectual labor.
Much like the first and second industrial revolutions happened to have small niches where better science was required to speed up progress, so does the internet revolution.
But much like the driver of the first and second industrial revolution was a pointless quest for spice, so is the engine of our current era a pointless quest for click generation and on-demand indoctrination.
We are none the wiser or happier, indeed one could argue our environment is making us dumber (well, at least one could if our data point was every single military-aged man in Norway, but I'm not sure how representative that is). We must go further and further back in history in order to consider our station in life to be "favorable" compared to that of our ancestors.
I don't think a solution to our current predicament doesn't exist. Nor do I think this is a particularly novel or desperate situation.
Progress seems to inevitably come from happy accidents and small niches of insane people drudging away at seemingly pointless problems. There never was and never will be a kingdom founded on the pursuit of a better life for those who are yet to be born.
Indeed, I think that we live in a time more hopeful than any, where the tech economy has allowed smarter-than-average people such as Bill Gates to accumulate large amounts of wealth with which they can fund utopian projects.
Even more so, we have spare time more than ever and a lot of computational resources available to almost everyone.
But I think the message I'm trying to convey, is that we shouldn't forget there is software which has and will continue to be critical for the advancement of the "noble" goals of humanity, be those spreading into spaces or alleviating suffering and bringing joy here on Earth.
It includes things like the linux kernel, llvm, rustc, wikipedia, autograd, coq , and thousands of other such projects. Almost all of them worked on by volunteers or by people taking below-market wages and working long hours, funded by the charity of governments and tech giants.
But for every person working on world-changing projects, there's 100 more working on perfecting banner insertion into mobile apps and web pages. For every person working on important software, there's 100 working on the 9th iteration of an online-store frontend. For every SpaceX and DeepMind, there's 100 Ubers and Airbnbs.
Progress, it seems to me, is not the focus of our society or the driver of our economy. It has never been so and never will be, a few people tried to make it such and the only thing that came to be of it was poverty, hunger, and the murder of hundreds of millions. The reasons for why this is so elude me and could be better laid out by better thinkers.
For now, we are stuck in this imperfect world, where dozens of millions write software that is about as pointless to the well-being of future humans as the 15-19th century spice trade was for us.
This also means that there's plenty of avenue for making a significant contribution, and one shouldn't get disheartened by their only meaningful work being a weekend project. Our society is built on visionaries working on pointless tat, with a few interesting weekend projects that gave birth to what an 18th century Londoner would view as a utopia beyond their wildest dreams.