Why I code (Part 3)

In parts one and two I said coding is a skill and an activity valuable for cultural and personal reasons.

By "code" here, I mean a deep technical relation with computers, not just an understanding of commands or syntax, but some structural and procedural understanding of what a computer actually does.

As a powerful form of language it enables distinct modes of thought and gives us a lens through which to see the technological world.

That's culture.

That's actually as important as the constructive uses of code. Programming languages may come and go, but the practice of formal language itself is distinct. In the end all programming languages are the same thing. By the time you've learned your fifth or sixth language (easily achievable by one's mid 20s), the dots connect and you see them as dialects of a bigger picture. It's as timeless as mathematics going back to the logical thinking of Confucius, Pythagoras and Medhatithi Gautama in the fifth century BCE.

So on an intellectual level, coding is "good for you" like eating Broccoli, whether or not it earns you money as an intellectual labourer, or directly solves problems for you. That is a personal benefit.

Those who are never taught it, lose that faculty, or choose to throw their minds to the crows by using "AI", will be excluded from influence on the future of the digital world. Now, it may seem to them that they remain fully in control, chatting away with LLM bots and having large parts of their designs pre-fabricated for them. It's coding, but just faster, right? No. This is not the same argument once made for GUI interfaces, high-level languages and integrated development environments (IDEs) that create a layer of abstraction.

Claiming that "AI" is simply a new level of abstraction looks like a good argument on face value. It implies that developers "simply need to adjust" and will soon come to love it like a new IDE. But even visual no-code IDEs are still transparent and examinable in principle. Something else is at stake here and it concerns fidelity, security, and ultimately who is really writing the code and for whom?

"AI" users are forced to take "on trust" everything that they should (and currently can) verify for themselves. What you cannot verify you cannot trust. Since these systems are made, distributed and controlled by hostile entities with their own political agendaa there is a big problem. Technical people who cannot code are easily manipulated by salespeople, blind policy-makers, wishful thinkers and outright malevolent powers. Young coders learning their skills have their horizons fixed by formative experience of being "close to the metal". Deny them that grounding and you cut them off from their roots.

It seems best to think of "AI" as a powerful high-level wish fulfilment engine, so long as your wish lies firmly within an existing, mapped, templated and approved set of ideas. The "guardrails" are not for the "AI", they're for your mind. Trying to mediate coding with "AI" is like squeezing every creature in the ocean into a tibe of "fish paste". It's the intellectual equivalent of highly processed junk food.

Contrary to the slop propagandists, it is those who "embrace AI" who will be "left behind". They'll be first to experience technological weakening, because - as you should know by now - "AI" is like a drug that damages the mind. You get cognitive malnutrition. Sure, you can promise yourself to "never use the product", but look how that worked out for Tony Montana,

Without digital literacy all technology is opaque magic, a weapon, and if you're not on the "staying smart" side of it, a means of enslavement. As such 'other people's technology' deserves zero trust if not outright suspicion and contempt.

You can approach this on an individual level, with a wild-west hacker ethic of "Be smart or be pwned", carrying around knowledge as a gun in your pocket.

Or better, you can work with others for mutual protection against private technological power designed to dumb you down.

Decent societies aim higher. High trust societies with strong education and long-term planning have better all-round standards of living. They understand it's important to value and cultivate trust. Trust in technology happens only through openness, literacy and accountability. Trust can not be claimed, assured or forced upon people by behavioural templates. The US, China are regimes that allow monoliths of private, secret technological power and are storing up a future of pain and disharmony.

The benefit of understanding and practising code at the individual level is therefore inseparable from higher social and political consequences.

Richard Stallman entreats you to share code, voluntarily and as an act of conviviality. He frames this as a relation between individual developers, never saying the quiet part out loud - about the billions of people who aren't developers - that Software Freedom is a social and political responsibility "of those who can" and who value freedom and democracy, towards those "who can't". It's a commitment to keep alive the common knowledge necessary for a technological society that's compatible with democracy.

This matters because technocratic tyrants have proudly said their part out loud, and very clearly; That they do not believe a technological society is compatible with democracy. It's not that they are wrong and they are stupid - a technological society is very much compatible with democracy - but they are signalling that they are eager to see democracy destroyed so they can stay rich by imposing their bankrupt creed of technofascism and surveillance.

People say some dumb and dangerous things when they think they have history on their side and are riding a chariot to a glorious future. The last time people felt cocky enough to publicly speak like that, we hung them at Nuremberg. I will not stand for a bunch of jumped-up American college drop-outs insulting the values three generations of my family fought for. I sincerely hope we will see the broligarchs at least jailed.

In place of "I was just following orders" the unacceptable excuse that won't save them next time around will be "But there was a market for it".

Anna Rowlands, a member on the panel of The Vatican's Magnifica Humanitas launch recently wrote that "the text is neither anti-technology nor anti-industry.". I agree with the former, but that is not my reading and I find the latter "diplomatic" quip to be dishonest. To the extent that "industry" is now openly anti-democratic and comfortable with ending all human life on Earth for money any realistic ethics must come out as anti-industry.

There are alternative ways to remake industrial society, and when its over-reach gets too much as in previous eras, it can be very effectively dialled back.

So, there is a turd in their punch-bowl.

Lawrence Lessig's prescient understanding that "code is law" is too often narrowly interpreted. Code is a super-structural reality that cannot be wished away, hidden or avoided. It is the ability to understand, rewrite, out-think, define, circumvent, or explain code that is the source of pragmatic power.

Code is law. Write it.

I'm not talking about the power of hacking as an intrusive and destructive tool. The fact that we've built a digital society so unbelievably fragile that a small coordinated group of well placed system admins and hackers could bring the entirety of it to a screeching halt in twenty four hours, is worrying. No. The bigger gut-punch is that, ultimately all power and money must defer to the trousers of reality - that knowledge is king. Who writes the code runs the joint.

Important results can't be faked. Hare brained ideas that wreak havoc in the real world are easily dreamed up by those who who can't code, but they depend entirely on other thinkers to implement. Jeff Bezos is one of the few who actually has an engineering degree from a decent university, but he's in a minority amongst "bros" who are mainly skilled at getting other people to do their homework.

One can fairly make the argument that Isambard Brunel or Alec Eiffel were not great iron workers - and it's valid to distinguish an architect from a labourer. But code is a little different. It really matters that you understand systems if you want to adapt and run them for a long time. That usually needs a team for planning and sustainability. Big systems must constantly transform and renew. This creates a situation in which "tech-leaders", and even CTOs are not leading from the front in any meaningful way. They spend all day thinking about how to impress investors and massage metrics to "show growth".

Imposters who ride the coattails of temporary advantage; who bullshit, misappropriate, dabble, pontificate, cosplay as engineers or simply rely on an enormous pile of money to buy respect…are clutching tenuous power that will be easily taken from them.

Having to rely on other people, to share strategies, to ask for other opinions, consent and commitment naturally leads the most monomaniacal to dislike underlings. They'd like to get rid of other people entirely. The ideal "tech bro" company has no employees. It is simply a computer program designed to make money. This can work for a small niche website, but anything more complex that involves real people, serious things, customer support and so on, fails badly.

Business is inherently limited by people. And that is a good thing, because it means there are limits to how far pure money-seekers can push their greed and madness.