Be a loser

I think by now very few of you will have missed the controversial keynote session DHH delivered at the Rails World 2026 conference. Don't worry, I'm not here to remind you that DHH is a shameless fascist.1 That he uses keynote speeches as if they were sponsored talks.2 That he seemingly forgot that he was talking at a Rails conference.3 Or that he bragged about writing a lot of code in 2026 while at the same time bragging about not writing any single line of code.4 None of that.I'll instead focus on a few key flaws in his apparatus of (poor) arguments and the language he used to construct them (if anything was actually constructed by him, but more on that later).And I'll start with a seemingly unrelated angle. DHH, when he's not promoting himself on a stage, is also a rather successful car racer.Is racing cars an efficient means of transportation?The core assumption in DHH speech is that writing code is fundamentally and exclusively an economical activity. You write code to build stuff that accrues economic value. It's a very utilitarianist view of the discipline and a quite depressing one. Which is surprising for someone who used to tell everyone how passionate they were about the craft of writing good code. His argument can be boiled down to a simple statement: I used to be passionate about writing code, but now that machines can do that faster than me, I'm no longer interested.5 Unsettling, isn't it? But let's go back to DHH's favourite offline activity: racing cars. And let's apply the same economic-utilitarian perspective on it. From that angle, racing cars on a circuit makes absolutely no sense. You spend a lot of time and insane amounts of energy to go from point A to... point A. Yes, you arrive exactly where you left off, a few minutes or hours later. Of all the possible ways to achieve that result, it seems to me that only a loser6 would choose racing multiple laps at insane speed as a means to get there. The most reasonable person would simply stay in point A. The tech-bro version would probably take an Uber to do a single lap while they're posting pro-AI statements on X from the comfort of the back seat. Some of you might already be screaming from your chairs, arguing that the purpose of racing cars is not to be an effective means of transportation for people of good. Or rather, that driving cars can serve many different purposes: utilitarian, aesthetic, competitive, self-determination through betterment, you name it. And you'd be right.You see, DHH opened his talk by reminding us about how photography apparently replaced painters. Except that he focused on a specific type of painting: portraits. He, deliberately or not, completely forgot about all other forms of painting that existed before the invention of photography and continued to exist. Furthermore, while he seemed to imply that the advent of AI-assisted coding will replace all forms of humanly written code by using the unfortunate painting analogy, he didn't seem to realise that very few painters decided to become photographers as a consequence of the technical transformation. If anything, many painters felt "liberated", as they could focus even more on the artistic dimension of painting and less on the purely mechanical and utilitarian dimensions.Someone who races cars in the age of Uber is not a loser.A painter who doesn't give up the brush for a camera is not a loser.So, why would a programmer/developer/software engineer who keeps writing their own code be treated as one?DHH doesn't offer any compelling argument besides the usual and weak because I said so.But let's move on to another flaw: the pretended counter-arguments.Shallow counterarguments (p-doomerism)At some point towards the end of the speech, DHH mentioned that there are some reasonable concerns with this technology. I was almost pleased when he showed a picture of a nuclear plant. That didn't last long.The nuclear plant was a parallel, not a warning.The only counterarguments to GenAI adoption that DHH mentioned were nothing other than the set of escathological tropes used by the TESCREAL folks: the existential risk for humanity represented by a self-improving superintelligence. Something that doesn't exist and is only possible in the mind of some weird futurists.While he mentioned the ridiculous doomerist predictions, for which he used nuclear energy as a parallel, he forgot to mention all the harms that the race to AI buildup is causing to real humans today.No mention of the impacts on the environment, the looming financial bubble, the cost inflation for hardware components, mounting workplace anxiety and distress, skill atrophy, or how the internet has become a scrapyard of single-use slop.He didn't mention how reinforcement learning from human feedback, or RLHF, relies on exploiting low-wage gig workers in Kenya, Venezuela or other devastated countries.If you want to educate yourself on such harms, I do recommend starting with the egregious Empire of AI by Karen Hao. I covered it last February, in case you want a bit more context before committing to a great read.📚 Books I read in February 2026Another case of a tight race between two books for the highlight spot, and some good old sci-fi to complete last month’s trio.Sudo Make Me a CTOSergio VisinoniThe p-bloom vs p-doom is a false dichotomy. They are just the two faces of the same coin. A coin made of pure speculations and sci-fi-infused beliefs. They're a convenient distraction, though, as they allow us to close our eyes on the real, proven, material harms that are being inflicted on humans today.It looks as if DHH and his fellow tech bros are completely fine with those harms as long as they don't personally have to deal with them.Don't just take my words for it. Sudo Make Me a CTO is a reader-supported newsletter. If you enjoy the content and would like to see more of it in the future, please consider upgrading to a paid tier. Thanks in advance! Upgrade to Paid