Malus – Clean Room as a Service

I mean in the context of AI - we're already seeing the conflagration of SAAS, and software jobs are going kaput. It's my deeply considered opinion that the faster this happens, the better, because it'll force a reckoning with impending AI job loss across the board.

We need to deal with the issues now. The worst possible outcome is a gradual drip-drip-drip of incremental job losses, people shuffling from job to job, taking financial hits, some companies pretending everything is fine, other companies embracing full-bore zero employee work. The longer it goes on, the more wealth and power gets siphoned up by corporations and individuals who already have significant wealth, the bigger the inequality, and the bigger the social turmoil.

Software, graphics design, music, and video (even studio level movies) should cope with this now. It's not going to stop, AI isn't going to get worse, there's not going to be some special human only domain carved out. The sooner we cope with this the better, because it'll set the foundation for the rest of the job loss barreling down on us like the Chicxulub asteroid.

Certain views of OSS and its relation to commercial software always seemed to be fraught with highly voluntarist and moralizing attitudes and an intellectual naivete.

An interesting aspect of this, especially their blog post (https://malus.sh/blog.html ), is that it acknowledges a strain in our legal system I've been observing for decades, but don't think the legal system or people in general have dealt with, which is that generally costs matter.

A favorite example of mine is speed limits. There is a difference between "putting up a sign that says 55 mph and walking away", "putting up a sign that says 55 mph and occasionally enforcing it with expensive humans when they get around to it", and "putting up a sign that says 55 mph and rigidly enforcing it to the exact mph through a robot". Nominally, the law is "don't go faster than 55 mph". Realistically, those are three completely different policies in every way that matters.

We are all making a continual and ongoing grave error thinking that taking what were previously de jure policies that were de facto quite different in the real world, and thoughtlessly "upgrading" the de jure policies directly into de facto policies without realizing that that is in fact a huge change in policy. One that nobody voted for, one that no regulator even really thought about, one that we are just thoughtlessly putting into place because "well, the law is, 55 mph" without realizing that, no, in fact that never was the law before. That's what the law said, not what it was. In the past those could never really be the same thing. Now, more and more, they can.

This is a big change!

Cost of enforcement matters. The exact same nominal law that is very costly to enforce has completely different costs and benefits then that same law becoming all but free to rigidly enforce.

And without very many people consciously realizing it, we have centuries of laws that were written with the subconscious realization that enforcement is difficult and expensive, and that the discretion of that enforcement is part of the power of the government. Blindly translating those centuries of laws into rigid, free enforcement is a terrible idea for everyone.

Yet we still have almost no recognition that that is an issue. This could, perhaps surprisingly, be one of the first places we directly grapple with this in a legal case someday soon, that the legality of something may be at least partially influenced by the expense of the operation.

Let's not spam HN with AI slop please.