'Real techies don’t worry about forced eugenics. I learned this from a real techie in the cafeteria of a software company. The project team is having lunch and discussing how long it would take to wipe out a disease inherited recessively on the X chromosome. First come calculations of inheritance probabilities. Given a population of a given size, one of the engineers arrives at a wipe-out date. Immediately another suggests that the date could be moved forward by various manipulations of the inheritance patterns. For example, he says, there could be an education campaign. The six team members then fall over one another with further suggestions. They start with rewards to discourage carriers from breeding. Immediately they move to fines for those who reproduce the disease. Then they go for what they call “more effective” measures: Jail for breeding. Induced abortion. Forced sterilization. Now they’re hot. The calculations are flying. Years and years fall from the final doom-date of the disease. Finally, they get to the ultimate solution. “It’s straightforward,” someone says. “Just kill every carrier.” Everyone responds to this last suggestion with great enthusiasm. One generation and—bang—the disease is gone. Quietly, I say, “You know, that’s what the Nazis did.” They all look at me in disgust. It’s the look boys give a girl who has interrupted a burping contest. One says, “This is something my wife would say.” When he says “wife,” there is no love, warmth, or goodness in it. In this engineer’s mouth, “wife” means wet diapers and dirty dishes. It means someone angry with you for losing track of time and missing dinner. Someone sentimental. In his mind (for the moment), “wife” signifies all programming-party-pooping, illogical things in the universe. Still, I persist. “It started as just an idea for the Nazis, too, you know.” The engineer makes a reply that sounds like a retch. “This is how I know you’re not a real techie,” he says.'

Ellen Ullman, Life in Code

Well, none of the AI-generated code now seeping into those companies' codebase is copyrightable. So we're about to see a lot more free-use code hit the scene, when the chickens come home to roost. (Not the same as open source, but still a shakeup of the current status)

That's a very good question that few people consider. Here is the answer: https://knuckleheads.club/how-google-distorts-the-market/

I used to work at a large global company that was part of a uopoly in its industry. We had a lot of mandatory training about bribery, how its forbidden to offer bribes and receive bribes. A lot of training and processes around data retention, for privacy and also for lawsuit discovery... I guess the training was mandated because of previous egregious violations. A F500, older company likely comes with a long history of mandatory remediations for past transgressions.

> mandatory remediations for past transgressions.

Weirdly, this can actually cause companies that didn't do the bad thing (bribes, etc) to voluntarily, proactively adopt widespread, comprehensive, mandatory, highly-audited training to tell every single employee NOT to do that bad thing. That way, if some rogue employee(s) do that thing, having the audited paper trail proving how very hard the company tried to NOT do that thing, can be invaluable in defending the company. In the event of prosecution it can lower the penalties but even better, it helps in diverting potential prosecution into a settlement / consent decree - as well as negotiating lower penalties and settlement amounts.

This was actually the source of one onerous training requirement at BigCo, which I learned about over drinks with our Chief Legal Officer. Apparently, one company (not us) having an extreme punitive penalty slapped on them generates scary WSJ headlines every compliance training vendor uses on Slide #1 in their sales pitches.

There's no legal requirement to maximize profits. That's a common misconception.

Both could be true. I think Google is milking their main asset and tanking it, before the party ends.

I recomment reading "The Management Myth" by Matthew Stewart, on this topic

Ah, the "colonize the stars solution".

For site-search, I recommend making it a much better value for money. Sites with 300 pages don't really require a search engine. Imagine people with 30,000-300,000 pages where 99% don't ever change (so don't need a re-crawl) and who get 10,000 total visitors a month. That's a wholly different point on the searches x crawl frequency x number of docs matrix that you can still serve profitably.

We've just increased it to 1000 pages, now. We'll see how that goes!

Thanks for the input! We're going at it by cost, which is heavier on crawling and storage than the search itself (though that also costs money, just much less). Crawling and allowing 30k pages indexed for 6 months for €30 sounds like a loss at a glance, but we're open to experimenting with that for someone who needs more pages and the same amount of searches. Just reach out!

I'm creating a piece of "home cooked software" - a multi-player web game to play with my team at your monthly Social. Made for a specific team, for a specific occasion.

Its janky, and I made several time-saving decisions that'll cost me more time in the end, but I'm having fun.