Why? It's going to be my kids first computer.

Computers today are so absolutely hostile as they are simply attention-sucking sale-terminals. They spend all their time popping up unwanted notifications which are just advertising.

And the interface of modern devices is actually horrible for learning. Some stuff may be intuitive, but the biggest issue is that every slight movement, accidental tap or gesture is linked to something so for kids it's too easy to do something that exits the current program or bring up some sidebar. It's impossible even for me to connect "what did I just do?" with the sudden change in context. It makes it really hard to connect cause and effect. And don't even get me started on how dangerous apps like YouTube are for kids. The recommendation algo seems to surface click-farm scam content in no time. Or weird dopamine traps.

So my kids will start with a device that isn't constantly trying to sell things, they will learn to understand simple systems which has deterministic behavior.

It's like a hidden curse of LLMs -- they're so good at parsing intended meaning from non-grammatically-correct language that we don't have to be very good at clear communication.

Eventually all LLMs will be controlled by humans uttering terse gutteral grunts. We will all become neanderthals, with machines that deliver our every whim.

Especially because the companies telling us the first premise is true are the companies which need investors to prop up their business.

I mean, it is possible the first premise is true, but the absolutely bonkers credulity in it really mystifies me. It is an incredibly unlikely thing to be true and we should be demanding quite extraordinary evidence to back it up. But based on some neat tricks by current LLMs, some people are all in.

The great secret of the industry is that it's mostly run by clueless incompetents. And their only strategy is to copy from every other product because they don't have original ideas or enough conviction to lead in a new direction. They just want to show up, punch thier timecard and go home, so we get the laziest suite of product implementation possible.

So the richest, most prestigious schools where legacy admissions are a gateway to the upper classes, will keep the policy.

If you're high up at a company like Meta, you likely have a compensation package worth millions a year.

The question is what are they being paid for if not to be "better" at steering the ship than others? They always tell us they are brilliant leaders who bring more value to the company than others could or would.

So if they're just following the market like everyone else, and having to react with large reversals, then to me, it starts to poke some pretty large holes in this idea that they are somehow the best of the best. It starts to look like their only real skills are self-promotion and career advancement. Not because they're better at operating the company, but because they're better at office politics.

This is nothing new of course, this is the way most organizational structures have worked since the dawn of time. The people with power are given deference and privilege commensurate with being elite, but really they're just average at doing their actual job and kind of guessing their way through it. I'm not saying Meta is special or uniquely culpable for this mistake here. I'm saying it's a sad fact of life and maybe, just maybe, if we all start saying out loud this truth, that this is something we could change as a society.

So the answer is, when an executive is held accountable for disrupting this many people's lives. When they claw back bonuses they have probably received for hitting or setting those previous hiring targets.

I don't disagree with you at all, I have found that I turn to LLMs to answer questions that I would have just searched with Google before.

It feels like a case of companies creating a problem to sell you the solution. The problem in their eyes is that they couldn't squeeze any more money out of search. So they bring us LLMs to replace it at what is sure to be a much higher cost. But they had to torpedo search to force users to use LLMs.

Which for many engineers who got into this industry because they loved solving problems, it can be quite a shocking realization.