In the winter when the sun might set at 4:30 pm, it’s pretty ridiculous to try to stop eating before sundown.

The way that sort of invention often works is:

1. Inventor thinks they have a world changing idea
2. Inventor spends their own time and money to build a prototype
3. Inventor shows the product off to the world.

If it truly is a world changing invention, step 4 is “world is amazed, inventor can’t keep up with demand”. There are also frequent cases where the world goes “meh, not for me”. Now occasionally those are when an invention is ahead of its time, and years or decades later the inventor is vindicated. The other case is when the invention really isn’t good, and there simply isn’t and will never be demand for it.

Somehow, the AI bubble is built with people ignoring the feedback from people that keep saying “meh, not for me”, and the various “inventors” burning more and more of their money trying to change people’s minds. Has that ever worked?

Has there ever been a true world-changing invention where the “inventors” had to beg the public to use it?

Capitalism says that the market won’t reward people making those things, and the companies might fail as a result.

But, we’re no longer in a capitalist world. We’re in a corporatist world where it’s closer to technolfeudalism where it doesn’t really matter how bad your idea is, because you aren’t out to make a profit, instead you’re out to extract rent.

This is the thing I hope people learn about LLMs, it’s all hallucinations.

When an LLM has excellent data from multiple sources to answer your question, it is likely to give a correct answer. But, that answer is still a hallucination. It’s dreaming up a sequence of words that is likely to follow the previous words. It’s more likely go give an “incorrect” hallucination when the data is contradictory or vague. But, the process is identical. It’s just trying to dream up a likely series of words.

I’d rather program a normal way than try to wrangle some of the abominations I’ve seen in excel sheets

That’s the way I also think about learning fancy spreadsheet stuff. Spreadsheets are good for putting data into a graph. They’re good for basic numeric stuff where there’s a simple pattern that repeats. But, pretty soon you’re in a situation where you should either have a real database or a real program. If you’re doing a lot of manipulation of data, you should have a program with loops, conditionals, errors, exceptions, etc. and most importantly with comments. If you’re storing a lot of data, you should be using a real database, not hundreds of lines in a spreadsheet.

If, at the end, you do want something visual, and don’t feel like dealing with a graphics library, you can always export the data to a CSV and import that into a spreadsheet.

Let’s just say AI truly is a world-changing thing.

Has there ever been another world-changing thing where the sellers of that thing had to beg people to use it?

The applications of radio were immediately obvious, everybody wanted access to radios. Smart phones and iPods were just so obviously good that people bought them as soon as they could afford them. Nobody built hundreds of km of railroads then begged people to use them. It was hard to build the railways fast enough to keep up with demand.

Sure, there have been technologies where the benefit wasn’t immediately obvious. Lasers, for example, were a cool thing that you could do with physics for a while. But, nobody was out there banging on doors, begging people to find a use for lasers. They just sat around while people fiddled with them, until eventually a use was found for them.

In case someone thinks “Oh, he’s just an alt-right troll, he doesn’t actually believe in that stuff”, meet Elon’s grandfather:

He became involved in Canadian politics, backing the technocracy movement, before moving to South Africa in 1950. Over the course of decades, Haldeman repeatedly expressed racist, antisemitic, and antidemocratic views.[1] In South Africa he was a supporter of apartheid and promoted a number of conspiracy theories.

https://en.wikipedia.org/wiki/Joshua_N._Haldeman