Not much is working so far.... no world records were broken.

Only in New Zealand!

Friend, I see that you're fed up with the madness. It's tough to know that the world isn't quite as it's portrayed to be! It's tough to raise kids knowing that things could be so much better! Don't feel alone! There are many of us in the shadows who are fed up with these lies.

We all know the drill: The Earth is round and the Sun is in control of daylight, but not everybody gets the same amount at the same time. The capitalist politicians are just trying to optimize it for everybody, by giving out time to each according to their needs.... I don't know about you, but that idea had never sat right with me. Why can't we all get the same daylight all year?

I know what you're thinking, I know how this sounds, but trust me, I've been there. We've all been there. Please hear me out... What if there was an alternative? What if.... the Earth was Flat? Wouldn't that be great? The sun's rays would shine on everybody at the same time. Everybody would get the same daylight, the same darkness. No more privileges! When it's Noon in one place, it's Noon everywhere! The world is simple, and kids can finally stop learning arithmetic, with those confusing modulo 12 operations. We'll never need those world clock iPhone apps ever again!

Friend, here's my card. It's time to act. We meet every Wednesday afternoon at 3pm UTC. There's coffee and cakes. We just discuss things. How to make the world a better place. How to make the Earth completely Flat. For the sake of the children. All our children. Who hate arithmetic. Even the weirdos who don't. I'll see you there? Good.

It's not a bad analogy, except the database in this case is read-only:)

The way to make the database writable is to modify the training data, ie pages on the web etc. No human looks at the training documents to find potential issues with it, the models simply fit their errors to minimize the discrepancies.

To do it properly you'd have to:
1) Identify the sources of training data (main percentages, to get a handle on the scope)
2) Construct fake but free data sources (using any random LLM - true data is not the point) and replicate them widely.
3) Wait until the public model updates, then test the output to pinpoint areas where degradation was successful.
4) Rebias the fakery to reinforce the degradation.

Seems to me like "everyone with a brain" turned out to be wrong in the choices they made since then.

That's ok, there's a word for the outcome of that: "quiet quitting". Followed by "OMG, the Chinese are beating us!"

Nonsense, plenty of us here make that kind of money before bonus. What you don't get is that when the Titanic sinks, both the poor and the rich drown at the same time. And nobody likes living in a city where you can't cross a certain street because everything beyond is gang territory. And even fewer like some rich blowhard lining his and his friends' pockets without doing any work while fucking the country.

You're only doing first derivative prediction, aka "what if everything stays exactly the same, but this one single thing changes?" The real world operates in cause-effect, force-feedback, action-reaction modes. In other words "what if one single thing changes, how will everything change to adapt and exploit the change?"

With an ever growing number of unhappy and poor residents whose time is free from the constraints of working a job, new possibilities emerge to cannibalize the system. The usual single phrase used for this kind of society is "failed state".

What is life without beer, steak, and side of potato wedges.

Beer, steak, a side of potato wedges and SPAM, thankyouverymuch!

We should call them cycles.

Example usage: "Oh oh, looks like the AI is back on it's cycle again!"

I only watch pirated streaming content! Does that mean I can get paid 15% of Netflix's contributions?

Just ask Grok to make it bigger and say it's real!

All mathematicians build math by leveraging what other mathematicians have done

No.There's no leveraging. The purpose of research papers in mathematics is to advance the frontiers of knowledge. That requires, by definition, original contribution. It's about as far removed from "leveraging" as you can get.

Engineers (of the software kind, typically) don't always understand the difference between using something someone else has done for solving a problem, and creating a new original solution to a problem. The former is "leveraging".

For example, if you take out a pocket calculator and calculate 2343842374983275489374+7=??? then you have leveraged the work of the calculator's designer. You have also solved a numerical problem that nobody else has previously discussed. But you haven't done original work. Even if the above calculation was not published by anyone before.

Do you the correct words, and you will not make such elementary errors in public.

I think you missed the example I made: 1) make a bet in a casino. Now you have a stake. 2) make a second bet about the outcome of the first bet with the fellow gambler sitting at the table next to you. Now that second bet is not, ipso facto, a hedge. But it is a bet about an event that you have a real world stake in. So a real world stake is not a classifying feature.

Hedging is imperfect and costs money, but losing the money doesn't make you happy. You are only happy if the sum of the outcomes is positive, which includes the money lost on the hedge if any. But you might not be happy even then, if the sum of the outcomes is *insufficiently* positive. So happiness with a loss is not a good criterion either.

Professional gamblers care about edge. They will happily lose bets as long as they think they had edge. This is because they already priced the loss into the game and believe it is worth it. This is still a gamble, and in fact some gamblers will be *more* happy every time that they lose, because they believe, perhaps erroneously, that this makes the next win more likely.

You don't have control over users. For my ancient Free projects, I've always hosted them on my own personal site. However because they are GPL, some people "helpfully" host a copy on GitHub. I don't particularly like it but they've done it. I don't update their copy or fix any bugs there.