Welcome back to the board!

[bobalot]

That's it.

Also, fuck whoever DDOSed the board.

[LadyTevar]

I'm still wondering WTF they'd DDOS **THIS** Board.

[bilateralrope]

I'm still wondering WTF they'd DDOS **THIS** Board.

Could be all those AI companies checking the board far too often for new data to shovel into their AIs. I've seen articles about that causing problems elsewhere.

[bobalot]

I'm still wondering WTF they'd DDOS **THIS** Board.

Could be all those AI companies checking the board far too often for new data to shovel into their AIs. I've seen articles about that causing problems elsewhere.

That might explain the number of visitors being "40550 on Mon Mar 17, 2025 8:08 am".

Another thing "AI" has ruined.

[Alferd Packer]

I'm still wondering WTF they'd DDOS **THIS** Board.

Could be all those AI companies checking the board far too often for new data to shovel into their AIs. I've seen articles about that causing problems elsewhere.

It is 100% this. Other less-than-active forums from my youth of which I'm a member see the same thing happen, and one admin said it was an AI company scraping the whole forum.

In that forum's case, I question how effective a model will be trained if it's based on hundreds of thousands of posts from two decades ago, consisting largely of teenagers arguing about politics and calling each other homophobic slurs, but what do I know.

[Lord Revan]

I was wondering why the board was down granted never considered DDOS.

[bobalot]

I'm still wondering WTF they'd DDOS **THIS** Board.

Could be all those AI companies checking the board far too often for new data to shovel into their AIs. I've seen articles about that causing problems elsewhere.

It is 100% this. Other less-than-active forums from my youth of which I'm a member see the same thing happen, and one admin said it was an AI company scraping the whole forum.

In that forum's case, I question how effective a model will be trained if it's based on hundreds of thousands of posts from two decades ago, consisting largely of teenagers arguing about politics and calling each other homophobic slurs, but what do I know.

AI is largely useless and makes no money.

[bilateralrope]

In that forum's case, I question how effective a model will be trained if it's based on hundreds of thousands of posts from two decades ago, consisting largely of teenagers arguing about politics and calling each other homophobic slurs, but what do I know.

No need to question the effectiveness of LLMs. It is already well demonstrated by things like:
- Regular hallucinations. Like all the lawyers getting in trouble for filing an LLM written document to the court that cites cases that do not exist.
- Failing to correctly count the number of r's in strawberry


Right now, they are in the most dangerous stage of any level of automation. When they get things right often enough that people stop checking to make sure they are doing things correctly. Leading to a lot of nasty surprises.

[EnterpriseSovereign]

Even though I first heard this phrase with relation to AI-enhanced CVs, it actually applies to anything written with AI: "own every word", i.e. check everything it's written, instead of just assuming it's perfect. "To err is human, but to really foul things up you need a computer"

[LadyTevar]

Even though I first heard this phrase with relation to AI-enhanced CVs, it actually applies to anything written with AI: "own every word", i.e. check everything it's written, instead of just assuming it's perfect. "To err is human, but to really foul things up you need a computer"

I've not heard that saying in a long time

[EnterpriseSovereign]

Even though I first heard this phrase with relation to AI-enhanced CVs, it actually applies to anything written with AI: "own every word", i.e. check everything it's written, instead of just assuming it's perfect. "To err is human, but to really foul things up you need a computer"

I've not heard that saying in a long time

Nor have I, I will admit I had to look it up to remind myself of the exact wording but it's even more relevant now than when it was first uttered more than fifty years ago.

Like the story I've been working on I often have the AI generate a dozen iterations as I refine the instructions each time since no single one is completely perfect.

[Batman]

'Perfect' is the mortal enemy of 'good enough'.
The 'to err is human but to really foul things up' thing actually seems to predate computers, I've seen this used with 'lawyers' and 'politicians', too.

[EnterpriseSovereign]

'Perfect' is the mortal enemy of 'good enough'.
The 'to err is human but to really foul things up' thing actually seems to predate computers, I've seen this used with 'lawyers' and 'politicians', too.

Reminds me of the old joke where a doctor, an architect and a politician are arguing over whose profession was the oldest:
The doctor said Eve was created from Adam's rib, which was a surgical procedure
The architect said that before that, order was created from chaos, which was a form of architecture
The politician said that somebody created the chaos first