@stefano That gets me thinking, if I were AWS looking to scrape sites, pushing more sites onto AWS would make scraping them a lot cheaper and easier since that would be traffic entirely happening on my network and my hardware (plus they could theoretically simply directly ingest the web content instead of scraping it)

@stefano Why does everyone in corporate default to "just throw more money at the problem"?

@stefano this sounds a lot like a conversation I had at $DAYJOB where management wanted reports of wins encountered when using chatbots but didn't have any way to note the losses—time sunk into fruitless things, fixing problems with LLM-generated code, getting wrong answers, accidentally leaking confidential data? Nope, not measured, only the successes. 😑

@stefano I don't believe that's the case, although it's not impossible.

From a budget perspective, it's still not income from AI so you can't use it to convince investors on keeping burning money on it.

I'm more afraid that this may be used as an argument to build some type of "Personal internet ID" granted by some authority, as proof you are not a bot.

@stefano I think it's fair to say that those companies profiting from increased load don't mind it.

Sort of like the general store in a western frontier town during the gold rush, they profit the most. Regardless of success people needed to buy picks and shovels and food. They all got very rich whereas less than 1:10,000 diggers did.

@stefano in my college when i was introduced to ai, i found it very interesting and cool, now i hate it in every way possible...