Mastodawn

nixCraft 🐧Jan 4

This chart shows the total number of Stack Overflow questions asked each month. As you can see, AI summaries in Google and AI coding tools have nearly killed the site. It is only a matter of time before the site shuts down completely. The golden age of independent news, blogs, forums, and specialized sites like Stack Overflow is over. Whether this is good or bad, only time will tell. Personally, I think we are now restricting all internet traffic to just a few Gen AI apps https://data.stackexchange.com/stackoverflow/query/1926661#graph

Independent sites cannot compete with companies like Google, Microsoft, or OpenAI, which have unlimited money. Even Wikipedia has mentioned this problem. They are losing a massive amount of traffic from Google. These AI apps do not give credit or send users back to the original source on the web. As a result, people are stopped from creating new Q&A or editing and adding information to Wikipedia.

@nixCraft [citation needed]

Till Kleisli Jan 4

@a2800276 there is a link in the post... @nixCraft

I'm probably too thick to see it, this is what I was responding too:

"Even Wikipedia has mentioned this problem. They are losing a massive amount of traffic from Google. ... As a result, people are stopped from creating new Q&A or editing and adding information to Wikipedia."

Till Kleisli Jan 6

@a2800276 oh, sorry, I thought you were referring to the first post about stackoverflow.

danyal barakah Jan 4

@nixCraft it's really hard to have such conversations when still those companies are the ones running social media. we have to unlearn old media use methods, let's all run towards independent sites 🙌🏾✨.

TeqFu - Internet Solutions++Jan 8

Cheers to that concept. Every client should be a server to reduce gatekeeping. So on with independence and self-hosting. That's what we've chosen!

danyal barakah Jan 12

@TeqFu true.

@nixCraft I was watching a video the other day, and the creator while carrying out a technical process, turned around and said "while I do this, I'll let ChatGPT summarise the method' then read out a method... instead of reading the passage from the technical manual, and citing the author. I guess the point I'm trying to make is that it's not just websites, many people now lazily rely on LLMs as their primary source of information

Onno (VK6FLAB)Jan 4

@robchapman was this a blacksmith discussing induction forges by chance?

Jens Oliver Meiert Jan 4

@nixCraft, they have unlimited money *and still* they steal and are *allowed to steal* the training data from everyone else (which is all of us).

Seen from a property perspective, what we permit here is scandalous. Uninhibited information exploitation for uninhibited wealth extraction.

And then, we do care—act—about so little that we cannot say we wouldn’t be part of the problem. We’re sheep being sold our own wool to.

@j9t @nixCraft We’re being fed mutton.

lazyraccoon83 Jan 4

@nixCraft perhaps that is the exact solution - legislation that forces all search based LLMs to provide hits, include commercial content or exposure provided by said website mandatory, or maybe just direct compensation in a b2b model. This data could be used in a class action lawsuit.

The Penguin of Evil Jan 4

@lazyraccoon @nixCraft it's been said from the start by copyright maximalists that 'AI' search should be giving answers like 'according to XYZ there is a relationship between A and B, would you like to purchase the paper', or 'I think the song you mean is Foo, would you like to play it on spotify.

Now it seems however we are going to go from no credit to 'Here are five songs from our advertisers that are vaguely related to your query'.

@nixCraft

Of course there's no game. But do Google and a community of, just to say one, programmers or people passionale about anything, really need to measure the same metrics? All went to shit when said communities sold their soul to VCs since they only know the metric of money.

I personally see the future with a reflourishing of small (probably hyperfocused) communities run for the knowledge and experience sharing and not for profit.

Carsten Hoffmann Jan 4

@nixCraft while you have a point, what you write dos not apply in full intent to all of them. ChatGPT for example gives citations and links to web findings so the user can jump to the original content when performing web research.

@morl99 @nixCraft

Yeah, they can jump to an Ai-slop filled website optimized for Ai crawlers. Or to an advertised website? Or there is any way to tell what is what if you never had to scroll three of four pages of search results on old search engines? You gotta develop some instinct and sensitivity, why should you stop at the first citation by an LLM and call it a day?

stefan (stefbun)Jan 4

@nixCraft yeah consolidation of information / power. great news (for the shareholders)

@nixCraft so it consumes itself to stagnation. Societal intellectual cannibalism.

@nixCraft I think Wikipedia and SO are good things to compare.

Wiki got it's annual funding targets met early the last two years. It's strategy is to be good enough that people pay to keep it going.

SO, even though I'm fond of it, follows the VC model of somehow turning lots of users into revenue and that hasn't really panned out.

There's a cohort of people who value scrutinised information and they're funding Wiki.

@nixCraft hmmm so without anyone creating new reliable information (and there is always new stuff to ask and write about, things change), how will LLM apps maintain relevancy? This doesn't seem like a good development for them either. Some kind of equilibrium between predator and prey is required...

Alex Stabing Jan 5

@nixCraft AI-Qaeda

TeqFu - Internet Solutions++Jan 8

@nixCraft THIS is a problem, because AI often quotes popular opinion over facts, even in legal advice, quoting jurisprudence above actual law. A huge percentage of people will never verify until their eyes are opened.

We must keep factual sites and archives alive despite the AI overreach until humanity catches up.

Our kids need us to be immovable despite AI hype, slop, and bias, if we shall ever want to maintain some bit of truth.