I fixed Louie Anderson's iphone settings once while we were waiting for a flight. He gave me his e-mail and i kind of wonder what would have happened if I'd reached out.

Make sure you name, save, and enable in the left menubar, and reload your docs page to test.

Style to your hearts content! Mine looks like:
/* Hide "Refine" pop-up when text is selected*/
#docs-instant-bubble {
display: none !important;
}

/* Hide "Ask Gemini" button*/
.appsElementsSidekickEntryPointRoot {
display: none !important;
}

(2/4)
Step 2: make a custom style for docs.
* Go into the Stylus extension menu and hit plus for new style
* Change the "Everything" dropdown above the text box to "URLs starting with", and put in docs.google.com/document/ for docs

Fair. I'm not arguing at all that they didn't accept dirty money.

Giving them faster speeds *also* saves server load, which makes Wiki better for everyone. A version of this fast, complete dump is also publicly accessible.

Getting a broad base of Wikipedia editors is definitely an issue and one of the reasons I am engaging here. We need more women! We need more non-CS folk! If people think wiki is selling out, the groups less willing to join the conversation are going to be the folks we need most.

The big companies aren't the ones ddosing wiki. It's mostly smaller open source models, many from China. I have the graphs if you're curious :) I'm guessing the big companies are interesting in keeping a known clean data source fast and accessible.

For context, I'm also a writer, and i do a lot of academic work warning against ai content proliferation. AI content in wiki is something I'm personally actively working against and have been for over a year.

Wiki are the people getting the money, not giving it. There isn't extortion. Crawling the content for model training is allowed under the CC license the content is under. Wiki has had years of heavy AI crawler traffic. The only change post deal is that Wiki has convinced some folks to help pay.

In case it helps, this isn't "AI generated articles". This is wikimedia negotiating with large companies to get paid for the bandwidth they use crawling the site. All they are giving the companies are faster downloads and some custom data formatting. The data itself is public.

Congratulations!

I will die on the hills of sweet peppers with the onions but otherwise 100% with you.