/4 Why build a German speaking bot network if they don't post strong opinions (yet)?

While this is largely speculation, one should consider that Germany holds elections soon! By growing reach now, such an AI network will then be able to significantly steer what people see on Bsky! See #romania.

This is my life story. I'm not particularly saying you should read it, but if you're bored.
(cw suicide)
threadreaderapp.com/thread/17775...

I don't even use their OS and still I have to be aware of their maze to install free software!

My puppet monkeys want to know what's wrong with being a talking giraffe?

the way it works is that your terminal emulator sends escape codes (like "ctrl-[[M#x>ctrl-[[M }D") for various mouse actions and then the application has to explicitly implement support for interpreting those codes.

it's not so different from how the mouse works in a GUI in a way

Whoa, thanks for the pointer! I tried to read up on it and learned that executing `printf '\033[?1000h'` starts reporting mouse clicks. Then just click around to see what gets sent!

Do you know (or can point to an explanation of) why it works though? Have terminal tools built in explicit support for this or is this some hacky side effect the terminal emulator provides?
I've used it increasingly recently and am still baffled it works at all.

While I agree on the whole "copyright – which doesn't even prevent so-called AI companies' theft – is a stupid stop-gap solution, we need actual financing for artists", I think the focus on "quality", "process" or "merit" is both a losing-strategy and makes questionable assumptions:

I like art that is both provoking/requiring "deep, critical thinking" and is socially/politically engaged. Other people don't.
If artists in either camp could economically get by just fine we wouldn't care so much if genAI was really good at the latter – but the latter is important economically.

But modern readers don’t seem to want to do this labor, instead preferring texts that give them “instant answers,” [...]"
www.smithsonianmag.com/smart-news/c...

But saying "genAI = shit art" helps nobody here.

To close, some anecdotal evidence: in a study people preferred genAI "poetry" because "Understanding poems written by humans requires deep, critical thinking—and that’s a big part of poetry’s appeal, the researchers write in the paper.

...which gets us nowhere. It's somewhere between "both sides ignore each other" and "increasing existing societal divides". There will be popular genAI images that pass as "art" and that will be economically viable in some sense. In our current configuration it will hurt already precarious artists.

Second and most importantly: this argumentation presupposes not only a personal preference but some bigger and applicable standard or measurement of what is "good" (as opposed to "shit"). This leads to the usual divide between an inner-art-world-elitist understanding and the popular understanding.

This applies especially to art – there's such a long history of "you can't possibly make good art with/without X" until it becomes the accepted avant-garde. (Admittedly, I'm extremely doubtful sloppy generated images will become capital-a art, but that's for different reasons.)

I think it's also a losing strategy because unless you can derive from first principles why a technology can't possibly work, you'll always look like a popularly-called luddite (in the derogatory sense, not the historic sense) who'll be shown wrong in the future.

First, focusing on the current limitations of a technology is somewhere between naïve and disingenuous: genAI (or some other algorithmic system) might very well progress on whatever metric you could apply – unless it boils down to "human/brain license". It's gotten a lot better, it will continue.

While I agree on the whole "copyright – which doesn't even prevent so-called AI companies' theft – is a stupid stop-gap solution, we need actual financing for artists", I think the focus on "quality", "process" or "merit" is both a losing-strategy and makes questionable assumptions:

Cory Doctorow, who I consider to be a 'protocol' advocate in some way, doesn't buy Bluesky's 'protocol' approach at all, and thus won't join anytime soon:
pluralistic.net/2024/11/02/u...

Bluesky is currently a platform not a protocol and it's controlled by a single organization. Who is in control?

I'm chair of palast.jetzt / @palast-jetzt.bsky.social , to rebuild the #PalastDerRepublik, and member of thisisfake.team, a #VR collective.

As a side-project I'm building structure.love, an information management tool currently best for bookmarking.