Tessera
Folgen

Coleen (she/her)

@thatsilvermagpie.bsky.social

Software developer, content writer, photographer - it's almost like I'm multi-passionate!

113 Posts58 Follower22 Folgt
Coleen (she/her)@thatsilvermagpie.bsky.social · last wk.

This text is over 100 years old and does contain views, especially on people, that don't really jive with modern times and I do not endorse. But if you're looking for a perspective on intellectual property, turning society upside down, and how long we've been going round about the same things...

Read it. The Conquest of Bread by Peter Kropotkin.
Coleen (she/her)@thatsilvermagpie.bsky.social · 2w ago

Before anyone says it, yes, individuals should also be punished for getting material through AI. Deterring every step in the process and every avenue is important. No one should fear their pictures are in a training data set and that kind of repetitive victimization. No one should fear deep fakes.

Unsolicited opinions. That AI companies are spending big to defend themselves in court and exempt themselves from accountability rather than fix models so they don't generate deep fakes, especially of kids, is repellent.
Coleen (she/her)@thatsilvermagpie.bsky.social · 2mo ago

Real universal healthcare, paid parental leave, the ability for disabled people to live in safe ways they don't involve having to never have more than 2k in their possession....the list goes on. These things are the literal floor for what comparable gdp nations (and others) offer their citizens.

Unsolicited opinions. Social safety nets should be stronger. Yes, all of them. Yes, right now. That they are not says a lot about what the US values, and it certainly isn't it's citizens.
Coleen (she/her)@thatsilvermagpie.bsky.social · 2mo ago

AI resource needs will hit historically "poor" communities that have been hit before, usually the elderly, the disabled, the ones who can't move, and who are already disadvantaged by policy choice first. It will exaggerate existing cracks. It's already happening. #AI

Who AI Hits First. By now, we’ve all seen the news about power and cooling (usually water) requirements for the centers that run AI. The data needed to make the models “better.” That all comes from somewhere and someone.In the US, more than 50% of large-scale data centers sit in water-stressed areas. The UN estimates global AI will use as much water as 1.3 billion people by 2030. Stanford HAI put out a report about who’s investing in AI, and it reads like who’s being fed the most propaganda about it.Thing is, we already know whose taps will run dry first to feed this and whose electricity prices will go up fastest. That part is already evident. But it won’t stop there, since aquifers don’t stop at state lines and power grids are all connected. We’re all going to bear it, though unevenly.From a practical standpoint, stopping genAI development at this stage, or AI development in general, is challenging. But we can push to be more efficient, focus on fewer things, not be a mass surveillance tool, etc. This is a we thing though, and that means actual work.