We could probably do better.
he/him or they/them
My desire to use Mastodon is basically at zero now because post-"Claude", hanging out here means inevitably I'm going to have a conversation with someone who tolerates, or even uses, "generative AI". And what's the point of being in a community where that's a risk. Interestingly* there is absolutely no chance of this on Bluesky, because there are artists there
* And oddly, given how "AI"-brained the *admins* there are
- The process is, and its products often are, stupidly & tragically wasteful. Anything that seems elegant (1) is that way only bc it had a large number of generations in which to improve, and (2) typically contains fractal stupidity just under the surface.
- Globally, it produces/enables suffering of a scale & depth that is far beyond anyone's ability to fully comprehend.
- It is not your friend.
- It is nobody's friend.
- It's a huge mistake to even anthropomorphize it.
Things about evolution that people should keep in mind when considering LLM-based development:
Yes, a process that feeds the output of an RNG into a selection function (like a local environment with a fast predator, or a test suite), _can_ cause the formation of a product that kinda-sorta works (for some definition of "works")...
However:
What an inspiring, powerful letter by Emily Tucker, Executive Directory of the Center on Privacy and Technology, Georgetown Law.
“But the great thing is, you don’t have to go along with this, and I urge you not to. You can refuse to use the chatbot. You can tell your professors that you don’t want them to use it or to require you to use it. At a minimum, you can demand that they assess the work of those who actually want to do the work themselves differently from the work of those who want the chatbot to do it for them. You can organize against “AI” requirements in degree programs and against data products, surveillance systems, and automation in all aspects of your university experience. You can create student groups dedicated to the rejection of all these things, and to the imagination of what you would like your education to be like instead.”
“An Open Letter to Georgetown Students, In Response to Recent Announcements by the University about “Generative AI””
RE: https://sauropods.win/@futurebird/116272840734165707
the costs of allowing generative "a.i."-accelerating hardware to continue to exist are too goddamn high
I've linked to the time code where the error happens.
It's possible to tell that the video is AI generated, but something about how it breaks down here is very disturbing. No one in the comments seemed to notice.
I was scrubbing around trying find out what article they asked the AI to summarize to make this.
Anyway here is the creepy moment.
https://youtu.be/8yTO9nQdN7w?list=TLPQMjIwMzIwMjYkZfxbXaBybw&t=1141
Nothing irritates a bully more than when someone else demonstrates the courage or honor that they lack. They will have an immediate need to prove that any such acts were either a sign that the person was duped, or that the sacrifice was motivated by a desire for public acclaim, not any principles or moral values.
To confront true selflessness is terrifying for the bully. Nothing could be more alien. Nothing more clearly calls into question their own self worth.
No! It's virtue signaling!
mcc
Dave Rahardja
myrmepropagandist
Erik Johnson