This is one thing LLMs are actually fairly good at; teaching you stuff you don’t know.

While AI’s are much worse teachers than humans are, they’re far better than trying to figure out complicated things on your own.

I’ve gone through many iterations of “Hey LLM, teach me how to do XYZ.” Then I try it. If it works, I ask for more, if it doesn’t I tell it to fix it.

That’s not always a viable path (eg don’t try to learn medicine or stunt driving that way) but when mistakes are safe and cheap, it’s great.

That would be the Mercury Mistress.

https://www.reddit.com/r/LiveFromNewYork/comments/1ql7ur6/the_mercury_mistress_1998/

Obviously NSFW. Pardon the reddit link. I wasn’t able to find a better one.

It takes work to turn them into useful search engines. Set up librewolf. Drop in at least a search and a fetch agent. Wrap it in a some prompts to make sure you get actual links.

On my extremely non-scientific scale I’d rate that setup somewhere between barebones Kagi and pre-enshitified Google in terms of results.

I’m hopeful that an actual agent framework will do a better job that’s significantly more work. If you just log onto a SaaS LLM you’ll get very unpredictable search results.

It would be pretty messed up if the punishment came before the crime.

My point, in this case is that the crime has been ongoing. Bullying and harassment aren’t new. What’s new is that we now have a lot of it recorded. This gives us an unprecedented ability to prosecute a class of crimes that people have been committing with impunity.

The biggest downside is that it would likely be short-lived. If there were a bunch of headlines about people getting giant civil judgements, bullies would probably learn to be more careful about generating evidence of their crimes.

That would likely initially focus on attempting to hide the videos from the general public but I suspect they would eventually learn that doesn’t work; courts can compel discovery. They would probably also try to modify their on-screen behavior so it doesn’t technically qualify as a crime but I expect they would mostly fail at that.

That said there would likely always be some cohort of dumb criminals who continue to record their misdeeds. We still see cases where people discuss their crimes over unencrypted email and we still see cases of criminals taking selfies with their crimes

Can someone explain why this isn’t actually a solution to an old problem? We used to have issues with various forms of harassment that were not addressed because they relied on unreliable testimony. Now we actually have people creating hard evidence of their actions.

The behaviors described in this article are banned in every school I or my children have ever attended. How difficult would it be for the parents of any of these victims to hire a lawyer to absolutely wipe the floor with these people? Don’t these videos basically put the perpetrators heads on silver platters?

That’s not what my comment says. Give it an other read and you’ll see I addressed that.

The overall technology of EV is better than ICE but that doesn’t mean every EV is better than every ICE.

Cars may be better than horses but no horse ever exploded the way a Ford Pinto would.

I just headed over to .ml and did a local search for “mamdani”

I see a bunch of the more conservatives Democrats and their supporters complaining about him and conservatives of all stripes are really hoping he’ll fail.

I’ve been called a tankie several times and I rather like Mamdani. I see him, AOC, and Sanders as the best hope for the future of the US.

Where are these tankies that hate Mamdani so much? What do they hate him for?