count_dongulus @ count_dongulus @lemmy.world

Posts

2

Comments

284

Joined

2 yr. ago

Summarization is one of the things LLMs are pretty good at. Same for the other thing where Wikipedia talked about auto-generating the "simple article" variants that are normally managed by hand to dumb down content.

But if they're pushing these tools, they need to be pushed as handy tools for editors to consider leveraging, not forced behavior for end users.

If nonviolence gets you want you want, you don't resort to violence in the first place. Did the author account for this and consider whether resistance categorized as violent began as nonviolent?

Just donate plasma. Works great, gives some other schmuck your plastic.

It's a bit strange to think about, but our brains seem to have adapted to information accessibility today by more readily remembering how to find the information instead of the information itself. (See Betsy Sparrow et al)

If you lived back then, chances are you'd just straight up remember more things without needing to go look them up again. But, you might also just remember what book you found it in.

I have wondered if this is part of the reason why ancient orators were apparently capable of reciting hours of dialog from memory. They simply had to. Libraries and books weren't generally accessible. They had to rely on memory, and thus became very trained on it.

"If you won't let us have Ukraine, let us have the Baltic states instead! 😡😭"

Defense, foreign relations, cross-jurisdiction crime, the usual things. But civil law and local criminal policy overridden locally, if voters desire?

I guess I'm thinking about a situation where let's say one region wants to trade with some other country, and another doesn't like that, then tough luck. Or same sex marriage, vehicle emissions rules, etc. That sort of thing. Seems like in places such as the US, voters from the other side of the country can override what your local citizens want if they get enough other external voters to side with them.

Humans apply judgment, because they have emotion. LLMs do not possess emotion. Mimicking emotion without ever actually having the capability of experiencing it is sociopathy. An LLM would at best apply patterns like a sociopath.

At the end of the day, most internet forums exist as entertainment. It doesn't really matter whether the text is from real people or not, what matters to Reddit and other platforms is that real users show up to pump numbers and get served ads. It's kinda like how pro sports is not latently about athleticism, it's just another entertainment platform to get butts into seats and eyes onscreen to make a buck.

If bot content gets the eyeballs, so what? Guess the content being fictional doesn't really matter. People don't like being lied to, but then nobody is claiming everything you read on the internet is true in the first place.

Anyway, to all the people who imagined humanity's future would end up looking different than WALL-E: sorry to disappoint. You'll get a hover recliner, at least.

Sorry I'm not into bracelets

Oh no, there are only three rebrandings of the same hot dogs from the same hot dog manufacturing plant instead of five 🪦

Pulsar mouse, Keychron keyboard

They're...asking a country about this? What, do they expect China to say "we're preparing to invade Taiwan in 2026"? This is just fishing for bullshit answers so politicians can pretend like they buy it and ignore actually dealing with the military buildup by taking action.

Shot an invasive squirrel with a new mini crossbow someone gifted me instead of my usual compound bow. The squirrel's definitely dead now, but the mini bolt got stuck in it before it ran off onto the neighbor's bigass wisteria bramble. With the compound bow and a blunt head arrow, they die instantly on the spot and I can toss them. I did the math just now, the mini crossbow is an order of magnitude weaker. Oops.

Anyway, gonna get stinky over there in a few days.

Move to an undisclosed remote location and start posting massive crypto bounties on the heads of the shittiest people in the world. Like, $100M a pop. Pay a digital sweatshop to spam social media with AI generated posts and memes about it until the whole world is aware. Then wait. See if anyone is able to collect.

When you're involved in drugs, you are pretty likely to regularly interact with sketchy people who make you feel unsafe for one reason or another. That goes for both dealers and consumers, though moreso on the dealer side.

You can use an inkjet printer. They don't have that secret yellow dot metadata.

I've gone through Visual Studio, JetBrains IDEs, Vim, Atom, Sublime, VSCode, probably others too, but frankly VSCode's simplicity out of the box coupled with great plugin support is hard to beat. Folks who complain about VSCode not having some feature like to ignore that being relatively simple by default is a good thing. You can always add or enable what extensions you need to tailor it to your language and workflow of choice. Even if you're used to Vim keyboard centric editing...guess what? There's a well supported OSS extension to give you that functionality.

The power of being able to use one IDE on a diverse team across various languages is huge. You can even commit extension and settings defaults to a repo to immediately get new cloners up to speed with whatever workflow and tooling defaults are good starting points on a per project basis, but still leaving them the option to ignore/override as needed without dictating a team-wide workflow change.

Well yeah, they have no natural aversion to lying and manipulation.

lol @ dumbasses who didn't look at how their area was zoned before buying a house.

Pound sand, crybabies.

But it legally doesn't. That is why AI has not taken over in high liability fields. Morons are testing the waters and learning that AI mistakes make no difference in a court room, and if anything are grounds for further evidence of negligence.

The big bet now, I think, is whether those popup insurance policies regarding coverage for losses relates to AI usage end up profitable. If so, that is what will lead to truly dystopian stories like "AI piloted passenger jet crashes, United Airlines fined x million dollars but happily continues using AI pilots because insurance covered the fine and it's just a cost of doing business"