Schadrach

We have healthcare, the best healthcare actually

We actually do. We have some of the best healthcare in the world if you can afford it, and healthcare stats that make it absolutely clear just how few people that is.

Robot butler - robot waiters already exist, so it wouldn’t be terribly difficult to repurpose one (although they’re only sold to businesses as far as I could tell)

There’s an Asian AI lab that’s demoed an early version of an AI-driven humanoid robot domestic servant. There are suggestions in might hit market within a few years and cost about as much as a decent used car. Figure those estimates are always too optimistic and something like 2035 and $15k is a possibly realistic estimate assuming nothing goes catastrophically wrong.

No it generally makes sense to teach kids to not cite Wikipedia. Though it is consistently checked and updated you can look at the wiki link and drama for the Israeli genocide just to see a perfect example of why it shouldn’t be cited.

Wikipedia is generally terrible for anything that was politically controversial since Wikipedia has been a thing. A lot of why is very intentionally buried in layers of bureaucracy and wikilawyering to make it look like totally reasonable, neutral point of view decision making. One of the big routes to viewpoint control on Wikipedia is arguments about notoriety and what is or is not a “reliable source” and what sources are sufficient to discuss a topic.

Given the Lemmy view on AI, I wonder how many folks are now uninstalling the game and demanding a refund because it’s suddenly transformed into “AI slop”? Or demanding it be delisted from Steam since they didn’t disclose their use of AI on Steam?

To quote UNSONG Chapter 17: No Earthly Parents I Confess (https://unsongbook.com/chapter-18-no-earthly-parents-i-confess/ and yes it’s chapter 17 despite the URL, and I’m sure there’s something significant about that but I am unsure exactly what offhand, but everything in UNSONG is significant):

"Picture a maiden lost in the hills.

“Maiden” can mean either “young woman” or “virgin”. Its Greek and Hebrew equivalents have the same ambiguity, which is why some people think the person we call the Virgin Mary was actually supposed to be the Young Woman Mary – which might change the significance of her subsequent pregnancy a bit. People grew up faster, back in the days when they spoke of “maidens”. Mary was probably only fourteen when she gave birth.

I am a kabbalist. Words matter. Nowadays we have replaced “maiden” with “teenage girl”. A maiden and a teenager are the same thing, but their names drag different tracks through lexical space, stir up different waters. Synonymity aside, some young women are maidens and others are teenagers. The girl in our story was definitely a maiden, even though it was the 1970s and being a maiden was somewhat out of fashion."

I remember an internet joke site decades ago doing one that was sports cars or ED drugs. Don’t recall the name of the site though. Quick search didn’t turn it up.

Depends on model tuning. Basically, you can tune a model to hallucinate less, or to write more human-like, but not really both at the same time, at least not for a model you could expect most users to run locally. For this sort of application (summarizing text), you’d tune heavily against hallucination, because ideally your bullet points are going to mostly be made up of direct paraphrase of article text, with a very limited need for fluid writing or anything even vaguely creative.

Doesn’t do that for me. I have to hold left click on a link for over a second to trigger it.

And yeah, pretty decent. It can produce a basically summary of a fair amount of text pretty quickly and generally accurately. It’s not an expert wordsmith, it won’t give a deep and thoughtful analysis of the poem you pointed it at or anything, but that’s not the use case. The use case is “give me the key bullet points of this article so I can decide if I should give it more attention.”, and it does that job pretty well.

Do people making that argument also find ad blockers even ten percent as horrible as this? They both ultimately have the same effect, which is your web browser not maximizing someone else’s profits by denying them a revenue opportunity.

I’d be curious if the link summarizer in Firefox runs a model locally or calls some remote API. Most current machines ought to be able to run an appropriate LLM model for that task.

They aren’t against the concept of having pets themselves they just don’t like that pets are specifically bred for domestication when millions of pets are put down in shelters because they couldn’t find homes.

…and yet PETA shelters have higher kill rates than many/most others.

So it’s entirely optional, you only encounter it by interacting with the browser in an atypical way, and the thing it does is a thing that AI is actually pretty decent at (summarizing text)? Sounds like they couldn’t stop themselves from joining the dick hammering bandwagon, but decided not to hit it too hard.

except this can still represent revenue for (some) artists…whilst being able to practice (to a degree) and being paid for it.

…which is just another way of saying that that work should be protected from automation. That’s what arguing that tools that automate doing a thing will make people doing that thing less valuable as a paid labor is doing, it’s arguing for protection from automation.

Also, do you have a source regarding that or is it from what you experience online?

Experience and a bit of hyperbole. Ask anyone who takes art commissions how often they get asked to do lewd to outright pornographic images though and you’ll be surprised by the answer. But commissioned art/graphics for very specific carefully described things is all AI image generation is ever really going to replace, if only because that’s the core of what it does - take a prompt and a big block of white noise and sort of digitally chisel away the bits that don’t look like prompt until the result looks enough like prompt. The other obvious productive use case would be for rapid prototyping of visual design.

But then I’m old enough to remember people complaining that photoshop was destroying art. I’d be shocked if we couldn’t find record of people back when it was new claiming photography was going to destroy art likewise.

Fuck the creative process, the journey it takes you on, and the necessary introspection and connection to the world that needs to occur for it. Fuck the joy you can find in effort, failure, and in finally having an epiphany. Fuck being able to hone your skills without depending on a corporate tool that can be taken away from you at a minute’s notice. All of it be damned, you can now (allegedly) get to the same end result quicker without the effort (or pleasure, or self-discovery, or personal growth). We all know that the end result is what counts, and nothing else.

Is anyone claiming you can’t do that? The use case for most AI slop is soulless corporate graphic design that’s about as worker bee as it gets and the rest is mostly people using it in place of hiring an actual artist for their weirdly specific niche pornography. And the guy wanting you to draw him something involving dolphins with a foot fetish isn’t deeply concerned what your personal journey regarding the nature of creatures without feet who are obsessed with them sexually looks like.

At it’s heart, the non-environmentalist AI hate is mostly people who thought their jobs were impossible to automate trying to protect those jobs from being automated.

Google gives you an option as to how autonomous you want it to be. There is an option to essentially let it do what it wants, there are settings for various degrees of making it get your approval first.

Unironically this. I’ve only really tried it once, used it mostly because I didn’t know what libraries were out there for one specific thing I needed or how to use them and it gave me a list of such libraries and code where that bit was absolutely spot on that I could integrate into the rest easily.

It’s code was a better example of the APIs in action and the differences in how those APIs behave than I would have expected.

I definitely wouldn’t run it on the “can run terminal commands without direct user authorization” though, at least not outside a VM created just for that purpose.

…and he very, very carefully threaded the needle and chose his words to avoid perjury. It’s why he had questions like asking them to define “sexual relations” (the definition they gave didn’t include oral, so he did not have sexual relations with her by their definition) and what the definition of “is” was (specifically does it mean currently or does it include at any time in the past).

Detect Magic telling someone “it’s chowder” is a cop-out, same as a DM saying “you failed the skill check because you looked suspicious.” If a spell exists to reveal a magical aura, use it to reveal an aura, not to sass the player.

My answer in that case is “You detect no aura” from the non-magical chowder (or maybe they do detect one if it was flavored with prestidigitation), unless it’s an edition where the effect is a cone, and they are sitting across the table from their friend blinged out in magical gear, in which case they are definitely detecting an aura. Several of them. And they’re going to have to take time, focus, and make checks to recognize that none are coming from the chowder.

Paranoia, the game where every character is technically engaged in a crime punishable by death at basically all times, and you’re given a number of clones because you are expected to die…a lot. Also the R&D gadgets, like the personal disintegrator which does exactly what it says on the tin - disintegrates your person.

Like forcing the players into an encounter where all their toolkits are nerfed. Close quarters for casters, magical monsters that can’t be harmed by melee, or NPCs that are way OP for the group and they stick to the Monster Manual to the letter.

When I GM, it depends on just how narrow and just how powerful your particular toolkit is. I’m not going to ensure that you can do whatever your thing is at absolutely every opportunity, and if your schtick becomes well known, enemies capable of planning will plan around it when feasible. The more narrow your schtick is, the more scenarios you might encounter where it does not apply simply by chance (for example, if you’re a flying archer every room in a dungeon won’t gain a minimum 30’ high ceiling to maximize your use of that). The more disproportionately powerful your schtick is compared to other party members, the more likely I am to specifically come up with occasional scenarios meant to make it not apply so someone else gets to shine.

Sometimes I will signpost something is a very bad idea, and if you do it anyways (or do something else absurdly dangerously foolish) I’m not going to pop up a guard rail to save you at the last moment - retrieving your body from somewhere adrift on the astral and your soul from the gemstone the archdevil you pissed off is keeping in his treasury to try to save you is the next adventure hook.

You encounter a huge, elaborate tome, on a concealed lectern, in a library connected by a hidden door directly off the bedroom of a powerful wizard, you detect magic and get extremely powerful auras of conjuration, transmutation and evocation maybe “I flip it open to a random page and start reading aloud, I’ll sound out any words I don’t recognize” is not, in fact, a wise decision. The copy of “Words You Mispronounce And Die: A Primer For Apprentice Wizards” you saw on one of the shelves on the way there, the references to a cursed grimoire of terrible power, the book being bound in the skin of an angel covered in burns and scars, etc, etc should have maybe hinted at that.

Only way to fight back in the War on Christmas, which like the Palestine and Israel is mostly just an excuse for the alleged victim (Israel) to attack and take territory from their neighbors.

Here Christmas has invaded so far that there are occasional sightings before the equinox.