News:

That's okay, I know how to turn my washing machine into a centrifuge if need be.

Main Menu

Open Bar: Curbside Pickup Only

Started by Junkenstein, July 09, 2020, 06:38:37 PM

Previous topic - Next topic

Suu

Quote from: Cain on January 10, 2023, 10:37:57 AM
Meanwhile I'm using the AI to shitpost with my RP group:



The redundancy annoys me.
Sovereign Episkopos-Princess Kaousuu; Esq., Battle Nun, Bene Gesserit.
Our Lady of Perpetual Confusion; 1st Church of Discordia

"Add a dab of lavender to milk, leave town with an orange, and pretend you're laughing at it."

altered

Quote from: Suu (parody account) on January 13, 2023, 12:41:42 AM
Quote from: altered on January 09, 2023, 12:03:21 AM
Funny enough, I think academia is going to be the first place to get well and TRULY fucked by AI for the combination of 4 important aspects:

1: Academic papers are usually kind of samey (with good reason), which makes writing them very automatable.
2: Feeding source material and new research in and getting a paper out is something most academics who are worth a damn think they would enjoy doing, because it'd reduce the time spent on writing papers and lower the barrier to entry for getting helpers on board.
3: The bleeding edge of AI is largely coming out of academia, so will likely have a lot of access to academic source material.
4: AI models that have high contextual awareness are the norm now, which means you just have to turn the temperature down real low and have it focus on turning data into words with zero creativity.

We already have shit like computer aided chemical synthesis design and automated theorem provers out there literally churning out almost wholly computer generated papers, and have had them for years now. Academia is already primed for this sort of thing and people would hardly notice at first, even in the "hard sciences". It might even have good side effects. Then again, it'll probably concentrate institutional power even more with good politicians who can't do science worth a fuck.

We already have software to catch them, and honestly, some of them are very, very "off" when you read them. It's a regurgitation of facts but with no argument. It's fucking WEIRD. That, and they cite incorrect 99% of the time. I know all of this will be fixed eventually and I'm out of a job to a bunch of Zoomers who speak entire sentences with one word, but it's not making the waves people thought it was going to.

Also, I'm in the humanities, so our papers are generally more flowery and have a detectable voice. I'm sure dry STEM professors with no sense of critical thinking skills will think their students are geniuses or some shit reading this garbage.

Oh, I was definitely thinking STEM there.
"I am that worst of all type of criminal...I cannot bring myself to do what you tell me, because you told me."

There's over 100 of us in this meat-suit. You'd think it runs like a ship, but it's more like a hundred and ten angry ghosts having an old-school QuakeWorld tournament, three people desperately trying to make sure the gamers don't go hungry or soil themselves, and the Facilities manager weeping in the corner as the garbage piles high.

Suu

Quote from: altered on January 13, 2023, 09:12:26 AM
Quote from: Suu (parody account) on January 13, 2023, 12:41:42 AM
Quote from: altered on January 09, 2023, 12:03:21 AM
Funny enough, I think academia is going to be the first place to get well and TRULY fucked by AI for the combination of 4 important aspects:

1: Academic papers are usually kind of samey (with good reason), which makes writing them very automatable.
2: Feeding source material and new research in and getting a paper out is something most academics who are worth a damn think they would enjoy doing, because it'd reduce the time spent on writing papers and lower the barrier to entry for getting helpers on board.
3: The bleeding edge of AI is largely coming out of academia, so will likely have a lot of access to academic source material.
4: AI models that have high contextual awareness are the norm now, which means you just have to turn the temperature down real low and have it focus on turning data into words with zero creativity.

We already have shit like computer aided chemical synthesis design and automated theorem provers out there literally churning out almost wholly computer generated papers, and have had them for years now. Academia is already primed for this sort of thing and people would hardly notice at first, even in the "hard sciences". It might even have good side effects. Then again, it'll probably concentrate institutional power even more with good politicians who can't do science worth a fuck.

We already have software to catch them, and honestly, some of them are very, very "off" when you read them. It's a regurgitation of facts but with no argument. It's fucking WEIRD. That, and they cite incorrect 99% of the time. I know all of this will be fixed eventually and I'm out of a job to a bunch of Zoomers who speak entire sentences with one word, but it's not making the waves people thought it was going to.

Also, I'm in the humanities, so our papers are generally more flowery and have a detectable voice. I'm sure dry STEM professors with no sense of critical thinking skills will think their students are geniuses or some shit reading this garbage.

Oh, I was definitely thinking STEM there.

Get ready for rocket scientists that drank their way through grad school getting jobs at NASA and SpaceX.
Sovereign Episkopos-Princess Kaousuu; Esq., Battle Nun, Bene Gesserit.
Our Lady of Perpetual Confusion; 1st Church of Discordia

"Add a dab of lavender to milk, leave town with an orange, and pretend you're laughing at it."

Scribbly

Quote from: Scribbly on January 12, 2023, 06:43:35 PM
I think I've just been banned from the local 'find roleplay groups' facebook group for ... posting an ad for my game?

Currently pondering what level of revenge to take if this turns out to be correct.


ETA: It has now been confirmed.

Apparently I am not well-known enough as part of the community to be advertising my games, and the person who confirmed this got some static from them for doing so.

Excellent.

I reached out on Twitter to ask for an explanation and woke up to a 5 tweet long rant about how entitled and arrogant I am. Apparently the gentleman in question had gotten the idea I was advertising a paid-for product? I'm not sure how. Trying to explain that no, you weirdo, I'm just using the group like everyone else, just seemed to make him angrier. He also repeated like three times that he built this 500 person facebook group from NOTHING so I should be showing him RESPECT for that achievement.

Eventually I blocked him but this has preyed on my mind most of the day at this point.

I'm going to have to think of something elaborate.
I had an existential crisis and all I got was this stupid gender.

chaotic neutral observer

I am beginning to perceive a pattern, where people outside the STEM fields see AI as this revolutionary technology that is going to upend everything, while the people inside the STEM fields have an attitude more akin to "here's another tool for the toolbox; it will improve your productivity in some ways, and occasionally give you a terrible headache."

There was a similar dichotomy of attitudes when electronic microcomputers were becoming popular, but we didn't get a utopian paradise, nor did it take away everyone's jobs, or make us slaves to the machines.  Some jobs went away, some were created, some changed; but overall, we're still in perdition, just with some aspects running orders of magnitude faster.

I don't write academic papers, so automating that process does nothing for me.  However, I do occasionally read papers from electronic engineering academia, and in that context, my interest is whether or not the paper contains information I need to solve a problem.  The form of the paper affects its readability, but the underlying research is what determines its utility.

In its current incarnation, OpenAI is obviously and painfully incapable of producing academic papers of any value to me (see my travelling salesman example, above).  An AI that could synthesize existing papers together to produce novel conclusions might have use, but the results would have to have practical relevance.  There are an infinite number of true yet meaningless papers that it could construct, which means that someone who knew what they were doing would have to be presenting a problem space to perform the synthesis over, or sifting through the mud to find the gold nuggets.

An AI that could, on its own, determine what kinds of papers someone in industry would find useful, and then synthesize those (effectively, an autonomous research engine) would be verging into singularity territory, and that would be pretty cool.

Quote from: Suu (parody account) on January 13, 2023, 01:30:28 PM
Get ready for rocket scientists that drank their way through grad school getting jobs at NASA and SpaceX.

I thought all rocket scientists drank their way through grad school?

But in practical terms, AI isn't going to provide a vector for students to cheat their way to a degree, because practically everyone in the domain will know what the tools are capable of.
There will be periods of transition, to be sure, but the assignments will just change form.  Instead of "perform research X, and write a paper on it", it becomes "use RocketAI to model research X, and generate a paper from your results using PaperWriter-2.1".
Desine fata deum flecti sperare precando.

altered

The issue with AI in STEM is that, like I said, we already have near-automated paper factories in the form of total synthesis design and automated theorem provers. They're junk that does occasionally produce good stuff, but (for instance) a lot of total synthesis papers can't be replicated because it's all computer generated. Good STEM papers are pretty interchangeable as far as surface-level stuff goes, they translate readily between languages because they're practically standardized to hell and gone. I think when someone trains a model to receive input data and spit out a passable paper, they could lie about it and no one would even know.

I don't think it's going to work for students, like CNO said. I don't think third-rate students are going to be getting doctorates out of a future academics AI. I think you have to have someone who knows how to DO research, and present data. The people who will get doctorates out of it would have gotten them anyway. Either they can do the work and just automated the part they don't care about, or they were going to get nepotized in or buy their way to their scholarly ambitions already, and this was one way to make it less obvious and embarassing for everyone involved.

What I mean is that all of the WRITING part of putting out articles is not long for this world. And I think that's, to abuse some dead people, going to do more for a Thomas Edison than a Nikola Tesla, which isn't a good thing in my mind. It might incidentally have some good side effects, lots of the stuff with AI right now is left-field "wow, what the fuck?" surprises after all, but following the concepts as they exist to fruition looks grim.
"I am that worst of all type of criminal...I cannot bring myself to do what you tell me, because you told me."

There's over 100 of us in this meat-suit. You'd think it runs like a ship, but it's more like a hundred and ten angry ghosts having an old-school QuakeWorld tournament, three people desperately trying to make sure the gamers don't go hungry or soil themselves, and the Facilities manager weeping in the corner as the garbage piles high.

Cain

Quote from: Suu (parody account) on January 13, 2023, 12:46:03 AM
The redundancy annoys me.

Yeah, there's definitely a few "tells" that it is generated.

That said, I've found it a useful tool for quickly generating NPCs or even small plot ideas. On it's own it came up with the idea of a hidden Ayleid city (?) in the heart of the Clockwork City (??) which the factotums need to hire the PCs to get into because the defences prevent them from being able to breach it (???) because the Worm Cult (????) were trying to steal an artifact from it first.

Which is batshit, but you can definitely work with it. Correcting it and re-defining parameters and exclusions make it a great brainstorming tool.

I also made it give me a summary for an entirely made up 1890s period "techno-thriller" involving sabotage of the telegram system, which was fun.

Suu

Quote from: Cain on January 15, 2023, 06:02:47 AM
Quote from: Suu (parody account) on January 13, 2023, 12:46:03 AM
The redundancy annoys me.

Yeah, there's definitely a few "tells" that it is generated.

That said, I've found it a useful tool for quickly generating NPCs or even small plot ideas. On it's own it came up with the idea of a hidden Ayleid city (?) in the heart of the Clockwork City (??) which the factotums need to hire the PCs to get into because the defences prevent them from being able to breach it (???) because the Worm Cult (????) were trying to steal an artifact from it first.

Which is batshit, but you can definitely work with it. Correcting it and re-defining parameters and exclusions make it a great brainstorming tool.

I also made it give me a summary for an entirely made up 1890s period "techno-thriller" involving sabotage of the telegram system, which was fun.

It may also turn out to be a worthwhile tool to help overcome writer's block. You stuck but have some ideas? Throw in some prompts and see what comes up.
Sovereign Episkopos-Princess Kaousuu; Esq., Battle Nun, Bene Gesserit.
Our Lady of Perpetual Confusion; 1st Church of Discordia

"Add a dab of lavender to milk, leave town with an orange, and pretend you're laughing at it."

Cain


chaotic neutral observer

So, word is they're now planning on sending tanks to Ukraine, and I says to myself, I says:  "Why not earlier?"  They've been doing the same kind of gradual ramping up in other ways, as well; HIMARS now, Patriots later.
And I can think of a few possible reasons:

1. Bureaucracy is slow.
2. They're worried that an an abrupt escalation will spook Pootin into going nuclear.
3. The object of the Powers that Be is not to break Russia quickly and abruptly with overwhelming force, but to bleed them out slowly, causing them much more damage in the long run.
4.  ...and on the opposite side of that coin, Ukrainians are the "good guys" right now, but maybe they don't want to assume that will continue indefinitely into the future.  Flooding Ukraine with modern weaponry has the risk of creating a new regional mini-superpower, and if we provide the weapons gradually enough, the accumulating damage to Ukraine's infrastructure will keep a lid on their future ambitions.

I tend to prefer the "Bureaucracy is slow" reason, because it's the simplest.
Desine fata deum flecti sperare precando.

altered

Bureaucracy is indeed slow.

If anything from 2-4 is true, I'd put all my money on 4 just because of the lingering influence of that pigfucker Kissinger. 3 is strategically sound but doesn't build much support (and so is a bad move from a political standpoint), and 2 is more of a civilian-brain take on the situation (functionally, Putin is at the point where if he would ever press that button, when he does isn't really affected by NATO aid because all his shiny NBC warfare-equipped tanks are smoldering dumpsters on the Ukrainian border -- Russia can't fight very well in NBC conditions now, and NATO still can, for whatever that's worth).

But Kissinger put so much rhetorical energy into this two-superpowers idea. It's why everyone was/is talking scared about China forever, and why Japan was demonized in the 90s for being the technology hub of the globe, and why US policy in the Middle East is to repeatedly fuck every dog in sight. If there's any reason other than bureaucracy, its almost certainly that Kissinger's stink still has the frontal lobes of the US military entranced.
"I am that worst of all type of criminal...I cannot bring myself to do what you tell me, because you told me."

There's over 100 of us in this meat-suit. You'd think it runs like a ship, but it's more like a hundred and ten angry ghosts having an old-school QuakeWorld tournament, three people desperately trying to make sure the gamers don't go hungry or soil themselves, and the Facilities manager weeping in the corner as the garbage piles high.

Cain

The public statements are two, but I suspect 3 with an eye to 4 are also strong considerations. There's some...questionable sentiments on the eastern fringes of the EU (Hungary and Poland stand out) and a highly militarised Ukraine would not help those situations any.

Suu

And then there are days wherein after I break a toe and probably a metatarsal with it, I just go get a random tattoo out of a vending machine. The end.
Sovereign Episkopos-Princess Kaousuu; Esq., Battle Nun, Bene Gesserit.
Our Lady of Perpetual Confusion; 1st Church of Discordia

"Add a dab of lavender to milk, leave town with an orange, and pretend you're laughing at it."

Scribbly

Not gonna lie, these past couple days have been brutal.

Basically everyone I know is now telling me to flee the country and as much as it goes against my nature I am having to give it real thought. This is not doing great things to my mental health.
I had an existential crisis and all I got was this stupid gender.

altered

Be careful about choosing where you go, if you do. There's a lot of places that are a bad day from as bad off as or worse than the UK.
"I am that worst of all type of criminal...I cannot bring myself to do what you tell me, because you told me."

There's over 100 of us in this meat-suit. You'd think it runs like a ship, but it's more like a hundred and ten angry ghosts having an old-school QuakeWorld tournament, three people desperately trying to make sure the gamers don't go hungry or soil themselves, and the Facilities manager weeping in the corner as the garbage piles high.