|
Perth Comic Arts Festival (a tiny comic convention in Australia) just announced a total ban on AI art across their entire event https://pcaf.org.au/ai-statement/
|
# ? May 8, 2024 09:01 |
|
|
# ? Jun 3, 2024 09:34 |
|
Snowglobe of Doom posted:Well this is loving cursed "allergies yeah" comedy gold also weird ai yankovich was right there
|
# ? May 8, 2024 09:39 |
|
Snowglobe of Doom posted:Well this is loving cursed everything about that is awful
|
# ? May 8, 2024 10:42 |
|
Saint Isaias Boner posted:"allergies yeah" comedy gold https://bsky.app/profile/alyankovic.bsky.social/post/3kqgpkkndhn2g
|
# ? May 8, 2024 11:06 |
|
*AI does something incredible and almost magical* Performative Online Moron: "B-b-but what about LE JOBS aka my lovely furry commissions????"
|
# ? May 8, 2024 15:20 |
|
Spidder posted:*AI does something incredible and almost magical* Oh my goodness, can you please educate everyone (in detail) how amazing and magical it is? I'm sure it will be an amazing read.
|
# ? May 8, 2024 15:29 |
|
If you're dumb enough to engage with a post that bad as though there's a real sentient person behind it AI probably really is gonna blow your mind
|
# ? May 8, 2024 15:35 |
|
A Wizard of Goatse posted:If you're dumb enough to engage with a post that bad as though there's a real sentient person behind it AI probably really is gonna blow your mind I want to hear details come on I'm starving for content, this thread is so one sided.
|
# ? May 8, 2024 15:38 |
|
He kinda owned you, guys.
|
# ? May 8, 2024 15:40 |
(USER WAS PUT ON PROBATION FOR THIS POST)
Somebody fucked around with this message at 15:53 on May 8, 2024 |
|
# ? May 8, 2024 15:51 |
|
New Randy Travis song out...
|
# ? May 8, 2024 16:26 |
|
The Management posted:LLMs are *incapable* of reasoning. They regurgitate words based on a statistical model of language. That means that they can generate the right words that one would expect in a response, but they can’t provide the actual solution unless they’ve been trained on it. Oh, totally. Thing is this screw-up does not happen in general, it usually does its job when I played around with it. Getting it to count wrong and sort wrong took spaces and the ; as a separator. I'd guess the ";" sometimes pushes it in a programmy direction and then spaces count as letters? So your data looks fine forever, everything is correct and then there's the tiniest change in the data and suddenly it is sometimes wrong. And people can understand this is bad, they just need to be told about it. I wrote an email to my bank because they sent me some AI-Backpfeifengesicht in their newsletter and I always advise people to complain if any company that's important to their life starts using AI. If we all work together we can kill it for sure. lol
|
# ? May 12, 2024 14:55 |
|
AI is really BAD at tasks like that. I once generated a list of random numbers and asked chat gpt to tell me the highest number that appeared in the list. Not only would it get it wrong, but would choose different numbers when asked the same question again. It also can't do "what numbers are missing from this series" or "what is the next number in this pattern" questions well.
|
# ? May 12, 2024 17:30 |
|
https://youtu.be/Bvgt9y_o2tw?si=jucjrWsejhTPU8zS
|
# ? May 12, 2024 17:48 |
|
funny how in just two years the public perception of "AI" has gone from something painfully literate and pedantically exact like Data in Star Trek, to just mean some plausible sounding bullshit or messed up anatomy somehow no sci-fi writers predicted this
|
# ? May 12, 2024 17:49 |
|
RabbitWizard posted:Oh, totally. Thing is this screw-up does not happen in general, it usually does its job when I played around with it. Getting it to count wrong and sort wrong took spaces and the ; as a separator. I'd guess the ";" sometimes pushes it in a programmy direction and then spaces count as letters? So, what's interesting is that yes, LLM's have a bitch of a time doing these kinds of problems. However, if you enter these kinds of problems into the actual OpenAI chat window and ask it to solve it it will write code with the data in it and then execute a sorting function or whatever the problem calls for and it generally solves it correctly.
|
# ? May 12, 2024 19:47 |
|
Tarkus posted:So, what's interesting is that yes, LLM's have a bitch of a time doing these kinds of problems. However, if you enter these kinds of problems into the actual OpenAI chat window and ask it to solve it it will write code with the data in it and then execute a sorting function or whatever the problem calls for and it generally solves it correctly. maybe companies that use AI shouldn't be allowed to be limited liability anymore. When a computer runs everything who is left to take on that liability gently caress. I thought you said LLC not LLM, I don't actually know what LLM is
|
# ? May 12, 2024 22:05 |
|
Large Lady Models
|
# ? May 12, 2024 22:09 |
|
Post pics of LLMs here.
|
# ? May 12, 2024 22:10 |
|
Jordan Peterson say that Large Lady Models don't get him hard so they shouldn't be in magazines
|
# ? May 12, 2024 23:35 |
|
Tarkus posted:Jordan Peterson say that Large Lady Models don't get him hard so they shouldn't be in magazines he doesn't get me hard, so he shouldn't be in magazines
|
# ? May 12, 2024 23:47 |
|
Clean your room. Or get addicted to benzos
|
# ? May 13, 2024 01:25 |
|
Houle posted:Clean your room. Benzos? Remember when quaaludes was a thing. I don't
|
# ? May 13, 2024 01:31 |
|
The thing is that the programs aren't 'reading' what you type. They're selecting keywords in combination and outputting what kind of text is most commonly associated with those keywords. Hence why it gives stock answers to those questions without seeming to be able to notice the key differences, because it can't.
|
# ? May 13, 2024 05:15 |
|
I found a compelling use for copilot: When it pops up unwanted in your windows install simply ask it for the instructions on how to disable copilot in the registry and follow the instructions.
|
# ? May 13, 2024 08:14 |
|
personally i like that selena gomez has tried to sell me timeshares and enroll me in a le creuset cookware giveaway in the last 24 hours, and that machine learning will ultimately be a cool breeze of trustworthiness across our scorched media landscape
|
# ? May 13, 2024 09:30 |
|
the heebie-gbs posted:personally i like that selena gomez has tried to sell me timeshares and enroll me in a le creuset cookware giveaway in the last 24 hours, and that machine learning will ultimately be a cool breeze of trustworthiness across our scorched media landscape her makeup range is pretty decent assuming an ai didn't sell it to me
|
# ? May 13, 2024 10:24 |
|
ReelBigLizard posted:I found a compelling use for copilot:
|
# ? May 13, 2024 11:18 |
|
Snowglobe of Doom posted:Perth Comic Arts Festival (a tiny comic convention in Australia) just announced a total ban on AI art across their entire event I can't wait for public sentiment to swing this way until all creative events make the same ban
|
# ? May 13, 2024 12:32 |
|
RabbitWizard posted:The first thing I thought would happen with AI is killing specific kinds of spam mails. Check for words like "SEO", "lottery", "inheritance" or currency values above 12345$, if found write a ton of stuff back relevant to the mail. But this isn't making money I guess? I think you'll see implementations of this fairly soon. I was playing around a few months ago with sifting through my email with a small LLM. It was slow because my computer sucked but it caught the vast majority of the spam that wasn't caught already. Oddly enough, as far as I can tell, spam hasn't gotten much better over the last year. Not as much as I thought it would have. Then again I don't see as much as say some it guy would.
|
# ? May 13, 2024 14:25 |
(USER WAS PUT ON PROBATION FOR THIS POST) Somebody fucked around with this message at 14:50 on May 13, 2024 |
|
# ? May 13, 2024 14:49 |
|
051324_2 posted:(USER WAS PUT ON PROBATION FOR THIS POST) I agree
|
# ? May 13, 2024 14:52 |
|
He actually said something really poignant and funny probably.
|
# ? May 13, 2024 15:08 |
|
RabbitWizard posted:The first thing I thought would happen with AI is killing specific kinds of spam mails. Check for words like "SEO", "lottery", "inheritance" or currency values above 12345$, if found write a ton of stuff back relevant to the mail. But this isn't making money I guess? the money would be in using AI to spam mail even harder unfortunately
|
# ? May 13, 2024 15:43 |
|
kntfkr posted:He actually said something really poignant and funny probably. My understanding is sometimes it is literally just the n word. These machines have a long way to go before they learn to play nice!
|
# ? May 13, 2024 16:47 |
|
ReelBigLizard posted:I found a compelling use for copilot: Hey robot, pull your own guts out while I watch and clap.
|
# ? May 13, 2024 16:57 |
|
naem posted:the money would be in using AI to spam mail even harder unfortunately I watched a video that reasoned if 10% of all electricity use on Earth goes to running the internet, and 50% of the internet is spam bots, then 5% of all electricity use goes to bot junk. And AI's going to up that dramatically. So dead internet theory is just a little early! It's going to end up being mostly AI screaming ads at each other and we'll be reserving a significant portion of Earth's resources for that purpose!
|
# ? May 13, 2024 17:49 |
|
Das Boo posted:I watched a video that reasoned if 10% of all electricity use on Earth goes to running the internet, and 50% of the internet is spam bots, then 5% of all electricity use goes to bot junk. And AI's going to up that dramatically. that number sounds high, but i dunno, maybe. when i worked for a company that designed and built grinding mills (like a front loading dryer that tumbles rock and ball bearings), i was told an interesting stat: somewhere between 2-5% of the worlds total energy consumption (not just electricity, but total energy, including burning fossil fuels) is used in mining new rock. around half of that is used in just one process at the mine: grinding or comminution in the mills. So something like 1-2.5% of the worlds total energy consumption went strictly to grinding rock. this was a lesson in why incremental improvements in grinding efficiency yielded huge returns not hard to imagine when you see a mill like this: and know its filled with hundreds of tons of 3-6" ball bearings and huge chunks of rock anyway. just reminded of that with that claim of the internet using 10% of the worlds electricity (again, not doubting that necessarily)
|
# ? May 13, 2024 18:02 |
|
hot cocoa on the couch posted:that number sounds high, but i dunno, maybe. when i worked for a company that designed and built grinding mills (like a front loading dryer that tumbles rock and ball bearings), i was told an interesting stat: somewhere between 2-5% of the worlds total energy consumption (not just electricity, but total energy, including burning fossil fuels) is used in mining new rock. around half of that is used in just one process at the mine: grinding or comminution in the mills. So something like 1-2.5% of the worlds total energy consumption went strictly to grinding rock. this was a lesson in why incremental improvements in grinding efficiency yielded huge returns So if they turn it on when they are standing there, they have a very bad day right? The 2nd is inside the 1st picture and that spins?
|
# ? May 13, 2024 18:10 |
|
|
# ? Jun 3, 2024 09:34 |
|
lol. yes.
|
# ? May 13, 2024 18:11 |