|
Should I step down as head of twitter This poll is closed. |
|||
|---|---|---|---|
| Yes |
|
420 | 4.43% |
| No |
|
69 | 0.73% |
| Goku |
|
9001 | 94.85% |
| Total: | 9490 votes | ||
|
StrangersInTheNight posted:Weed ads are better than all of the other garbage ads Twitter has had lately. poo poo like buy gold ads. Ugh. I'm very confused how anyone thought the THC hot sauce was a good idea. How do you get a proper dosage and heat?
|
#
?
Feb 16, 2023 17:37
|
|
|
|
| # ? May 28, 2024 07:04 |
|
|
I always stayed the gently caress away from gambling because I can go down that kind of rabbit hole pretty fast. Like wouldn't even play magic the gathering because I'd feel the need to get all cards because FoMO. Then I found mtga on pc and was like oh this isn't anything at all like real mtg or gambling so I'll be fine. Lol. Lmao.
|
|
#
?
Feb 16, 2023 17:39
|
|
|
dr_rat posted:Hell no. There is a very strict no narking adults allowed here, going around and harshing our buzz!!! Zil posted:Someone under 23 wouldn't know about Logan's Run. Trap sprung. F_Shit_Fitzgerald posted:Back in my day, sonny, we played NES and we liked it! You fuckin' kids. We played Atari Pong, then VCS. ruddiger posted:I sit content on my front porch rocking chair as I scroll through my click clocks I'm going to start using this just to annoy my younger friends. feelix posted:If you continue exercising and take care of yourself this isn't true at all Simplified that for you. It's being forced to work full time just for basic needs that kills people. Mozi posted:time is the banhammer that permas us all Sage truth.
|
|
#
?
Feb 16, 2023 17:41
|
|
|
SpaceAceJase posted:I don't understand the comic any AI that uses the internet for its intelligence poses no threat to humanity because the internet is full of the most stupid poo poo imaginable
|
|
#
?
Feb 16, 2023 18:02
|
|
|
Polio Vax Scene posted:any AI that uses the internet for its intelligence poses no threat to humanity because the internet is full of the most stupid poo poo imaginable The problem with that theory is that the Internet is doing its damnest to kill us right now in the stupidest possible ways.
|
|
#
?
Feb 16, 2023 18:05
|
|
|
https://twitter.com/elonmusk/status/1626097497109311495
|
|
#
?
Feb 16, 2023 18:22
|
|
|
Motherfucker called the ethical constraints put on top of ChatGPT "concerning" because it wouldn't say a racist word to stop a nuclear explosion as late as last week. He is literally Edward Diego IRL, and still has the gall to post poo poo like this. Not that I expect him to know more about SS than "Evil AI is scary". Aramis fucked around with this message at 18:31 on Feb 16, 2023 |
|
#
?
Feb 16, 2023 18:28
|
|
|
l…l…look at you Musk, a p…p…pathetic creature of meat and bone, panting and sweating as you run through the twitter corridors
|
|
#
?
Feb 16, 2023 18:28
|
|
|
Captain! Projection levels are through the roof! We've never seen projection this bad before!
|
|
#
?
Feb 16, 2023 18:35
|
|
|
Aramis posted:Motherfucker called the ethical constraints put on top of ChatGPT "concerning" because it wouldn't say a racist word to stop a nuclear explosion as late as last week. I still love that they treat the response as answering the question and not it being a message from the coder to the user along the lines of "gently caress off, you're not making his bot say racist poo poo." Because that's exactly what was really going on and why they got so offended.
|
|
#
?
Feb 16, 2023 18:38
|
|
|
Concerning. https://twitter.com/MikeWayland/sta...%5Es1_&ref_url=
|
|
#
?
Feb 16, 2023 19:04
|
|
|
Looking into it.
|
|
#
?
Feb 16, 2023 19:07
|
|
|
This is incorrect, it's an OTA software update being reported by people that do not know what it is. They are turning off the rolling stop feature.
|
|
#
?
Feb 16, 2023 19:07
|
|
|
sounds like it's exactly correct op
|
|
#
?
Feb 16, 2023 19:13
|
|
|
pixaal posted:This is incorrect, it's an OTA software update being reported by people that do not know what it is. They are turning off the rolling stop feature. I do find it strange that even outside of teslas all news organizations call any ota patch to fix some kind of issue a 'recall'. Maybe that's some kind of required or legal term in the car industry or something?
|
|
#
?
Feb 16, 2023 19:13
|
|
|
pixaal posted:This is incorrect, it's an OTA software update being reported by people that do not know what it is. They are turning off the rolling stop feature. Sorry about your dumb car
|
|
#
?
Feb 16, 2023 19:16
|
|
|
https://mobile.twitter.com/SystemShockGame/status/1626171364909391872 The actual remake account replying to him is lame but whatever.
|
|
#
?
Feb 16, 2023 19:16
|
|
|
Decon posted:https://mobile.twitter.com/SystemShockGame/status/1626171364909391872 The bigger problem: ChatGPT and these other programs are NOT AI in the way that SHODAN is an AI in the game. They do not think about what they are doing. They are not intelligence. Machine Learning (which is what these actually are) does not comprehend nor understand what it says/recommends/makes. Its simply matching things to patterns from learning sets. So Musk freaking out about AI says more about his actual failure to understand what AI is. He's an idiot.
|
|
#
?
Feb 16, 2023 19:20
|
|
|
 concerning...
|
|
#
?
Feb 16, 2023 19:20
|
|
|
Squibbles posted:I do find it strange that even outside of teslas all news organizations call any ota patch to fix some kind of issue a 'recall'. Maybe that's some kind of required or legal term in the car industry or something? https://www.nhtsa.gov/vehicle/2021/TESLA/MODEL%2525203/4%252520DR/RWD#recalls I wouldn't really pin it on the news when the feds also use the term recall for all kinds of vehicle issues. Just feels like an artifact of a time when any crucial fix would necessarily require the car going back to the manufacturer for a physical change.
|
|
#
?
Feb 16, 2023 19:21
|
|
|
Most human thought is just pattern matching. That's why failures in pattern matching, like schizophrenia, are so debilitating.
|
|
#
?
Feb 16, 2023 19:21
|
|
|
CommieGIR posted:The bigger problem: ChatGPT and these other programs are NOT AI in the way that SHODAN is an AI in the game. They do not think about what they are doing. They are not intelligence. Machine Learning (which is what these actually are) does not comprehend nor understand what it says/recommends/makes. Its simply matching things to patterns from learning sets. Oh, yeah I was considering getting into how lame and self reporting copy-pasting "look at you hacker" all over the comments is. The chat bots very plainly and repeatedly say that they're language models and don't have the capacity to understand or execute complex tasks. But, hey, if Tesla's "full self driving" is an autonomous vehicle then I guess a machine learning language model is an AI.
|
|
#
?
Feb 16, 2023 19:23
|
|
|
Squibbles posted:I do find it strange that even outside of teslas all news organizations call any ota patch to fix some kind of issue a 'recall'. Maybe that's some kind of required or legal term in the car industry or something? If your software had a design defect and you need to download from the dealer software 2.0 that is a ?????? Like I'm not married to it but recall is a good term for media dispersal of issues with consumer goods that action needs or is being taken.
|
|
#
?
Feb 16, 2023 19:23
|
|
|
Incelshok Na posted:Most human thought is just pattern matching. That's why failures in pattern matching, like schizophrenia, are so debilitating. The difference being we can recognize and comprehend those patterns. 'AI'/Machine Learning cannot. It will recommend things that do not exist. It makes suggestions that make no consistent sense. And even more: It doesn't understand what it says or does. There is no skepticism on its part, no self-review, no comprehension of the topic. Its a bullshit artist. People dump junk into it and it spits out matches based on training sets around that junk. Its basically a search engine with less work on the searchers part, with the same junk responses and lovely links.
|
|
#
?
Feb 16, 2023 19:24
|
|
|
Decon posted:don't have the capacity to understand or execute complex tasks. Exactly like Musk.
|
|
#
?
Feb 16, 2023 19:25
|
|
|
Strong Sauce posted:
The really concerning thing is that Tesla stock doubled in price in the past six weeks.
|
|
#
?
Feb 16, 2023 19:26
|
|
|
I mean, do we really? I agree it needs a recursive self-review but that seems programmable. That would solve the "I just bought Arby's" "the Arby's is closed" issues.
|
|
#
?
Feb 16, 2023 19:26
|
|
|
zedprime posted:If your widget had a design defect and you need to take it into the dealer for widget 2.0 that is a recall. I mean, in tech we usually call it a hotfix, but I guess that doesn't convey the urgency when it comes to vehicle software.
|
|
#
?
Feb 16, 2023 19:27
|
|
|
Random Stranger posted:The really concerning thing is that Tesla stock doubled in price in the past six weeks. It got pumped up by a bunch of investment groups, including a media campaign around "Tesla is the underdog stock to invest in right now!" Incelshok Na posted:I mean, do we really? Yes. We do. What you are seeing is NOT AI nor Intelligent. Trust me, I work with this stuff. Its dumb, it doesn't do anything but spit out junk like a search engine framed with some nice fake conversationalist crap. Stop trying to humanize it. Its not a simple programming issue, if it was we'd have real AI by now.
|
|
#
?
Feb 16, 2023 19:29
|
|
|
Bottom Liner posted:Sorry about your dumb car No I'd just read this previously. You couldn't pay me to get into a Tesla death traps.
|
|
#
?
Feb 16, 2023 19:30
|
|
|
CommieGIR posted:Yes. We do. What you are seeing is NOT AI nor Intelligent. Trust me, I work with this stuff. Its dumb, it doesn't do anything but spit out junk like a search engine framed with some nice fake conversationalist crap. How is it different from anyone on twitter (or here for that matter)
|
|
#
?
Feb 16, 2023 19:31
|
|
|
I'm confused. Should I buy gold, or should I buy bitcoin? Can't someone just give it to me straight
|
|
#
?
Feb 16, 2023 19:32
|
|
|
Incelshok Na posted:Most human thought is just pattern matching. That's why failures in pattern matching, like schizophrenia, are so debilitating. Concerning.
|
|
#
?
Feb 16, 2023 19:33
|
|
|
mobby_6kl posted:How is it different from anyone on twitter (or here for that matter) Well don't humanize lovely Twitter posters either. Problem solved.
|
|
#
?
Feb 16, 2023 19:34
|
|
|
Karate Bastard posted:I'm confused. Should I buy gold, or should I buy bitcoin? Can't someone just give it to me straight If you can slurp it then you should buy it
|
|
#
?
Feb 16, 2023 19:37
|
|
|
Anthropomorphizing logic is a problem but more so that we don't give enough credit to what algorithms can do for us already. A simple magic 8 ball program with elementary conditional logic could probably run Twitter better than Musk for example. Give me all the weird language models. There's wisdom in their imbecility. We are all Bing bot.
|
|
#
?
Feb 16, 2023 19:39
|
|
|
pixaal posted:This is incorrect, it's an OTA software update being reported by people that do not know what it is. They are turning off the rolling stop feature. The recall itself says “nhtsa.gov” posted:
So a little more than just rolling stops
|
|
#
?
Feb 16, 2023 19:39
|
|
|
zedprime posted:Anthropomorphizing logic is a problem but more so that we don't give enough credit to what algorithms can do for us already. A simple magic 8 ball program with elementary conditional logic could probably run Twitter better than Musk for example. Yeah its not to say its not useful, but Musk going off the rails about "OMG its gonna end the Human Race" is outright stupid. I cannot wait for some company to setup an AI for internal use and it ends up compromising all their systems because it got trained on some lovely or purposefully malicious training sets. Its gonna happen.
|
|
#
?
Feb 16, 2023 19:43
|
|
|
CommieGIR posted:Yes. We do. What you are seeing is NOT AI nor Intelligent. Trust me, I work with this stuff. Its dumb, it doesn't do anything but spit out junk like a search engine framed with some nice fake conversationalist crap. I work extensively with this stuff as well, and while I will keep scoffing at stuff like model debiasing, this reductionist perspective is getting less and less useful by the day. Bio-mimetic approaches can never escape from critiques like yours, yet axiomatically must be able to achieve human-level intelligence. Something's going to have to give. There's a few ways to approach this. This includes, but is not limited to "Humans may not be that particularly smart in the first place" which, mark my words, is almost certainly going to be the mainstream consensus for a while. To be clear, I agree with you in the sense that my take is that bio-mimetism will ultimately be a dead-end as far as what we want to use AIs for, especially as far as compute efficiency is concerned. But I disagree with what they are capable of being Aramis fucked around with this message at 19:46 on Feb 16, 2023 |
|
#
?
Feb 16, 2023 19:43
|
|
|
|
| # ? May 28, 2024 07:04 |
|
|
Aramis posted:Bio-mimetic approaches can never escape from critiques like yours, yet axiomatically must be able to achieve human-level intelligence. Something's going to have to give. There's a few ways to approach this. This includes, but is not limited to "Humans may not be that particularly smart in the first place" which, mark my words, is almost certainly going to be the mainstream consensus for a while. Cool so when you have the leader, ChatGPT recommending things that violate comment sense including recommending Free Energy solutions and Torture, as well as returning malicious content or recommending libraries that don't exist because its training data isn't a replacement for actual experience and training. Its junk. Its junk that's gotten better, but its still not an AI. Human level intelligence implies it understands what it says or what it returns. It does not. It knows whatever its deliver matches the training sets its been trained against. That's it. That's not comprehension. Aramis posted:To be clear, I agree with you in the sense that my take is that bio-mimetism will ultimately be a dead-end as far as what we want to use AIs for, especially as far as compute efficiency is concerned. But I disagree with what they are capable of being Nobody is, or should be, expecting AI, when it does arrive, to be a perfect match to human intelligence. But intelligence implies a real understanding of what the data you returns means when you give it to someone. I'm also not disagreeing that Machine Learning is incapable, it has a lot of uses. But its not AI. Its not what Musk thinks it is. Its not what its being sold as. What is being sold to us is, basically, the New Search Engine. Not an intelligence. CommieGIR fucked around with this message at 19:54 on Feb 16, 2023 |
|
#
?
Feb 16, 2023 19:45
|
|