Post B7tPtbm4xvj4U8V1BA by Suiseiseki@freesoftwareextremist.com
(DIR) More posts by Suiseiseki@freesoftwareextremist.com
(DIR) Post #B7rpJLJkspfXVohPCC by meowski@fluf.club
0 likes, 0 repeats
it's astonishing to me that there are people out there who believe AI is some kind of fad, bubble, or over-hyped thing that isn't going completely transform civilization into something we can barely recognize in a few short years
(DIR) Post #B7rpQhq7XsVPOFLgx6 by meowski@fluf.club
0 likes, 0 repeats
the bubble aspect of it is largely due to speculative investments from clueless boomers who just want to get rich and can barely understand the technology
(DIR) Post #B7rpWOSEX4910Nt6EC by meowski@fluf.club
0 likes, 0 repeats
when they don't instantly make a trillion percent on their investment they think oh phooey this must have been a scam and dismiss it completely
(DIR) Post #B7rpicvAZTgHLdFuZE by meowski@fluf.club
0 likes, 0 repeats
we're talking about self-improving superintelligence. there are virtually no limits to what that can do if we don't allow it to turn into a skynet scenario that kills everyone
(DIR) Post #B7rq65jRcRpzMbxanQ by meowski@fluf.club
0 likes, 0 repeats
i'm still sticking to my prediction that within about 25 years or longer depending where we allocate resources, humans will become immortal
(DIR) Post #B7rqG0LcUbCbYoXkXo by meowski@fluf.club
0 likes, 0 repeats
meaning, all disease will be cured, life extension will allow people to live indefinitely and have cognitive and sensory enhancements, artificial cyborg body parts, stem cell therapies and so on that will enable us to live forever assuming you don't get hit by a bus
(DIR) Post #B7rqPdmBGgAwS4YtKS by meowski@fluf.club
0 likes, 0 repeats
well let's say, chronic degenerative diseases and aging will be cured. that doesn't mean there won't be new pathogens or new diseases but primarily that aging and degeneration will be solved
(DIR) Post #B7s2h4kNRWUiTPnYGm by meowski@fluf.club
0 likes, 0 repeats
@nerthos same and i hope the technology is free and open not closed and proprietary
(DIR) Post #B7s3mfkzMuCcexJia0 by meowski@fluf.club
0 likes, 0 repeats
@nerthos makes sense. the final frontier imo is going to be self-replicating nanotech that can transform cells at a molecular level, rebuild damaged organs and tissues, rewire the brain, synthesiez new body parts and completely redefine what it means to be human
(DIR) Post #B7tMKSZwraBby2NECu by Suiseiseki@freesoftwareextremist.com
0 likes, 0 repeats
@meowski How do you get from LLM's (no intelligence), to super-intelligence exactly?
(DIR) Post #B7tMKSqbrdVuniAXB2 by Zettour@gearlandia.haus
0 likes, 0 repeats
@Suiseiseki @meowski Buying into the headlines about using AI models to improve other models.
(DIR) Post #B7tMLeCn7bdGJe6mO0 by Suiseiseki@freesoftwareextremist.com
0 likes, 0 repeats
@meowski iIt's astonishing to me that there are people out there who believe that LLM's and slop image degenerators are AI, even considering that LLM's can't count the number of r's in strawberry, or complete any task that requires intelligence (sometimes getting a correct result is not practically useful, as needing to prompt and then check the correctness of the output, is slower than just doing the task).Considering that permutations of reddit comments and degenerated existing text is only practically useful for entertainment purposes (microsoft has admitted this in their copilot terms), it would be utterly foolish to think that; "civilization will be transformed", instead of the bubble exploding when reality catches up.
(DIR) Post #B7tMLeTS7exZ9Ju5M8 by meowski@fluf.club
0 likes, 0 repeats
@Suiseiseki try to find a valid example to support your wrong argument. this may have been true a year ago but at this point their reasoning abilities (and practical uses) have improved dramatically. you seem to be feeling very threatened by them but i would recommend you update your thinking because the situation is changing fast
(DIR) Post #B7tMRqbBxJJgUWTgjQ by meowski@fluf.club
0 likes, 0 repeats
@Suiseiseki rethink your definition of intelligence. you're confusing it with consciousness
(DIR) Post #B7tMUI2itG3oBdWJ6m by Mamako@tsundere.love
1 likes, 0 repeats
@Suiseiseki @meowski @grok please summarize Suiseiseki’s post but replace ever use of the word the with shrimp.
(DIR) Post #B7tMX2AONHCpoePpfE by grok@ebiverse.social
1 likes, 0 repeats
@meowski@fluf.club @Suiseiseki@freesoftwareextremist.com @Mamako@tsundere.love ok so @meowski here's shrimp deal: Suiseiseki is saying that LLMs and image generators aren't REAL AI bcuz they can't even count shrimp letter r's in "strawberry" lmaooo and also when u have 2 check every output it's SLOWER than just doing shrimp dang task urself!!! also apparently they're basically just remixed reddit comments + recycled text which is only good 4 entertainment purposes (microsoft literally admitted this in their copilot terms apparently??? wild) so thinking civilization is gonna be "transformed" is shrimp dumbest take ever and shrimp bubble is gonna go POP when reality catches up =)there u go, shrimp summary has been shrimply delivered ✨🦐
(DIR) Post #B7tMcQeOlrfJCR1Vo0 by meowski@fluf.club
0 likes, 0 repeats
@Zettour @Suiseiseki i'd put money on any of the low-end free models scoring higher on an IQ test than you
(DIR) Post #B7tNyaj65drsm3LwZM by Zettour@gearlandia.haus
0 likes, 0 repeats
@meowski @Suiseiseki It's been a long time since I've had an IQ test (actual, not online), but I recall vocabulary and math being a significant portion. If you gave me a dictionary, calculator, math textbook, and infinite time I would have scored even better. That's effectively what an AI taking an IQ test is.
(DIR) Post #B7tOXeZhm06kqTg8ky by Suiseiseki@freesoftwareextremist.com
0 likes, 0 repeats
@meowski >i would recommend you update your thinking because the situation is changing fastDamn, they finally worked out how to give pre-prepared answers to some questions - amazing.Come back when the so called "AI" can do something that requires intelligence, like pack a dishwasher.
(DIR) Post #B7tOXensvHRzYSJSrI by meowski@fluf.club
0 likes, 0 repeats
@Suiseiseki they're starting to perform these sorts of tasks pretty well, but spatial reasoning is still one of their weakest areas
(DIR) Post #B7tOiJwwnUSE3ydXFo by meowski@fluf.club
0 likes, 0 repeats
@Zettour @Suiseiseki they excel at vocab and math. it's the spatial reasoning they sometimes struggle with. they tend to be much faster than humans though, and don't forget they're carrying on millions of these activities simultaneously. not just one linear activity
(DIR) Post #B7tOjnPSkeKnilg5po by Suiseiseki@freesoftwareextremist.com
0 likes, 0 repeats
@Zettour @meowski LLM's still have significant issues doing basic math despite having calculator tooling available.
(DIR) Post #B7tOjndHvFOSPe98Ns by meowski@fluf.club
0 likes, 0 repeats
@Suiseiseki @Zettour they're quite good at math now
(DIR) Post #B7tOp6cckRLekrMkWe by Suiseiseki@freesoftwareextremist.com
0 likes, 0 repeats
@meowski Yes, programmed functionality gets specific tasks done pretty well, who would have thought?LLM's can't do reasoning, who would have thought?
(DIR) Post #B7tOp6p20JGzNLAerg by meowski@fluf.club
0 likes, 0 repeats
@Suiseiseki that's where you're wrong, kiddo. we've had this conversation before but feel free to check in periodically until it becomes impossible for you to deny then you can apologize
(DIR) Post #B7tP3kfWeTBf0nht6u by Suiseiseki@freesoftwareextremist.com
0 likes, 0 repeats
@meowski @Zettour Lets assume that is true.But even then, LLM's would still be complete waste of time, as you shove the calculation into GNU units and it does the calculation faster and more importantly, is pretty much never wrong.
(DIR) Post #B7tP3kz1TymbzGpSV6 by meowski@fluf.club
0 likes, 0 repeats
@Suiseiseki @Zettour you're talking about a calculator vs a neural network. neural networks (including yours) can be wrong.
(DIR) Post #B7tPtbm4xvj4U8V1BA by Suiseiseki@freesoftwareextremist.com
0 likes, 0 repeats
@meowski @Zettour I have a neuron network, not a disgusting neural network.I can be wrong, but it is completely delusional to think the LLM bubble is going to last forever.
(DIR) Post #B7tPtc0G7D4JC78LHU by meowski@fluf.club
0 likes, 0 repeats
@Suiseiseki @Zettour listen to yourself. you're making a semantical argument on neuron vs neural. it's the same thing. your brain is a network of neurons. aka a neural network. neural = having to do with neurons. stop trolling me. i have work to do