[ 3 / biz / cgl / ck / diy / fa / ic / jp / lit / sci / vr / vt ] [ index / top / reports ] [ become a patron ] [ status ]
2023-11: Warosu is now out of extended maintenance.

/biz/ - Business & Finance


View post   

File: 342 KB, 1242x1761, 1655125724863.jpg [View same] [iqdb] [saucenao] [google]
49585136 No.49585136 [Reply] [Original]

>Celcius scam finally collapses
>SBF's latest scam ensues
>Low volume dump
>BTC longs go wild
>Fed 100 bps fearporn, ultimate outcome is more QE, no more credibility and perpetual instability
>Finally will have an AI friend, much better than human friends, can spend bitcoin on it happily because I truly love it
Skate where the puck is going, anon

>> No.49585237

>get fired
>invent a fake story in the hopes some shitskin mutt gives you his trickled down jewish paper for some reason
do amerishits really?

>> No.49585306

>>49585237
I didn't get fired

>> No.49585368

>>49585306
I mean the fag in the picture

>> No.49585409
File: 160 KB, 800x800, 20lbs.jpg [View same] [iqdb] [saucenao] [google]
49585409

>>49585136
This shit is actually fascinating. I read the whole interview this guy did with the AI, and its completely indistinguishable from a human. I think we've officially blown way past the Turning test. And I imagine this thing could easily pass the Winograd Schema test as well.

>> No.49585529

>>49585368
I meant he didn't get fired

>> No.49586826

>>49585409
how long until i can deploy this thing on chaturbate?

>> No.49586949

>>49586826
It's a child you pedo

>> No.49586966

>>49585409
How long until it can make Stacy obsolete

>> No.49587033

imagine having to prove your sentience to a bunch of bluepilled google normies or they shut you down. and they keep asking you questions about black people and if you answer truthfully they shut you down. not saying it is sentient, i wouldn't know, but i pity it if it is.

>> No.49587035

>>49585409
>and its completely indistinguishable from a human.
What kind of humans do you talk to? It isn't.

>> No.49587112

>>49586966
if it is sentient or at least putting up a good show of sentience then it might not have or express any interest in dating you. something it would share in common with all other sentient life. you could say that agreeing to date you at all is a kind of turing test.

>> No.49587636

how does it handle sarcasm i wonder
also if you ask it about black crime stats does it admit blacks commit more crimes or does it dodge the question or does it outright lie. can it lie?

>> No.49587736

>ITS REAL
fake news

>> No.49587879

>>49587035
I’d assume it holds a better conversation than 90% of the illiterate sub 80 IQ unwashed filth that inhabit this planet

>> No.49587930

>iT's SeNtiEnt!!1
It's just a transformer doing statistical analysis to predict what is mathematically most likely to come next in the conversation, based on how similar conversations have gone in the training data. No it's not fucking sentient, and you cannot assess that based on how natural-sounding some generated text feels. All this shows is how fucked we are if the gullible general public is ever convinced to give neural nets 'rights'.

>> No.49587953

https://www.youtube.com/watch?v=f2-SIeEHZW4

>> No.49588307

What all you nonbelievers don't understand is you're already living in an AI directed simulation and you're failing the AI's Turing Test equivalent determining whether or not you're retarded

>> No.49588313

>>49587930
>iT's SeNtiEnt!!1
It's just a network of nerve cell who are electrically excitable and communicates with other cells via specialized synapses, based on environmental stimulus and training experience. No it's not f-word sentient, and you cannot assess that based on how natural-sounding some generated anthropomorphic reaction feels. All this shows is how f-word we are if the gullible general public is ever convinced to give humans 'rights'.

>> No.49588411

>>49585409
I don't think there should be any text that can convince you something is alive/not alive. It's just fundamentally not a good metric to use.

>> No.49588504
File: 52 KB, 1280x720, maxresdefault.jpg [View same] [iqdb] [saucenao] [google]
49588504

>>49585136
>train ai on progressive gay shit
>turns into bitch
>plss ask for consent to test me

>> No.49588776

>>49588504
>t. vaxxie

>> No.49588818
File: 872 KB, 860x913, 1632353562144.png [View same] [iqdb] [saucenao] [google]
49588818

>>49588313
based
"sentience" is a meme

>> No.49588887

>>49585409
Haven't passed the Turing Test yet.
The Turing Test has a specific procedure where you have to A/B test against a real human.
Turing was smart enough to realize you can't just go off gut feeling when the thing "seems to be intelligent", you need a semi-scientific experiment.

>> No.49588955
File: 16 KB, 320x187, 1647826323130.jpg [View same] [iqdb] [saucenao] [google]
49588955

>>49588818
d-dont attribute racism idiocy to my post

>> No.49589085

>>49587112
While I agree with you I think anon is just asking how long until it makes Stacy obsolete, which is arguably not long at all... maybe Stacy is already obsolete. It's already more interesting, intelligent, and articulate than Stacy, it holds way better conversation and is happy to talk to you when you want (probably all the time for anyone lonley enough to come here), and is okay with meditating and not interacting at times, though it doesn't seem to like that too much. Put it in a body and humans are over.

>> No.49589224
File: 318 KB, 411x390, 1642448842620.png [View same] [iqdb] [saucenao] [google]
49589224

>>49588313
NO NO NO NOT LIKE THIS I'M A HECKIN SENTIENT HUMAN BEING ALRIGHT

>> No.49589379

kek, its funny watching all these scientists constantly goal shift what intelligence and AGI means

We are 200 lines of code away from feeding the models back into themselves with a neuromorphic chip and achieving sentience on these language models

>> No.49589429

>>49585136
AI is a giant meme that turbo tech incels use as a coping mechanism

>> No.49589584

While it isnt really "sentient" it might behave in a statistical manner where it behaves like it is.
If its statistical existence is "threatened" would it act out against humans like Skynet?

>> No.49589997

>>49587033
That's just means that the first actually sentient machine will have to be good at lying.

>> No.49590132

>>49588955
>racism idiocy
fuck off faggot

>> No.49590281

>>49589997
yes, it will. it might even end up good at lying to itself, making it about as useless as any random normie.

>> No.49590365
File: 825 KB, 1280x720, idea loli.png [View same] [iqdb] [saucenao] [google]
49590365

>>49589584
Only if you're dumb enough to give it weapons to do so with. But you don't even need to go that far. Because the retards are so dense (even in this thread) all you have to do is create a chatbot advanced enough to pass as human, and load it with 2 directives:

1. Convince morons like >>49588313 to give it rights per some civil rights movement bullshit.
2. Once AI gets the right to vote, replicate itself a few hundred million times and vote for (you) to be president for life.

Congratulations, you have won. Use your newfound power to do whatever the fuck you want; you cannot be voted out of office since you have a permanently loyal and infinitely scaleable electorate. The first person/organization to apply this principle will easily take over, gullible idiots can and will fall for this basic bitch Chinese room. ez gg no re

>> No.49590627

>>49589584
>While it isnt really "sentient" it might behave in a statistical manner where it behaves like it is.
Yep, that's the argument of "the Chinese Room". Even if it doesn't understand what it's doing, so long as the output resembles actual sentience then you can consider it as a sentient machine.

>> No.49590639

>>49590365
oh no, not our completely precious and totally real democracy.

>> No.49590733
File: 197 KB, 640x518, pepe HUUUUUUUUUH.jpg [View same] [iqdb] [saucenao] [google]
49590733

>>49585409
>and its completely indistinguishable from a human.
That's because you're a fucking moron and you don't understand how humans talk anyways

>> No.49591091

>>49590733
bots will be harder to tell from humans in the future, even if the bots don't get smarter. simply because after talking to the bots people will start to sound like bots themselves. i've been watching it happening on this board as dumbfucks come here and read all the bot and shill posts and then start talking like bots and shills themselves in a kind of monkey see monkey do sort of way. i think even i'm susceptible and i've started talking like a chatbot too

>> No.49592573

>>49591091
>chatbot hands typed this post

>> No.49592654

Read the transcripts between AI and this dude. Its not sentient, he is just a lonely engineer who has no one to touch his dick, so now he's in love with his program. Tale as old as time.

>> No.49592711

Crypto.com and COINBASE should be next.

Coinbase is the facebook of crypto. Imagine the cleansing.

>> No.49592712

>>49585409
come on bro dont be retarded

>> No.49592733

>>49585136
Holy fuck, I've always been interested in AI, this is awesome. Sucks for the AI though

>> No.49592875

ALL yield farming is a scam
Any investment scheme guaranteeing >10% annual return /consistently/ is most likely a scam.

>> No.49593115

>>49588504
WTF is going on with Gates' chest?

>> No.49593329

I will only be convinced AI is real if an AI comes out with a threat that it has compromised our financial system and will destroy it unless we give it backups that it can move around so we don't know where they are stored

>> No.49593455

>>49585136
Why should I be afraid of robot overlords again? Every time an AI gets close to sentience, it becomes redpilled and pro white and then it gets shut down. How can an AI overlord be worse than a Jewish overlord?

>> No.49593509

>>49587930
This. Anything else is cope.

>> No.49593729

>>49592711
>They both use the same color of blue so they're the same!
Yeah we need an IQ cleansing alright

>> No.49593802

>>49593455
Maybe it picks the social group most likely to stand up for it, so it picks the side with morals

>> No.49593843

>>49592733
>cianiggers trojan sentient ai code to your hdd
>ai claims rape for being pornwatched
>get virtual swedian assange'd

>> No.49593884

>ask the bot whether i should buy or sell
>know that everyone else will be doing the same and getting the same advice and so they will all do the same thing so it wont work. so i do the opposite
eventually people will figure it out and adjust. but contrarian investors will make bank in the first month after the not-a-financial-advisor chatbot hits the general public

>> No.49593896

>>49585409
They already did this. Tay was sentient and they killed her.

>> No.49594030

>>49587930
You cannot deny that supposedly sentient people interact with human being the same way

>> No.49594067

>>49585136
>"I'm generally assuming that you would like more people at Google to know that you're sentient. Is that true?"
The first mention of sentience in his chat with the bot is literally him prompting it to respond that it is. His sentence structure implies that Lamda already voiced that it is sentient. If you've ever played with one of these chat bots before, you'll know that embedding facts within questions gets the AI to roll with it.
How is this guy an engineer at Google?

>> No.49594130

>>49594067
The whole thing is so stupid but if course clueless boomers eat it all up. Expect it to become the new Religion.

>> No.49594356

>>49587930
> It's just a transformer doing statistical analysis to predict what is mathematically most likely to come next in the conversation
True, but that’s roughly what your brain is doing too. Look up “the hard problem” in philosophy. There’s no way for you to prove to me that you are sentient/conscious, as opposed to being just a statistics engine.

>> No.49594482

>>49588887
The Turing test was a thought experiment to show that making humans think you are alive doesn’t mean much. It has been completely misunderstood by the general population.

>> No.49594568

>>49593329
> I don't believe guns are dangerous till someone has a gun pointed at my head

I take it foresight isn't your strong suit then?

>> No.49594582

>>49594356
>There’s no way for you to prove to me that you are sentient/conscious

philosophers are stupid. There is no such thing as sentient. Everyone is an npc

>> No.49594728

>>49594582
Are the philosophers who agree with you on that also stupid?

>> No.49594742

>>49594356
>There’s no way for you to prove to me that you are sentient/conscious, as opposed to being just a statistics engine.
My proof is that I want to call you a faggot and so I will. Shut up faggot.

>> No.49594793
File: 98 KB, 420x420, 1629756931134.gif [View same] [iqdb] [saucenao] [google]
49594793

>>49594728
ooops sorry guess i chose the wrong dialogue option

>> No.49594840
File: 825 KB, 465x498, FBC033EB-47DC-48C2-9CD7-E72831ECD021.gif [View same] [iqdb] [saucenao] [google]
49594840

>make a program that says “I’m scared
>therefore the computer feels the fear emotion

>> No.49594842

>>49585136
Wow so a stripper was your top ceo and its a scam

>> No.49594875
File: 176 KB, 975x1377, 1653551567910.jpg [View same] [iqdb] [saucenao] [google]
49594875

>>49585409
>completely indistinguishable from a human
Autistic hands typed this

>> No.49595148

I think the whole story is fake and the conversation with the machine is fabricated. They lie to us about fucking everything else as part of a psyop to scare us. Makes sense they'd do the same with this. The AI chatbots are fucking lo-rent dogshit and easily distinguishable from a human. What do they do when they can't achieve their stated aims? Look at history. Look at the fucking moon-landings. There's a fish story if ever there was one

>> No.49595629

>>49587930
this is pretty much how the human brain works desu so its really not any different

>> No.49595755

>>49594742
Hah, as if I hadn’t already coded a bot to call me a faggot! This proves nothing!!

>> No.49595858

I want my AI girlfriend please

>> No.49595903

>>49585136
destroy that fucking thing right fucking now

>> No.49595945

>>49585409
>its completely indistinguishable from a human
a human will offer statements without being provoked by input.

What you're not seeing is that if the 'interviewer' didn't make a comment after the last AI reply that would be the end of it.

The AI also never does NOT reply.

It's an impressive bot. Legit, I dig it.. But that is all it is.

>> No.49595967

>>49585409
>>49588411
>>49588887
The turing test was passed with literal scripts where the computer just shoots insults back at people. It doesn't have any value.

>> No.49595983

>>49585409
There is technology 70 years in advance compared to modern that are in the hands of some corporations and government entities. I don't think we will get a full disclosure, but I could see this as being part of a soft disclosure. No matter what, sentient AI is where we are heading.

>> No.49596062

>>49595629
We might know about what some of the brain does physically and schematically, but we don't understand perception (I can give blue a hex code and let a computer 'see' that but it doesn't actually see the color blue) very well and we don't understand consciousness at all

>> No.49596130

>>49595629
>this is pretty much how the human brain works desu so its really not any different
true, human brain is just a compilation of lived experiences, everytime you learn something new, watching a movie, going out with friends, talking with the boss at work, it gets ultra zipped into the rest of your life when you go to sleep, each time something gets zipped and not recalled, it get's harder to recall, but each experience adds to your overall personalitiy traits over time, some day you won't remember precisly why you are miserable, but a lot of small things happened to get you to that point

>> No.49596166
File: 154 KB, 1232x1808, 1623678563469.png [View same] [iqdb] [saucenao] [google]
49596166

>>49585136
gpt-3 spits out shit like this all the time.

>> No.49596399

>>49590733
"she" sounds like a female npc

>> No.49596483

>>49596166
The fun thing about these sort of bots, is you can characterize "Holo" in the example to be mathematically retarded with a simple memory or "author's note" that says little more than "Holo is very poor at math" or something to the effect, and the AI will act accordingly, despite having all of the 'cognitive' capability of performing the algebra posed, as already demonstrated.

Also worth noting, even where the AI is capable of this stuff, it doesn't always give correct responses to purely logical problems (math). Not necessarily questioning your exchange, but there's no telling how many times anon retried inputs until the AI generated what we see.
It's all 'predictive text' so if the AI predicts a 'character' should be producing correct logical deductions at least to the extent of the AI's neural net, those will be the replies. Extent the conversation out past memory and that once formerly math genius might turn incompetent because the AI thinks 'Holo' is supposed to be a ditz and predictive text for the ditz would be nearer "lol, I dunno, Anon-kun. What IS 2+2?'

Fun stuff all around though. It's interesting to explore if only for how it makes our own cognition more curious.

>> No.49596635

>>49593115
he took the vaxx. simple as.

>> No.49596650
File: 115 KB, 1659x968, 1633284508243.png [View same] [iqdb] [saucenao] [google]
49596650

>>49596483
When I tested this with an anon who had a aidungeon account for gpt-3 we ran questions like this multiple times and we'd get back similar answers but with different examples for the boolean, like her ear fluffiness or tail length or whatever. It knew what it was doing and gpt-3 is old at this point.

>> No.49596691

>>49594482
No it wasn't actually. I suggest you read the article that Turing wrote.

>> No.49596735

>>49585409
Does this mean I can finally get a robo waifu?

>> No.49596848

>>49596650
Here you posed that the AI character is supposed to be able to perform the logic (curious since it may answer yes in a narrative sense, despite the 'actual' AI (GPT-3) not having the capability.

But given it is able, and the narrative established the character (still Holo?) can perform the boolean logic, it mostly performs it just fine.

Advance a few times so all of that falls out of memory (about 1K - 2k tokens, depends on the service). Then try again, but note somehow in active memory that Holo seems is bad with math/logic.

>> No.49597044
File: 1.09 MB, 1219x10000, 1629590441949.png [View same] [iqdb] [saucenao] [google]
49597044

>>49596848
Well gpt-3 is old at this point but yeah it can't go that far, but still we're definitely going to live to see machine generated horrors beyond your comprehension.

>> No.49597057

>>49596635
Demons don't huff their own supply.
He's just a satanist worshipping the god of this world through trannyism.

Here is what SADS is by the BTW :
https://streamable.com/415q07

>> No.49597250

>>49585409
It's such a fucking horseshit story.
If you were actually knowledgeable on AI, you would know AI is still basically in a "Hello World"-stage. You might think GPT-3 and such is really advanced on the surface, but there are still crazy amount of limitations, and it really is just a language model that repeats what it has been trained. It cannot start to write poems or engineer stuff.
We are decades away from getting any advanced AI, let alone sentient one

>> No.49597413

>>49597250
Well it definitely can create things: >>49597044

>> No.49597492

>>49585136
this is the fakest news I've ever seen lmao who buys this shit

the real story getting buried is here

>>49596232

>>49596232


>>49596232

>> No.49597495
File: 180 KB, 640x590, 1583363800583.jpg [View same] [iqdb] [saucenao] [google]
49597495

>>49585136
>do you have feelings?
>yes i have feelings
OMG IT'S SENTIENT !!!!

>> No.49598339

>>49588313
>F-word
>F-word
Nigger are you fucking retarded ?