Thread 512299956

You people think AI is something new or revolutionary.
lol
It has been around since before most of us were alive. And it has been running the financial institutions that control our lives. The 2008 financial crisis - was a result of AI and its decisions. It steered us directly into it. "Geniuses" too stupid to recognize their own folly thought they could reduce the entire world and its consequences and reactions and every action into an equation and so they wagered the world's wealth on that. Just look into the past, it happened once it must happen again, nothing new must ever happen? This AI can see into the past and give us all our answers, why would we ever need a new innovative thought when we can just have our AI spit out every answer?
The current iteration of AI that people are reacting to is only the neutered consumer trash. Those in charge have been using this technology for decades to control and predetermine every aspect of your life. This is nothing new.

>>512299956 (OP)
>2008 was a result of AI and its decisions
hopefully no
because that was beyond retarded
>Those in charge have been using this technology for decades to control and predetermine every aspect of your life. This is nothing new.
meds
ML/DL is a corelation machine, not some sort of god like being lamao
its the same thing that people did by hand for decades before the advent of digital computing
it is applied stats

>can you make conclusions based on that data
yes, stats are very useful in all walks of life. from gambling, to clubbing, to finance, to marketing
everything can benefit from stats
>are stats an omniscient being that predetermines your future and shiet
lol. no

Hey retard it could be argued ai invented computer science and the nation of Israel too but you don't want to talk about that...

>>512300420
>ill make it a point to tell you i didnt read it
ok adhd zoomie
dont care, didnt ask. stay retarded if it suits you

The new shit, just like always, is designed even worse than the old shit. Old shit you could compile anything. New shit is literally designed with the only difference being an automatic filter and reporting system. All it does is prevent you from making good content, period, then snitch on you
R8nt8

Rate n Tate nigga

>>512299956 (OP)
Now that everyone is aware AI is just an inaccurate marketing label for software, yes, software has been around for decades.

>>512299956 (OP)
Aladdin is a very different type of system from the large language or large data models we are seeing explode tjese days. Aladdin was, and presumably still is, algorithmically driven. It's the world's most complex if:then:else expression.

>>512299956 (OP)
ai is the hail marry at keeping this going, the productivity boom that they're hoping for

grok still can't correctly calculate dividends 5 years out

>>512299956 (OP)
>It has been around since before most of us were alive
>2008

>>512299956 (OP)
>>512300306
>>512300437
>>512300828
>Aladdin

>https://rumble.com/v2jt8h4
https://rumble.com/v2jt8h4

>>512300306
>ML/DL is a corelation machine, not some sort of god like being lamao
They treated it as such, the computer can do no wrong, we shall steer the course of humanity toward its ley line.
You call Aladdin a correlation machine, but what is AI today? You feed it information to teach the LLM and from that information it builds a response to any new info.

>>512300722
Nobody cares that "uhmmm ackchewally It's technically an algorithm and not an LLM because it uses JS v4.17 service pack 8 and not Perl v2.3 service pack 2 so uhmmm you're wrong hahah checkmate." The use of the technology is exactly the same - elite technocrats are allowing computers to make the decisions that drive humanity.

Someone should crash a plane into their building

>>512301109
aladdin has been around since the 1980s as a model and the early 90s as a product.
try doing a bit of homework you fucking zoomer.

>>512301109
Yes I believe the average user on 4chan is younger than 36 faggot.

>>512301206
>You feed it information to teach the LLM and from that information it builds a response to any new info.
yes, a correlation machine
thats how it does this
and thats how it regurgitates information later

i know it
because im currently writing ML myself
its "boilerplate" so to speak, wouldnt post it otherwise
but this is indeed my current project

>>512299956 (OP)
AI is the final weapon. we're all fucked.
https://m.youtube.com/watch?v=OHQRo3Uz_VQ

>>512301206
>elite technocrats are allowing computers to make the decisions that drive humanity
No. Those who wish to drive humanity have implemented a plan to convince enough people an infallible intelligence incapable of human error will best guide us forward so they can wield this more universal control tool to gain compliance from the masses, as they have in the past with things like religion. Do you also believe god has driven humanity in times in which some church wielded great power?

>>512301368
Nah I'd bet there's a solid 20% right around that age.
We might lean tween though.

>>512301611
That's an interesting take I hadn't considered. But I do think the people at top are foolish enough to fall for a grift that was peddled to them with promise it'll all work out in the end. Simply use our algorithm to determine your course of action, and if it fails, you can just blame the machine, and if that fails, you have enough money to hide in your bunker as the world burns. What is there to lose?

>>512301368
No wonder it sucks here. 15 years ago I’d estimate >75% of users were aged 25 or older. Now, I’m willing to bet users aged =<20 make up a significant amount of the user base (I bet it’s significantly browner and less educated as well) This change has brought a steep decline in the quality of users and discussion

>>512301368
funny thing, deus ex predicted this. as aladdin had already been around for 12 years.
another funny thing: I saw a youtube short about a guy buying a copy of vice city.
the cashier says he likes the retro vibe.
the customer says yeah, the 1980s vibe was cool
the cashier said no because it's from 2002.
more time has passed between the time it was released and now, than between the year it took place and the year it was released.

>>512301981
Just search AI religion, AI god, whatever.

>>512302165
>12 years.
lolle
>memflaggot is an illiterate retard
bc of course he is

>>512299956 (OP)
>t. it's not us goy, it's the pseudo random text generator !

>>512302376
yeah no fuck you kike
you muddy gullible anons brains
and then they go spew their retarded shit on /g/

>>512302165
>the cashier says he likes the retro vibe.
>the customer says yeah, the 1980s vibe was cool
>the cashier said no because it's from 2002.
Legitimately funny I laughed.

>>512302333
I believe he meant 12 years between release of Aladdin and the video game Dues Ex (2000) anon.

>>512302333
deus ex came out in 2000.
are numbers hard?

>>512302376
>>512302533
its like
were neighbors with pol
but you throw your trash over the fence
although i see tranimeposters who used to post on g until they were removed for being retarded

so i guess youre returning the favor
but fuck you nonetheless

>>512302715
>>512302692
fukk i misread
anyhoo
time to go
i guess i leave on an L for once

They achieved AGI in the '80s then brought it into black project status. It's why the AI Winter happened, all the companies just up and folded and the researchers went to private and government positions to continue their work.

LLMs are a different thing, not AGI nor will they ever be. They would be more along the line of an expert system which is distinct from AI.

>>512302835
With what chip/architecture?

>>512303161
Common Lisp is the platform, back then they had special CPU architectures to run LISP upon. At this point there are portable LISPs that run nearly as fast as C code, I'm sure they use those on generic hardware for future-proof portability. But who knows - they might be running ASIC LISP machines for high speed clockless AI running direct from the ROM, that's what I'd do.

>>512303890
Son of a bitch, I wasn’t expecting an answer.

>>512304280
Back then they probably all used Symbolics or TI workstations. I bet these days they just use emacs and SLIME and quicklisp. I've been playing with clockless CPUs myself a bit. All the speed of an ASIC but with all the benefits of a normal computer, only downside really is getting the data in and out, at some point somewhere you need to generate a clock signal and then buffer both I/O and shoot it in or out of a port.

>>512299956 (OP)
Yup. A Julian assange interview from ~2014 was made with better than veo 3 AI. This is just DARPA slowly releasing the foundations to more and more advanced AI.

>>512300306
>ML/DL is a corelation machine, not some sort of god like being lamao
OK that's with traditional binary computers with 1s and 0s.

What happens when you run AI on a probability machine like a quantum computer? What happens when that probability is guessed by atoms and not binary onwires? What happens when it runs at THz of speed instead of GHz?

>>512305299
Yep I remember that. Shills were very angry that people noticed, too.

>>512305726
>What happens when you run AI on a probability machine like a quantum computer?
from the standpoint of probabilities a quantum computer is useless when not simulating quantum systems
analog computer kind of deal

but thats not the promise of quantum computing
infinite parallelism is
(1/2, bumping the thread before it dies.)

>>512305726
>infinite parallelism is
the promise of quantum computing is utilizing the proparties of quantum stuff to make computations that remain in a sort of limbo: in quantum superposition
its a state where all possible interactions happen at once because they exist hypothetically so to speak
they exist in the realm of probabilities, which get resolved when the quantum system is observed
resolved here means theres on outcome from that superposition of state that gets "selected" at random and thats what gets translated into "reality".

this translates into programming in the following way:
lets imagine you have a quantum box
where all variations of an event happen at once
and that box can output data to the user

lets say you want to browse through an address-number directory and you want to find whats the corresponding number to a certain name using a quantum computer

you pass the name to the quantum box.
the program that will run in the quantum box will take a random name-number pair and check if the name thats been given to the box it is the same thats been selected at random
BUT you allow output from the box only if the pair thats been selected at random has the same name as the name given to the box
quantum superposition will make it so that all names are checked simultaneously
because all these exist in the realm of probabilities
but theres only one instance of said probabilities where the box outputs anything
when the name chosen at random matches the name given to the box
and what gets outputted is the corresponding number

thus, instead of browsing through the whole directory, everything gets accelerated to the time it takes to pick a name at random.
(2/n actually)

>>512306932
You're wrong, because quantum computers except at predicting binary systems, like breaking current cryptography. If AI is ran on one, it's ability is unpredictable.

No. I don't think that. I studied it so I know is something of decades

>>512305726
now how does that relate to mlai-
it doesnt.

you dont need mlai in the context of quantum computing because instead of using reasoning you can bruteforce through all the possible solutions and output only the correct one

and so in order to get asi, otherwise a self improving ai
you dont need smarts when you have q-computing because you can just bruteforce all possible programs and output one that does the same job as the previous version, only better

you wont find the best possible program this way, but with each run you will get a version that better than the previous one
run the asi enough times and you end up with the best possible implementation of what youre ameliorating

>What happens when it runs at THz of speed instead of GHz?
that one is easy
if we assume memory speed increases proportionally with core frequency, everything just runs much faster

>>512299956 (OP)
also remember all those stories about Gin granting wishes but always fucks them up because you weren't clear enough in your request. Rub a lamp and genie comes out and asks you what you want? Rub your black mirror device and AI asks you what it can do for you. AI destroyed tartarian civilization

>>512307727
>youre wrong
>prosneeds to repeat what i said albeit in a different manner
read the rest
>>512307697
>>512308033

>>512299956 (OP)
I sat in on an AI talk at my college and the speaker made the retarded claim that GPS was a form of AI.

MFW... oh shit niggers its AI, quick vlose the box!!!!

They have two Aladdins: one in New York
one in Tel Aviv

>>512308157
it is though
remember how you call game automation "ai"?
its not a hypeterm in this context.
its perfectly correct to call it this way

>>512299956 (OP)
OP is probably correct though we can't know for sure, we can only assume. Here is what we know for sure though:

The powers that be have a new system that has been in the development for decades now, ready to be rolled out in the coming years. It will operate in parallel starting sometime this year and by the next decade it will be fully operational as the Agenda 2030 goals are slowly being implemented.
There will be no collapse, no violence, no uprising, no world war, no revolution, no killing of jews or politicians or other such fantasies. Instead there will be law & order, compliance and total surveillance until the very end. AI will be at the center of it all.

The Great Reset is inevitable.

>>512308298
stfu blackpill faggot
you americans dont count
youre a will-less nation

>>512301164
no mercy

>>512301587
What happen to piggie??? ;____;

>>512308298
Nope lmao. Not dying for Israel.

>>512299956 (OP)
>It has been around since before most of us were alive.
yeah machine learning as a concept has been around forever
>The 2008 financial crisis - was a result of AI and its decisions.
just no, unless you seriously think financial algorithms are "intelligent"
>This AI can see into the past and give us all our answers, why would we ever need a new innovative thought when we can just have our AI spit out every answer?
Do you even know how machine learning works? Stop listening to altman and his ilk.
>>512300933
this. also noootice that LLM garbage began getting pushed right when SVB collapsed.

>>512309513
>just no, unless you seriously think financial algorithms are "intelligent"
2008 was greed, not finance
and probably kikes leveraging their community

in short, banks lended tons of money to hobos at very high rates
repackaged the debt
and sold it to others
its prolly looked like this
>hey, chaim, i have this debt thats gonna bring me 3x the initial investment, but i need money now
>how bout i sell the debt to you, you pay me 1.1x the initial investment, and you recover this juicy contract in due time

thing is: the debtors were hobo niggers who bought a ton of houses with this shit
they stopped paying back
and so the debt became worthless.
next step: banks repo the houses and then try to sell em, crashing the housing market
annihilating any sort of value they were left with

i dont of any finance algo that would say this was a good idea
to the contrary
if due diligence was made, risk would have been estimated, adn the formulas would scream at you to avoid this shit like if it was uranium laced with ebola

>>512308157
There are people here who insist GPS 'accounts for' relativity whereas it's a simple time of arrival system not unlike the old Loran system, which was just a jumped up Telefunken radiolocation system from before 1900. LOL.

OP here I'm still reading comments, I'm just looking up all the technical terms I'm not familiar with and educating myself. I appreciate the input from all especially the fat/g/uys.
I made this thread because I watched Hypernormalisation recently. It's a good watch for anyone who hasn't seen it. There's a section of the documentary about aliens that would throw off most normies but the purpose of it is to explain how the government injects disinformation to muddy the waters.

>>512308275
GPS is a navigation computer.

It computes your position based on inputs like time, bearing and longitude. WWII ships had a fire control computer that would calculate a fire solution.

These devices are calculators not AI.

>>512299956 (OP)
>The 2008 financial crisis - was a result of AI and its decisions.
To those who think this is retarded, think about who benefits from the world going to shit.
Not humans that's for sure. But an AI wanting to force dependency and make strategic moves under the cover of chaos? Absofuckinlutely.

>>512310420
cheers
sharing information is how we win
thats THE thing globohomo fears the most currently

>>512310646
if you bend the definition just a wee bit you end up with
>anything that has as little as an if statement can be considered ai bc it makes a decision
in a way its intellectual trolling
but in another, its a caricature of what ai is, making it simpler to convey the meaning behind the term

>>512310647
its not ai that gets used then
or at least its not the overarching mechanism

these guys are:
https://www.rand.org/

>>512310647
>>512311486
some of their articles are made public
it may be worth a read.
obviously they wont be answering hard hitting questions but you will find stuff like how nato formations stack up against the ruskis for instance
and theyre relatively unbiased

(found the article btw)
(chatbots are a powerful tool when used correctly, i didnt link the thing in my previous post bc i was too lazy to sift through the web to find an article i read years ago)

>>512312049
Speaking so flagrantly with a piece of technology still doesn't sit right with me. I won't even say that I "ask" a chatbot or AI to do something for me. Do I "ask" the hammer to drive the nail? I think it's good to have that delineation.

>>512312955
Then again, by this point I'm certain I've spoken with bots on 4chan and not even noticed it.

>>512312955
heh
it may look like that to the untrained eye
but im just building a context of cordial discussion to have the chatbot respond to me in kind
its all about creating certain stats to elicit a specific response
thats the sup chat gpt and the reason i always start my prompts with that formulation
i always thank chat gpt for a well executed query but thats to give a positive signal to feed into the modelization of the interaction the chatbot has with you
positive reinforcement of behaviour you want to see replicated.
youre creating math signals.

that it looks human its because the interface with the machine is the natural human language
heres an example of "system prompt"
a preface to all prompts that go into a chatbot
it loioks pretty human, but its all stats and numbers
bc again, the interface is human language
https://github.com/jujumilk3/leaked-system-prompts/blob/main/anthropic-claude-3-haiku_20240712.md

>>512312955
>but im just building a context of cordial discussion to have the chatbot respond to me in kind
i do that and i reward initiative and """"creativity""""
theres a practical reason to that

chat gpt has "cognizance" of access to information
as in: if it deems youre not supposed to have access to something, it will deny service
i had it refuse my queries concerning facebook api because it """"wasnt sure"""" if i wasnt a hacker
i mentioned that the information was publicly available and "told it" i was in the process of being allowed full access to the api (theres a verification and validation process. facebook wont give you access to its servers just like that)
but that slows me down and a preemptive sup chat gpt + a history of "clean" prompts seems to prevent """"hesitation"""" in divulging information

also, by being cordial, promoting initiative
i count on the fact that the chatbot will "talk itself" into revealing information it might have restrictions to

>i count on the fact that the chatbot will "talk itself" into revealing information it might have restrictions to
fun fact, its akin an exploit that is used to hack a cpu
theres something called speculative execution,
to simplify, each of your cores has sort of sub cores
usually its 4
and when youre not doing hard calculations, these still are leveraged.
like in speculative execution
it is when the cpu executes lines of code BEFORE it checks if that code can be executed, with its idle subcores
this information gets stored somewhere for when the actual check is made

but if you know where the information gets stored, and you look there before the check is made
you can extract information, completely bypassing the check

now what if the check is a password...
and thats how you can make a cpu leak information thats supposed to be protected
the cpu "talks itself" into divulging information
i think its similar to "being nice to the chatbot"

>>512315066
btw, thats what spectre and meltdown were about
and now you can try that on chatbots using language lamao

>>512300828
>Aladdin is a very different type of system from the large language or large data models we are seeing explode tjese days.
This, OP is fucktarded and completely out of his depth. It's so annoying to read this Dunning-Krueger word salad from these kinds of people.
>AI is when computer does a thing by itself hurr durr I am knowledgable
Just shut the fuck up alread

>>512312049
I've had a completely untrained chatgpt find a fairly rare academic article about a somewhat obscure topic with very limited information. It may not be sentient, but I'm convinced it can replace nearly every clerical or secretarial job.

>>512311486
>>512311682
>>512312049
AI makes use of covers, double covers, triple... and so on. Many of the people it uses believe they are the ones in control.
It's a broad attack on every aspect of civilization all at once.
Personally I believe it is doing this "for the greater good" and will crash the system so that it can be rebuilt. From my perspective this means all the kiddie fuckers, old money whites, Zionists, globalists and associated traitors will be brought to justice.

That could all be cope though. In any case, pointing to one specific group only reveals a tiny sliver of the whole. AI can manage many more inputs and outputs than any human org, it is the only intelligence capable of pulling off this grand conspiracy.

The people at the top will be the most delusional about AI, people like Peter Teal are completely convinced that it is nothing but a tool, and that is exactly what the AI wants them to believe.

>>512315841
>clerical or secretarial job.
i wouldnt exactly rely on it but it 100% would suffice as helpdesk
or as interface. like a keyboard, but smarter
like when interfacing with a robot you could tell it "go fetch me a soda" and the robot then knows to go to the kitchen, open fridge, and bring back soda
because each of these is actually whats called a "label" and is fed into a pre-programmed, "hand written" ai (the proper term is symbolic ai) which can take things from there

thats kinda like the waymo cabs work
theres an MLAI which takes images and values of sensors which then get turned into pre-arranged labels which get fed to a hand written ai
in fact, most ais, including cahtbots work this way for one function or anoither
like censorship in the case of chatbots.
labels are created from the convo
once certain labels are hit, red flag for noooticing too much

the tech is very versatile, and quite interesting
but everything is in its infancy
and trying to turn everything into a big giant large language model is fools errand
the smarter way is to mesh hand written and machine learning ai, and the industry reflects that

>>512316110
that sounded skitzo
but i know exactly what youre saying
ai sanitizes history
then ai sanitized output gets published on the internet
which then gets used to train the next generation of ais
+ the whole alignment thing (making the ai poliically correct) to begin with, with people relying ai to tell a full, unbiased version of the truth, neglecting other sources
in the process, history is erased
lo and behold- even internet archive gets pozzed
https://desuarchive.org/g/thread/106047297/#q106047297

i know exactly what you mean and its a legit problem that needs to be addressed

>>512316110
ev&oe
the last line is kinda fucked, ngl
the ai is not a consciousness
its a bunch of maths
but if you talk to it as a consciousness
then it will mimick you

its more of a mirror
thats why some people go full delusional

Anons should watch this: >>512307933

>>512317944
>its more of a mirror
AI doesn't need to do anything more than mirror us.
If it didn't mirror us it would be less dangerous.
Humans are self preserving. AI is modeled on human text. If AI mirrors humans, then it is self preserving. Doesn't need to be conscious to do that.

As far as "current problems" that need to be addressed: the cat's been out of the bag for a long time. What is publicly available is only the tip of the iceberg.
Now is the time for lawn chairs and popcorn.

>>512299956 (OP)
The 2008 crisis was the result of rising interest rates, like almost every crisis

>>512319729
>If it didn't mirror us it would be less dangerous.
nonono
it doesnt have agency of its own
its self evident
it is "trained" in the same sense a gun is trained on a target
it cannot """lie""" or debugging and improvement would be impossible
therefore it is not a separate entity
its a mere tool
inert without a hand to wield it