#✦│chat

1 messages · Page 231 of 1

plain dew
#

Did you bring some pizza?

polar rain
#

good night again!

shell quail
#

hi pizza

plain dew
#

Wassup?

wind lion
shell quail
#

have a good sleep

wind lion
#

good night li king

lean ether
#

anyway, goodnight chat cat_disco

shell quail
lean ether
#

time to play game

wind lion
lean ether
shell quail
plain dew
shell quail
plain dew
lean ether
#

we're sharing same house

plain dew
lean ether
#

same planet

#

mars

lean ether
plain dew
lean ether
#

I want to do a tour only lol

wind lion
plain dew
shell quail
#

so it will take some time

shell quail
lean ether
plain dew
#

Lets sleep then 🥱

lean ether
shell quail
shell quail
plain dew
#

Goodnight people! I will see ya all tomorrow if i can. Lol

shell quail
#

how did u know

lean ether
shell quail
lean ether
#

ain't zoom chatting

plain dew
lean ether
shell quail
plain dew
shell quail
#

but still we are fam

lean ether
#

yes

plain dew
lean ether
#

anything just tell me

shell quail
shell quail
lean ether
old urchin
shell quail
primal jolt
#

@warped hound i thought you'd vote for me

#

Nvm it

#

Smhhh

vapid hull
#

?afk studying do not ping i repeat do not ping

#

:|

stark pecan
rough gale
#

Hell o wut's going on here guys

stark pecan
#

we are fighting

shell quail
#

Hii

sage drift
#

come on

#

ffs

solemn pond
#

e

old urchin
#

HI

#

dead chat fr

wind lion
#

Gn

gilded granite
#

Ic pizza

gilded granite
old urchin
#

1 am i still havent had dinner and shower :<

gilded granite
#

Damn

#

Go eat

#

And then sleep

old urchin
#

where i can buy food now

#

😭

stark pecan
#

dont u have 24/7s

#

i get food at 3am

old urchin
#

I dont wanna walk 200m

wind lion
#

Order

stark pecan
shell quail
#

Helloo

old urchin
stark pecan
#

i was asking what kind of area... whatever

old urchin
#

Oh

stark pecan
#

ig that makes sense

#

dunno bout thailand

#

but maybe there's not much 24/7 stores in ur locals

old urchin
#

Ig wait for tmr

stark pecan
wind lion
#

Did you just leak your address

old urchin
#

Oh

wind lion
#

in a 500k server

old urchin
#

😔

wind lion
#

Be careful 😭😭

old urchin
#

Ok ty

stark pecan
#

that is crazy

old urchin
stark pecan
#

how are you from

stark pecan
#

what is this question 😭

primal jolt
#

No Shut up

#

I am tired

#

<@&1159293140440723499> get his ass

open moss
#

can you convert .pth voice models to .onnx or does it corrupt?

unreal juniper
winter void
#

Hi I'm a new AI user, I'm glad to join this community

plain dew
#

@opaque palm PLS UNBAN 😭 💔(ノへ ̄、)🥺

#

@opaque palm i just met a friend there pls

#

🥺

#

😢

#

@opaque palm pls...

primal jolt
#

Welcome to the server

winter void
#

Thanks !!

primal jolt
bitter hawk
#

Wtf is happening

sage drift
polar depot
#

hi

stark pecan
sage drift
stark pecan
#

he looks like a trusted adult

bronze plover
#

hi

bitter hawk
stable knoll
#

Where are you from

#

KNEW IT

#

I’m Cypriot

#

No

#

I won’t join your server

#

And vote for you Sammy

#

No?

#

Well either way I don’t give a rats ass

#

Tell your boyfriend to do it xei

#

@slender swift

#

Help your girl

#

Why are you deleting your messages 😭

#

You are weird

#

BALRIGHT

#

@slender swift she is for the streets

primal jolt
#

...

stable knoll
#

What a liar

primal jolt
primal jolt
#

I deleted the message from your dms sorry for that

primal jolt
#

You added me earlier that's why I texted
Mb for that

stable knoll
#

Why are you keep deleting

#

😭

primal jolt
primal jolt
#

Shouldnt have said it

primal jolt
#

Bye

#

Everyone

lucid ferry
#

live laugh love carti

#

w flight gifs btw

stark pecan
#

i cant quite put my finger on it

primal jolt
#

Mb

stark pecan
#

but u rub me the wrong way

primal jolt
#

..

#

Never heard it before

stark pecan
lucid ferry
#

how you been shrimpy

stark pecan
#

i have been better

#

but good enough

lucid ferry
#

so glad to hear

stark pecan
#

hows you

lucid ferry
#

decent

#

college classes started this week but we pulling through anime_pray

stark pecan
#

shucks

#

as for me, i just resigned from my job

#

feels amazing

#

always wanted to quit

lucid ferry
#

reminds me, planning to do the same

#

management sucks now and it’s spiraling down that path

#

just need a backup before i can actually put in my 2 weeks

stark pecan
#

sucks that i cant just quit either

#

been a rough patch these past 3 months

#

atleast im free now

lucid ferry
#

relax and enjoy the time before you start working harder again yk yk

stark pecan
#

west ppl hate the 2 weeks notice stuff but man, i really wish i only needed a 2 weeks

#

i think im gonna freelance for now lel

modern notch
lucid ferry
#

if it was easy to just straight up say fuck you to my higher up i would okay😭

modern notch
#

I am also ready to quit mine

#

I need a job in my field with more hours

lucid ferry
#

with me i have to either stay a bit longer before i can get experience in the field i’m in, or have a different job entirely

#

or just get connections and hopefully work somewhere else in the same field

#

though that’s unlikely

modern notch
#

My issue is I have a degree now and no one is hiring

stark pecan
#

i got not much options since i didnt graduate highschool

modern notch
#

U should go back to school if u can

lucid ferry
#

unfortunately

modern notch
#

I’m mad

stark pecan
modern notch
#

I have been looking for another job for 6 months now

lucid ferry
#

especially if people major in stuff like computer science or other things as such

stark pecan
#

graphics designing and working with tech in general

stark pecan
#

the demand is kinda high here

lucid ferry
modern notch
#

I need someone to hire me

#

Plz

#

I need another job

stark pecan
#

are u content with that?

lucid ferry
#

where have you applied to like any big stuff?

#

big companies

stark pecan
#

i dont think i ever wanna do corporate or retail again

modern notch
#

Ya I have through job sites and my college

lucid ferry
#

retail was gonna be one of my backups but then i just thought about it since ik it’s hell

lucid ferry
modern notch
#

0

#

Absolutely none

stark pecan
#

i hate interviews so mucu

#

this one interview i had was kind of a losing battle cuz the person working there somehow knew me

lucid ferry
#

evb be needing people but never give the opportunities

#

it’s stupid

#

could have a giant help wanted poster and never even bother giving an interview

stark pecan
#

i would love to do cosplay full time

#

or atleast

#

part time to the point i can solely rely on it

lucid ferry
#

are there jobs that actually have that

#

other than maid cafes

#

or mascot stuff

stark pecan
#

yeah its just mostly working for those stuff

#

and then there's the cosplayer for hire stuffs

#

which we either need to look out for or be popular enough that ppl reach out to us

#

i got a few friends who hit me up for those gigs so ig there's that

#

it pays to have a connection

lucid ferry
#

word

#

glad to see you found something at least

#

especially something you love to do

stark pecan
#

yeah I'm really glad

#

get to live my childhood

floral shadow
#

.

wary hamlet
#

hello

sage drift
#

hmmmm

#

gnight aihub

rocky leaf
#

hii

steel geyser
#

hey guys

#

How are we all feeling today

low patrol
#

What happened to astralabs?

steel geyser
abstract arrow
steel geyser
#

how do you think you can make money, maybe selling tupac songs?

abstract arrow
steel geyser
abstract arrow
#

i might make songs he wrote for other people n use his voice

steel geyser
#

ai is pretty illegal itself in some ways but its in a grey area right now since its new tech

#

you can get away with a lot

#

do you profit with ai often?

abstract arrow
#

naw

#

sort of do it for fun, i was in a alot of hip hop online blogs

analog holly
#

damn we hit a bottle neck

#

we’re not getting exponential growth

#

the iPhone 17 moment

steel geyser
analog holly
#

same

#

but we haven’t really seen anything innovative

#

both are bottlenecked

#

I think we need to transition

#

mobile AI

steel geyser
#

in the schools too

#

ai is a better teacher then most

analog holly
#

Shit can make u stupid too though

steel geyser
#

ai is great for our society

#

we should have ai fast food employees

#

they never get tired and never stop working

analog holly
#

ya

steel geyser
#

it will be billions more for companies

analog holly
#

except they’re incapable right now multitasking

steel geyser
#

and you dont have to pay them

steel geyser
analog holly
#

nothing’s ever free, bro

steel geyser
#

slow takeover

steel geyser
analog holly
#

what jobs is it gonna do?

#

it’s good at tasks, but I don’t know if it’s good at full jobs

steel geyser
#

everything

analog holly
#

bro, have you used shit?

steel geyser
#

it can replace most tech support jobs i mean it already has

analog holly
#

It’s good for a little bit, but not full-blown

steel geyser
#

replace accountants too

#

replace artists

#

replace editors

analog holly
#

I don’t know about all of that, dude

steel geyser
#

its good for our society

#

we dont need them anyway

analog holly
#

I think the labor of work just got shifted

#

somebody has a supervisor this shit

steel geyser
#

everyone can be plumbers and they will make more and be happier

analog holly
#

plumbers actually made good money

steel geyser
analog holly
#

I wish I was true I wanna believe it

steel geyser
analog holly
#

but I’m just not seeing it with the current capabilities lol

steel geyser
#

not currently but it is evolving exponentially

analog holly
#

that’s what I’ve seen earlier. We hit a bottleneck.

#

we’re not seeing that exponential growth we thought we would

steel geyser
#

when we get AGI it will take over most knowledge based jobs

analog holly
#

ChatGPT five was two years in the making

steel geyser
#

once we get self evolving ai its over

analog holly
#

Oh geez lol now how are we gonna do that?

steel geyser
#

it can already code

#

it knows more then any human on earth

#

once we run out of human made data we get ai to make its own stuff to train off of

#

most companies are working towards it

misty lance
steel geyser
misty lance
#

currently LLM is basically an Auto Complete on steroids with a MASSIVE databse

steel geyser
#

it can code and script

misty lance
#

no, it cannot lmao

#

which is why AI is not writing AI code

steel geyser
misty lance
steel geyser
#

you clearly aren't a real ai enthusiast

misty lance
#

it is generating tokens based upon a massive training database

analog holly
#

we can’t even figure out P verse NP

misty lance
#

it is essentially auto complete on massive steroids with a MASSIVE database

steel geyser
#

thats why companies are trying to create agi

misty lance
#

which is why if you ever push LLM to its boundaries by talking about something that it has poor training on, it will answer like it's an authority on it.. but you find out that half or more of what it generated s is false (but sounds true)

#

we have to create /mimic our brains and do it at a massive performant scale . that will be true "AGI"

steel geyser
#

do you guys think prompting is harder then what "real artists" do like drawing and writing and stuff

misty lance
#

but i think wemight a godo 5-10-20 years away from that

steel geyser
#

i think it is

analog holly
#

AI fanaticism and AI doubt are huge factors that plan to the industry they’re like plagues

misty lance
#

it might be more frustrating because the AI doesn't do what you want it to do lmao,but its not "more difficult" than human generated art

steel geyser
#

ai art is true art

#

we are directors just like a movie director

#

we deserve credit and fame just as much as them

misty lance
#

AI art is basicallycopying and pasting art that man has already created and trained it on

#

and applying wierd formulatic equiv of photoshop filters

#

which is why

analog holly
#

NP stands for “nondeterministic polynomial time.” It is the class of decision problems for which a given solution can be verified in polynomial time by a deterministic Turing machine.

#

Examples of NP problems include the traveling salesman problem, Boolean satisfiability (SAT), and graph coloring. Some of these problems are NP-complete, meaning they are both in NP and at least as hard as any other problem in NP.

misty lance
#

the poster in the 3rd child is messed up

#

the writing in the hat is messed up

analog holly
#

P stands for “polynomial time.” It is the class of decision problems that can be solved efficiently in polynomial time by a deterministic Turing machine.

In other words, if a problem belongs to P, there exists an algorithm that can both find the solution and verify it in a reasonable (polynomial) amount of time with respect to the input size. Examples include sorting numbers, finding the greatest common divisor, or determining if a graph is connected.

misty lance
#

and if you look at his face/skin its also messed up

steel geyser
analog holly
#

There are computing limits to learning

steel geyser
#

do you guys know any really active ai servers for fellow gentlesirs like myself

analog holly
#

Even with quantum computing

misty lance
steel geyser
#

this server is so quiet

analog holly
#

For example, if you had 30 million pieces of puzzles

#

How would the AI solve that?

#

Even with quantum computing, this would be extremely difficult

#

Possibly even impossible

lofty flint
misty lance
#

qwen edit

#

i believe

#

either that or Kontext

analog holly
#

P in the context of AI and its limitations of learning refers to the set of problems that can be solved in polynomial time, meaning they can be handled efficiently

misty lance
analog holly
#

NP in the context of AI and its limitations of learning refers to the set of problems for which a proposed solution can be checked quickly in polynomial time, but finding that solution may be extremely difficult and could require exponential time. For AI this means that many real world challenges such as planning, reasoning, or optimization fall into NP where the system can test if an answer is correct but may struggle to actually discover the answer efficiently. This limitation forces AI to rely on heuristics, approximations, and learning patterns from data rather than guaranteed optimal solutions because exact algorithms are often computationally infeasible for NP problems.

civic radish
#

op

analog holly
#

There are computer science limitations to learning

#

Even with highly advanced machine, machines and data centers

misty lance
#

Basically, we need to be able to mimic the brain and how it thinks and learns and processes

analog holly
#

The efficient genome assembly problem

misty lance
#

LLM is the first step - it has been modeled basically sort of how our brain keeps track of memories/inforamtion (but the LLM can't learn )

#

but it does not actually think how we do. we have to train it on "how it thinks" so it can generate tokens that align with thinking patterns that we can train it on but it can't actually think

steel geyser
#

nerds

misty lance
#

which is why sometimes if you follow deep LLM - you'll find that for some models and some problems actually get better answers by disabling thinking

analog holly
#

The efficient genome assembly problem shows AI’s limits because assembling a full genome from many short DNA fragments is like solving an NP hard puzzle. AI can help approximate solutions and speed up error correction or pattern matching, but it cannot guarantee fast optimal solutions since verifying an assembly is easy while finding the correct one is computationally difficult. This makes genome assembly an example of how AI must rely on heuristics and approximations when problems go beyond P.

unreal juniper
#

**Question of the Day ❓ **


What's one thing you wish AI could do that it can't do yet? 🙏

-# <@&1344706968417472563>

misty lance
#

(everything?)_ lmao

analog holly
unreal juniper
proven geyser
#

work for me

lyric goblet
#

да

lyric goblet
#

пизда

unreal juniper
#

@lyric goblet @lavish acorn be sure to keep the chat in english please

wispy vector
#

Hello extraterrestrials 👽

analog holly
#

You are talking about the Unique Games Conjecture. It is a major open problem in computational complexity theory introduced by Subhash Khot. The conjecture states that for a certain type of constraint satisfaction problem called a unique game, it is computationally hard to even approximate the best solution within some factor.

A unique game is set up so that each constraint between two variables specifies a unique matching of values between them. The conjecture proposes that deciding whether almost all constraints can be satisfied or whether only a small fraction can be satisfied is NP hard.

unreal juniper
#

@lavish acorn please, speak only english here

analog holly
#

I’m just stating the limitations

unreal juniper
analog holly
#

In computer science. Which are gonna limit AI

unreal juniper
unreal juniper
#

everything is crazy

misty lance
#

IKR

unreal juniper
misty lance
#

AI would prolly evict us from planet earth 😦

misty lance
lofty trench
analog holly
#

Genome assembly shows that limitations aren’t just about algorithms being “smart enough” they come from the hard limits of computers and data. If the data doesn’t contain enough unique information, no amount of computing power or AI can solve the reconstruction perfectly.

unreal juniper
unreal juniper
analog holly
#

Genome assembly proves that even with unlimited computation AI cannot solve problems when the data itself lacks enough unique information. This shows a ceiling on what superintelligence could achieve since many real world problems share the same limits of incomplete or insufficient data.

#

Reconstructing a shredded document when many strips are identical
Recovering an audio signal when parts are missing and the gaps are longer than the unique sounds around them
Trying to rebuild a 3D object from shadows when different objects cast the same outline
Decoding a text where large sections are repeated word for word and the breaks are lost
Restoring a corrupted hard drive when the repeated sectors look the same and no markers distinguish them

#

AI and computers are limited because they cannot recover information that isn’t in the data, meaning some problems will always remain unsolvable no matter how powerful the system is.

stark pecan
#

but what if ai gets eaten by a whale?

#

would it be able to solve it then?

analog holly
#

No

stark pecan
#

hmm

#

interesting

analog holly
#

Without enough observable information about what happened inside the whale no computer or AI could perfectly reconstruct the event.

stark pecan
#

what if it got eaten by a llama tho

analog holly
#

Same thing

#

At best it could model possibilities based on knowledge of whale biology or llama digestion and prior cases but it could never reach absolute certainty unless more direct data were available.

analog holly
#

The halting problem is a classic result in computer science that shows there is no general algorithm that can decide for every possible program and input whether that program will eventually stop running or continue forever.

#

Alan Turing proved this in 1936 by showing that any supposed universal halting tester leads to a logical contradiction when asked to analyze its own behavior.

#

This matters because it demonstrates a hard boundary on what computers can ever achieve. No matter how advanced, they cannot perfectly predict the outcome of all possible computations.

#

Too many factors

#

Think about something simple, like dropping a glass cup on the floor

#

For a human, we instantly recognize and we know what it is

#

A computer would need to run it through a bunch of data and algorithms in order to understand something so simple as the glass falling and breaking

#

For humans, we do this and recognize it subconsciously without even thinking we already know what it is and understand

#

Or sneaking up behind somebody to try to scare and startle when they least expect it

#

How do you think an AI would react?

analog holly
#

For humans the response is immediate without any conscious calculation. An AI on the other hand has no instinct; it only processes inputs according to programmed models. To react it must analyze sensor data, compare it against patterns it has learned, and then decide on an action.

plain dew
#

Finally someone else who understands.

Last time I brought that up, people brushed me off as a mad man.

stark pecan
#

felt the guilt of his creation

#

and he made an AI double of himself, a carbon copy

#

but stripped that clone of all human morality and emotions

#

and he expelled himself to the moon

#

while the clone lives among ppl

#

to figure out things for itself

analog holly
#

There’s two theories on this

stark pecan
#

and the clone starts to become its own person

#

having its own hobbies

#

likes and dislikes

#

and since its a 1:1 copy of That Man

#

it knows the things he's done

#

everything, every passing second

analog holly
#

Verifying program correctness means proving that a program does exactly what it is supposed to do under all possible conditions. Instead of just testing with sample inputs, verification tries to show mathematically that the code always meets its specification. This involves using formal methods like logic, proofs, or model checking to ensure there are no hidden errors, crashes, or unintended behaviors. It is hard because programs can have infinitely many possible states and inputs, making complete verification often impossible for complex systems.

stark pecan
#

except for the guilt and emotion

#

and the clone ultimately decides to surrender himself to the government

#

asking for the president's and the governments protection

silent current
stark pecan
#

protection, not for him

#

but for the tome...

silk forge
#

Is applio Google colab?

stark pecan
#

the book that possess all his research...

analog holly
#

A non-von Neumann architecture is any computer design that does not follow the classic von Neumann model where instructions and data share the same memory and are processed sequentially by a single control unit. y, parallelism, and problem-specific performance beyond what a standard sequential model allows.

#

In non-von Neumann systems the goal is to overcome bottlenecks like limited parallelism and the “von Neumann bottleneck” of moving data back and forth between CPU and memory.

stark pecan
#

the the man, the orignal creator is on the moon and starts a podcast

analog holly
#

many operations at once, neural networks that mimic brain-like distributed processing, and quantum computers that rely on quantum states rather than binary logic. These designs try to achieve higher efficiency

#

But there are also limits to quantum states

#

Because input could be very similar, and then you go back to the halting problem

#

Like these are barriers that are extremely hard to overcome

#

If they’re even possible at all

#

But would be required for a super artificial system

#

They’re literal hard limits

silent current
#

as well as the artists

analog holly
#

The Quantum PCP Conjecture shows some problems stay globally hard to verify even with quantum power, proving that AI and superintelligence will always face fundamental limits no matter how advanced they get.

#

The Quantum PCP Conjecture is an open problem in theoretical computer science. It asks whether checking properties of a quantum system can be done by looking at only small, local parts of it, similar to the classical PCP theorem for proofs. If true, it would mean even approximating solutions to certain quantum problems is just as hard as solving them exactly. This would confirm that some problems remain intractable even with quantum computers, marking a deep limit of computation itself.

#

PCP stands for Probabilistically Checkable Proofs. It is a concept in computational complexity that shows a proof can be written in such a way that a verifier only needs to check a few random bits of it to be convinced it’s correct with high probability.

The PCP Theorem says every mathematical proof can be transformed into this form. Its importance is that it links to hardness of approximation: if a problem has a PCP proof, then even approximating its solution is as hard as solving it exactly. This is one of the core reasons why certain optimization problems are provably hard.

silent current
#

quantum computing is still a decade ahead to be practically viable

analog holly
#

Imagine a puzzle with 30 million pieces

#

How hard that would be for human to put together

stark pecan
#

i think itd be pretty hard i think

fading peak
#

putting it together is easy

stark pecan
#

just my onion

fading peak
#

putting it together correctly is hard

analog holly
#

sheer size of the search space means the AI must use heuristics and approximations it can’t test every possible arrangement, because the number of combinations grows astronomically. If the puzzle has repeating patterns or pieces that look identical, the AI will still run into the same kind of bottleneck as genome assembly or PCP theory: the data doesn’t give enough unique information to guarantee a perfect solution without extra constraints.

stark pecan
#

hmm but what if we put the ai in a schrodigers box

analog holly
#

They start running into a paradox

silk forge
#

Is there a way to use applio on mobile

analog holly
#

exactly. These are what you could call hard limits not just engineering problems. Super artificial intelligence would still run into them because they are baked into mathematics and information theory.

stark pecan
#

that way we wont know if it has solved the puzzle or not unless we look inside it, meaning it solved yet hasnt solved the puzzle at the same time

analog holly
#

It just mathematically impossible

#

Like the bounds of physics itself lol

stark pecan
#

but also what if all the pieces of the puzzle have all been replaced through years, one by one. So is the puzzle actually the same puzzle as before? the puzzle of theseus

analog holly
#

Plus, we have the existence of one-way functions

#

A one way function is a function that is easy to compute in one direction but extremely hard to reverse. For example multiplying two large prime numbers is easy, but given only the product it is very hard to figure out the original primes. This property is central to cryptography because it allows secure encryption, digital signatures, and authentication. The “hard to reverse” part is not absolute proof but based on the assumption that no efficient algorithm exists to invert the function.

stark pecan
#

and what if sisyphus tried to put the puzzle together himself, but he cant get it right and has to start over when he's starting to see progress

analog holly
#

Taking shredded paper and running it through a shredder is easy, but reconstructing the exact original document from the strips is extremely hard. Cooking an egg is easy, but returning it back to its raw uncooked state is practically impossible. Blending fruit into a smoothie is easy, but separating it back into the original whole fruits is not realistically possible. Encrypting a message with a secure algorithm is easy, but decrypting it without the key is computationally infeasible.

stark pecan
#

putting him in an enternal process of completing the puzzle but never completing it

unreal juniper
stark pecan
#

is sisyphus happy?

unreal juniper
#

and that's prob the most irrealistic since it just predicts text

analog holly
#

even if an AI became conscious, it would still run into situations where the data and math themselves force limits. Those limits would act like built-in deficiencies, where no amount of intelligence can push past the lack of information or the exponential explosion of possibilities.

#

Mean that they would start developing mental problems like we seen people

#

Autism and going crazy and all that

stark pecan
#

damn

analog holly
#

Ya

stark pecan
#

first in my bloodline to see this

analog holly
#

I’m not saying that these things can’t be solved or resolve

#

But we have to be realistic with the current limitations of not just AI but computers themselves

#

These are extremely difficult hurdles for some of the best minds in the world to even get a clue of how to solve

#

Some of these have bounties on them

#

If anybody ever figures them out, you can get paid up to $1 million to figure out solutions to some of these problems

#

Even win a noble piece prize lol

#

I don’t think people realize the share amount of data and data centers needed for AGI

#

That’s so much information to process

#

Think about the traveling salesman problem

#

The traveling salesman problem asks this: given a list of cities and the distances between them, what is the shortest possible route that visits each city once and then returns to the starting point. It looks simple but the number of possible routes grows explosively as more cities are added. Finding the exact shortest path is NP hard, meaning no efficient algorithm is known to solve it for large cases. This problem is used as a benchmark for optimization, logistics, and to illustrate the limits of computation.

#

And the larger that list goes the harder it is to resolve

#

Adding more data or scaling up doesn’t remove the core difficulty it makes it worse. In the traveling salesman problem, every new city multiplies the number of possible routes dramatically. For AI, more data means more combinations, more states to evaluate, and more chances that the search space explodes beyond what can be handled. Scaling helps AI handle bigger cases faster, but it does not change the fact that some problems grow so quickly that no amount of data or computation makes them tractable.

#

Maybe these aren’t required for AGI

#

But anything further than that would be extremely complex and hard to solve

#

Maybe even impossible

#

We could prove this mathematically

#

The Goldbach Conjecture says every even number greater than 2 can be written as the sum of two primes. Nobody has ever proved it or disproved it, even though it holds for all tested numbers.

stark pecan
#

this is like putting on a 3 hour video essay from a channel with 12k subs by a random white boy talking about something he's incredibly passionate about and you listen to it while you work

analog holly
#

Well, I think there’s a lot of misconceptions in the AI community

#

A lot of bullshit floating around lol

stark pecan
#

funnilly enough, im styling a wig rn

analog holly
#

They’re hard to talk about because nobody’s really interested in them

#

The AI community is suffering from AI fanaticism I think

#

people in the community start to believe that if you just throw more data and bigger machines at the problem you can solve anything. From my view that is not true. There are problems that do not get easier with size or speed because the limits are built into the math itself. Things like the halting problem or the traveling salesman problem show that no matter how powerful the system is some tasks stay out of reach. What I mean is the belief in AI as an all solving force is exaggerated and ignores the real boundaries of computation.

#

OpenAI describes AGI not by technical breakthroughs or computational ability but by the amount of revenue it can generate, setting the benchmark at around 100 billion in profit.

#

This is the legal definition that they have written on paper and contract with Microsoft

#

“profits large enough to satisfy investors, set at a benchmark above $100 billion. Reaching that milestone would trigger separate investor-related terms in the contract. “

#

By that legal definition AGI is tied to financial output rather than pure capability. It means what gets labeled AGI in the Microsoft OpenAI contract could end up being an advanced version of today’s AI scaled up to generate more than 100 billion in profit rather than a truly general intelligence in the technical sense. In practice this turns AGI into a business milestone instead of a scientific breakthrough, so the first systems called AGI under that deal might look a lot like current AI optimized for revenue at massive scale.

#

At lower levels of progress you get big leaps from relatively modest resources, but as systems scale the returns shrink. Each new breakthrough requires far more data, larger models, and massive compute budgets, while the actual gain compared to the cost gets smaller. The higher you go the curve flattens

#

And that’s what we’re seeing with ChatGPT5

#

The cost of tokens is gonna go up

lofty flint
#

Guess I need to rev it up

analog holly
#

For sure

abstract arrow
#

I made this beat sounds real lol

analog holly
#

All the money is in the API and enterprise

#

So use up as much as you can lol

#

Have you guys ever jailbroken AI with images? It’s the funniest thing.

#

Stuff like this

stark pecan
#

why does ai always put this putrid mexican filter on top of everything

analog holly
#

And then you could ask it to generate you images of so hilarious

#

😂😂😂

stark pecan
#

hilarious

analog holly
#

Does anybody know how to login with agent mode?

#

But I can’t seem to login with the other one

#

😭

static walrus
#

dude someone called fuckin w-okada voice ai 😭

old urchin
#

Hi

stark pecan
#

Fi

misty schooner
slender swift
bitter hawk
#

Guuuyyyysss

#

Be civil please

#

Lets not get bannn hammar envolved please

slender swift
mental fulcrum
#

hello

bitter hawk
bitter hawk
#

This is a verbal warning @stable knoll

mental fulcrum
#

Hey, I was wondering if this server might be interested in adding a new AI-generated bot. It’s completely free. The bot can turn text prompts into short anime clips with voiceover.

static walrus
bitter hawk
#

Who knows

bitter hawk
silent current
bitter hawk
#

When was that

static walrus
#

also lusbert bot

static walrus
#

before u joined

#

nvm

#

u already joined

wind lion
halcyon lagoon
#

guys, is there any good website for creating image to video AI?
I used Hailuo but it didnt give me any more credit 😭

vapid hull
primal jolt
#

Of my feelings

#

IDKKK

#

good morning everyone I hope everyone has a great day today/ nice evening DANCE

polar rain
#

good morning

wind lion
#

Morning

analog holly
mental fulcrum
mental fulcrum
severe crest
primal jolt
#

Or for thinking

lofty flint
#

Stop thinking is bad for ya

primal jolt
old urchin
#

hi

polar rain
#

good night eddy

old urchin
polar rain
#

hi helloo

cloud hawk
#

who pinged me >:(

#

oh wait chat revival nvm

static walrus
#

Portal: Java Edition is a Minecraft project (created using resource and data packs only - no mods!), which is dedicated to recreating Portal 2's gameplay and mechanics inside Minecraft.
The mod releases on the 22nd August 2025.

The author of this mappack is BurningCakeDoesAche.

Huge thanks to the development team for providing me with early sh...

▶ Play video
primal jolt
cloud hawk
static walrus
cloud hawk
#

tbf and this isnt to downplay anything theyve done at all but like

#

it seems like itd be relatively easy to make apart from the cubes you use with the portal guns

static walrus
#

how the fu-

#

ITS JUST VOICE LINES

#

what are they gonna do to it 😭

exotic pawn
#

ok

cloud hawk
primal jolt
#

Uhhhh

#

So we are not talking today?

old urchin
cloud hawk
#

when im awake at 6am and wanna start a conversation

#

then randomly get the urge to continue at like 7pm

#

then maybe again at midnight

#

timezones and people make that awkward

polar rain
#

hello guys

polar rain
#

i've been online since earlier, i thought that this is not how it is usually

#

despite the timezones, i usually see people chatting in here at this time

#

so maybe it's just a coincidence that most people are busy at this time

primal jolt
#

I'm soo happy

#

like i did for now

#

I have to go back to it but mannn

#

I'm sooo happy

polar rain
#

ooh congratulations then!

primal jolt
primal jolt
#

TYYYYY!!!!

polar rain
primal jolt
#

Yesssss sirrrr

polar rain
primal jolt
shell quail
polar rain
#

hi zen!

#

how are you doing?

shell quail
#

wbuu

polar rain
#

i'm doing well

polar rain
#

it's just that i'm quite struggling since our internet connection is pretty slow today

#

i hate it

shell quail
#

sometimes my net acts drunk

#

😭

polar rain
#

most of the time it's okay

#

sometimes it's slow, rarely it's fast

shell quail
#

my net acts like that too

#

sometimes

#

and its so annoying

polar rain
#

it's annoying

shell quail
#

😭

#

ikrrr

polar rain
#

right, you get it

cloud hawk
#

why the hell is every community band leader called Lee

polar rain
#

the new secret garden decors are good!

#

the bundles

shell quail
#

maybe u are from mars too

shell quail
#

yesterdayyy

#

so pretty

polar rain
#

right, but it's expensive for me

#

it's almost 10$

#

at least for me

primal jolt
polar rain
#

ohh, i don't have a phone..

primal jolt
#

But bundle can't be gifted

polar rain
#

sam i sent you a friend request too!!

primal jolt
primal jolt
polar rain
#

that's sad

#

bundles could've been a good gift

primal jolt
#

Like it's less than 4 dollars i think

polar rain
#

woahh

#

oh yeah i see

#

i checked it, only the solos has the gift option

primal jolt
#

Yesss

polar rain
#

it's my first time seeing that this is how it actually looks like to gift someone

primal jolt
#

Lmaooo

#

I never gifted anyone either

polar rain
cloud hawk
#

light mode is crazy

primal jolt
cloud hawk
#

but fair

primal jolt
#

Sookn

polar rain
#

hopefully uhm..

primal jolt
polar rain
primal jolt
polar rain
#

i used to hate it too

shell quail
polar rain
#

but when i had my first nitro boost back then, i liked the themes on lighter modes

cloud hawk
#

also took me way too long to realise that was the peso currency

#

spent a good minute trying to think "what is that symbol for"

polar rain
#

so when it ended, it switched up to a normal lightmode

polar rain
primal jolt
#

I hope you get one soonn

#

😭

polar rain
#

so when i chat here, i only use my pc

polar rain
#

i sometimes miss just scrolling down on my phone tho...

primal jolt
polar rain
#

then i'll end up falling asleep

primal jolt
#

Hnmm

#

🍀

slender swift
#

Samm see

polar rain
#

the game quest still hasn't been installedd

slender swift
#

Heree too

polar rain
#

lmaoo

slender swift
slender swift
primal jolt
polar rain
#

estimated time remaining being a whole day is crazy

slender swift
polar rain
#

hopefully i can still complete this quest

primal jolt
polar rain
#

cuz that 700 orbs is still 700 orbs..

primal jolt
primal jolt
#

$ 20 is $20

polar rain
#

right, as someone who lacks money, those orbs are a good way to buy decors

slender swift
#

Yeppp

polar rain
#

have you finished the quest, prem?

cloud hawk
#

crazy they called a womens fortnite tournament the milk cup or whatever

primal jolt
polar rain
primal jolt
#

@slender swift you did ryt??

polar rain
#

was it easy to run?

primal jolt
polar rain
#

i'm afraid this might end up not running on my pc

#

i need to at least run it for 15 mins

primal jolt
#

Prem does you ask him

#

@slender swift TELLL HIMMM

polar rain
#

tell me tell me t-t-t-t-t tell me

primal jolt
polar rain
#

the last time i did the quest, despite having troubles with the graphics, i was able to play it a bit

polar rain
#

this is 90 GB, the last time was just around 29-30 GB.. uhm...

#

hopefully it will..

primal jolt
#

🙏🏻🍀

#

I have no idea about it

polar rain
#

we still have our last pack of buldak!

#

i'll eat it

#

i'll be back later

primal jolt
#

Eat well!!!

lusty quarry
#

Hi to everyone. 👋 I look forward to every new input.

primal jolt
#

Welcome to the server

cloud rock
primal jolt
#

I want a real person

#

Nice pfp

#

I watched that anime W ngl

cloud rock
#

i want to but im just jobless !!!

primal jolt
#

Welcome to the server btwwww

cloud rock
#

bro i joined almost month ago

primal jolt
cloud rock
#

now im active 😭

primal jolt
primal jolt
cloud rock
#

is there anything i can do ?

primal jolt
cloud rock
primal jolt
#

I'm not sure

#

I just tell people to be here and talk to me

#

Cuz I'm really shy irl

#

So i don't have many people around me anymore

#

So it's fine to atleast talk here

cloud rock
#

ohh that aint a issue tbh

primal jolt
#

I think soo

cloud rock
#

im so like you irl

primal jolt
#

It's not like I'm hurting someone

primal jolt
#

Do you know even our admin here is like uss

#

Not much social

cloud rock
primal jolt
#

You changed colour TH_yay

primal jolt
#

He ain't a social person irl

#

That's what I remember

cloud rock
#

ohh

#

well i hope id be of any help here 😉

primal jolt
#

Yeppss

primal jolt
#

What help you can doon

#

Tell me

cloud rock
#

anything is fine man

primal jolt
#

Hmmm

cloud rock
#

tbh im so jobless college guy

primal jolt
#

I was soo you 2 months ago

cloud rock
primal jolt
#

Yea but

#

Yeppss

primal jolt
#

If you wanna do it

cloud rock
#

oki!

primal jolt
#

Man i love this emoji

old urchin
#

No way someone make poll asking femboy dih size💀

primal jolt
#

How the fuck someone has 56

#

Crazy 😧

old urchin
#

Thai ladyboy xd

primal jolt
#

Ohh okiee

static walrus
slender swift
primal jolt
slender swift
primal jolt
slender swift
primal jolt
#

Okie

north sparrow
#

Hi

shell quail
#

welcome to the server

young pecan
#

Anyone who can make a dc server for free dm me

cloud hawk
wind lion
#

No

cloud hawk
#

yep i was waiting for someone to do that

wind lion
#

Aditya basu bad boy

severe crest
cloud hawk
#

oh we're allowed to do that

#

okayy

young pecan
#

Thx

wind lion
cloud hawk
#

talking about earlier

#

with the discord server thing

wind lion
#

Alr

old urchin
severe crest
#

@wintry trench just because you have disabled dms

Promotions of any kind is not allowed within the server, repeated violations of same kind might result in permanent ban.

young pecan
#

Can smb help me with a dc server ill give admin

#

I need commands

#

And i need 1 mod

wind lion
young pecan
#

Server is new

#

Created rn

#

So only me and 1 bot

#

But i have created all roles and all chanells aleeady

static walrus
#

why ask here o- o when this is literally has 500k members here

young pecan
#

Bc i think i can find proffesionalists here

static walrus
young pecan
young pecan
#

Ohhhhh

cloud hawk
#

apparently its level 5 now

static walrus
#

well that's boring

cloud hawk
#

my celeste gaming ass just thought that was a winged berry

primal jolt
stable knoll
humble mirage
#

Hi there I am a live streamer on twitch. I am roleplaying as a mafia person and want a real time voice as as a mob boss.

Is there a real time AI voice on voicemod replicate a mobster

If there is i will get the pro version today

Could some1 plz help with this requests 🙏

finite mist
#

which app can i use for voice changer free

static walrus
#

this text channel is not ai help

cerulean depot
#

ello

lean ether
#

hi

unreal juniper
plain dew
unreal juniper
unreal juniper
primal jolt
#

I'm soo eepy

polar rain
#

samm

#

i have a nameplate now

#

hann gifted mee

plain dew
plain dew
polar rain
plain dew
#

Hahaha im laughing

plain dew
polar rain
#

aahh alright

#

you ain't sharing that money anyway

lean ether
#

liking please delete ss

#

I don't want more people to ask for nitros

#

I can't afford

polar rain
#

oh wait alright

plain dew
polar rain
#

but they won't do it anyway

plain dew
polar rain
#

did someone ask you for a nitro? 😭

lean ether
#

many times

plain dew
plain dew
lean ether
#

some even offer something in return iykyk

plain dew
#

Sound rich too

#

That's why

polar rain
#

aaahh

lean ether
#

looking and being rich is a different things

plain dew
lean ether
#

mostly rich people dress average

plain dew
polar rain
#

looking rich and being rich at the same time is good

plain dew
plain dew
lean ether
plain dew
plain dew
polar rain
#

did mxiee already sent your nintendo prize?

lean ether
plain dew
polar rain
plain dew
polar rain
#

i forgot her surname

#

but i saw her on TV

plain dew
lean ether
polar rain
#

she's like one of the people that defines looking and being rich at the same time

plain dew