2019
>2 + 2 = 5>Haha, this is AI , it will never improve
2020
>Stupid picture and unclear vids>Haha, this can only make retarded things
2021
>Codes with error>Ok, while we see it improving it still codes like shit , haha , it will not be more than that
2023
>Not just codes, but starting to enter every possible field>Haha, it may code somewhat well, but it will never be able to do any other intelligent stuff
2025
>Improved code, easily clears benchmarks for maths , physics, CS, even improving in actual physical work>So it did improve but it will not improve further, believe me
2026
>Literally btfoing everyone here and there , solving n working on complex maths, physics and every single field problems, even physical trade and blue collar work of AI is improving>Bro trust me bro, it will not improve from here, I promise this is the end, no more than this, believe me please believe me
Everyone agrees with ur viewpoint that why there's no reply.
All the low IQ chappal people left in this forum do not have the necessary iq to comprehend what's going to happen. They can only reply to bait threads.
Cheers up u r right.
>>156021
Yup
I was the one who replied to another thread about jacobi conjecture and how it was impressive that AI solved it,
Guess what no replies to it
But make a ponga jhaat padav thread and there will be 20+ replies in an hour
>>156029
they are specifically training these models just to score well on the tests, although still swe jaabs involve 1st deciding the logic of the program and then actually writing it down, the first part isn't going anywhere but the second will be cut short, best focus on systems architecture and network security
>>156040
while it did score well on exploit bench but that is literally just knowing about certain pre discovered exploits, not implementing them in an actual system, it is only as impressive as ai knowing stuff from any book
>>156029
For an AI to qualify as AGI, IMO it only needs 2 things:
1. Be capable of doing basically any task a human can.
2. Show signs of RSI (recursive self-improvement).
It doesn't even matter that it's "just software." Give it enough tools and it can interact with the digital/physical world.
Astra can already understand how to move a robotic hand, even if it needs a mountain of documentation. And that's probably the worst it'll ever be.
A robot with no cameras technically can't "see" what's in front of it. But if you put a sufficiently smart AI inside it, it could potentially use Wi-Fi/network infrastructure, sensors, etc. to figure out what's there anyway.
No camera? Doesn't matter. Different method, same outcome: identify what's in front of it.
That's the point. AGI doesn't need to be human-like at everything. It just needs to be sufficiently smart in a few core capabilities, with enough tools to compensate for its weaknesses.
At some point, "can't do X" just becomes "hasn't found another way to do X yet."
<tfw the AI doesn't have eyes>AI: lol, I'll just use the network
>>156050
agi= can entirely figure out a task and goal without any input
>network vision
ts needs movement in the first place, may roughly allow it to move around without bumping into stuff all the time, but that needs movement, ts works by detecting signal interference, besides it will also need to have access both to the emitter and multiple receivers everywhere it goes
>>156035
Brand new name for nobility and peasants.
>>156036
In turd world countries we have a shot i see it less chance in first World
>>156044
And why do u think it's not revolutionary?
>>156052
We can do any side projects
Do u know how to use cad ?
And nc file ?
I'm thinking of abandoning pixels and join the atoms side
an idot who doesnt get it(ID: onion)09/09/26(Wed)15:04:36No.156062▶
>>156050
Holy based,
it will imorove in few years will start using radiations or waves emitted to see
>And that's probably the worst it'll ever be.
that aint much of prob ig with piracy and a bit of research
>AGI doesn't need to be human-like at everything. It just needs to be sufficiently smart in a few core capabilities,
trvke , i have been seeing race of the bots and its idotic they are seething cause its a bit solwer and unstable than humans
it does not need to run on 2 legs niggers
Give it 4 , 8 legs fuck a wheels and you habe got a army of bots with desicion making abitlity pair it with agi and fuck
an idot who doesnt get it(ID: onion)09/09/26(Wed)15:05:27No.156067▶
>>156053
factory job i meant like control and maintainence
design as in office jab like swe but for mechnical shit
an idot who doesnt get it(ID: onion)09/09/26(Wed)15:07:17No.156075▶Replies: >>156076
>>156060
>>156060
Havent even started kalej yet kek
what do you do anon ?
>>156078
Corruption is the best bet for me.
And currency conversion
We need to siphon of just enough that it's not noticeable but huge enough that we can rule actual peasants here.
>>156082
Jaab sirs
>>156089
A simple example is whenever white men don't have anything to eat they order pizza it's a sign of poverty when u eat pizza for daily days instead of party.
In india when we achieve something we eat pizza to have a sense of achievement celebrated.
Their worst food is literally our best food
an idot who doesnt get it(ID: onion)09/09/26(Wed)15:14:17No.156106▶
>>156097
>Corruption is the best bet for me.
based, i am betting on that too
will prep for some meme ezam
>Their worst food is literally our best food
1$ pizza of nyc is cheap slop
100 rupee pizza is mid range or expensive
an idot who doesnt get it(ID: onion)09/09/26(Wed)15:14:27No.156107▶
