programming

[warped narwhal]

yes

[obsidian mantle]

Why would it skyrocket

[warped narwhal]

cause only crucial makes it, so anyone else can just set the price to whatever they want and no-one will have any other choice than to buy it

[obsidian mantle]

Oh wtf

Isnt that illegal in 60 countries or smth

Or wait why would it be

[warped narwhal]

"ah yes this 4gb sodimm ram stick always cost £2000 "

[obsidian mantle]

Its not food

[opaque sigil]

i thought i'd browse through amazon to see if i'm just imagining things and nice

[obsidian mantle]

That isnt crucial

[opaque sigil]

indeed

there are other companies that make so-dimms

it's just that crucial is one of the bigger ones

and i guess more well-known companies

[obsidian mantle]

But then who is going to buy overpriced crucial over normal priced alternatives

Other companies are expected to raise the price too cause why not if majority got more expensive?

Also what about samsung

Or wait was it sony

I forgot

[opaque sigil]

samsung

[obsidian mantle]

There was a conversation here recently

About samsung being good

[opaque sigil]

their ssds?

[obsidian mantle]

Oh probably

Nvm then

[mighty thorn]

Guys

[opaque sigil]

samsung doesn't really sell consumer memory kits iirc, they target servers more or act as an OEM

[mighty thorn]

Someone should breed a species of bacteriophage to target human tissue cells

For trolling

[obsidian mantle]

Isnt that what common flu does or smth

[mighty thorn]

Not even close

[obsidian mantle]

Hm

Well gen ai overlords might refer to this later idk

[fast pagoda]

Wtf is going on
Are consumers just not allowed to build computers anymore

Is that what they want

[supple gust]

Ah you saw the news

The Micron stuff

[fast pagoda]

Or they really want Chinese ram market to fill the void

[mighty thorn]

Oh god

What happened

[obsidian mantle]

New virus?

[mighty thorn]

Give it to me straight

[supple gust]

https://investors.micron.com/news-releases/news-release-details/micron-announces-exit-crucial-consumer-business

[fast pagoda]

Micron quit

Consumer business

[fast pagoda]

So RIP crucial

[supple gust]

Yup RIP

[mighty thorn]

I’m literally

Bruh

[supple gust]

I was lurking the discords to spread the news and talk about it

[obsidian mantle]

You cant buy those or cant insert it because different socket or smth?

[fast pagoda]

Welp, I was a bit apprehensive about it, but I am so fucking ready for my new friend Changxin memory technologies (CXMT) to fuck micron up

[supple gust]

I bet most of the chips are going into the ECC Ram that Servers use

well all of them are now

rather than consumer ram

[fast pagoda]

Save me based China, friendship ended with South Korea

[obsidian mantle]

If there is demand there will be supply. Someone fills the shelf

[fast pagoda]

CXMT will

They are still behind but are not that far off

Dram isn't like x86

Dram is dram

[supple gust]

I've hear people talking about how they might just be planning to go all in on AI. And when the bubble pops just jump back into the Crucial Brand

Which I have doubts about

[fast pagoda]

If they ever get EUV going in china their DDR5 will be super fucking good anyways

[opaque sigil]

[obsidian mantle]

Dont we all have a computer already just use old computer

[fast pagoda]

CXMT parts are already better than equivalent Samsung/hynix and they're not using EUV for them while the Koreans did

[supple gust]

Its honestly so true. Because funny enough I was going to jump into that Virus talk with Bacteriophages. Complain a bit that I don't have access to stronger compute for more undergrad research work.

Because the ML work in biology is where the real stuff happens

[fast pagoda]

Folding@home my beloved

[supple gust]

Folding@Home?

Is that a protein modeler run locally?

[fast pagoda]

Pooled compute

For protein folding

[obsidian mantle]

Whats so special about that kind of virus? Will it not die to human immunity system?

[fast pagoda]

Like a Bitcoin pool but protein folding

[supple gust]

I have actually done two annotation projects for Bacteriophages

[fast pagoda]

You used to be able to set your PS3 to fold for it

[supple gust]

They are too specialized in infecting Bacteria to actually even effect human cells

[fast pagoda]

Folding@home - Wikipedia https://share.google/O3YEfSxJLwm2C4Hbl

[supple gust]

Nice will look into that later

thanks

[fast pagoda]

I used to contribute to the overclock.net folding team lol

[supple gust]

But yeah bacteriophages have no effect on human cells. Non Susceptible and Non permissive

[fast pagoda]

10+ years ago

[patent shard]

what I find baffling is the lack of optimization
they didn't use CUDA until 2020... wat?
and AMD also runs poorer than expected.. but I think they are just using opengl...?

[fast pagoda]

It's not scientific compute unless someone in the chain is rawdogging FORTRAN

[patent shard]

it's hard to know exactly what the performance is, because PPD is not reliable metric.. if a machine finishes a work load sooner, it earns drastically more points
I think they have it skewed up just a liiiittle too much with that

[supple gust]

Pretty big rustbelt for a lot of tools in biology

I know our work with DNAMaster for a long time was a huge nightmare because of how old and outdated it is

Before we switched mostly to PECAAN.

[opaque sigil]

i don't like biology, too much data

[supple gust]

But its fun. You get to look at your stupid little angry screwdrivers. Or atleast I do

[mighty thorn]

Which is why we should breed them to hunt human tissue

For the LOLs

[obsidian mantle]

What will it actually do

[supple gust]

What? That... then they would just be viruses. Because Bacteriophages are specifically viruses that infect bacteria. Plus you know the whole Prokaryotic and Eukaryotic divide

[mighty thorn]

Yeah but they are more fun and don’t have good defenses against them

Talk about a win win

[obsidian mantle]

Will it cause diarrhea by ruining sromach bacteria life

[supple gust]

In the human body... I imagine it would still cause a immune reaction? But since the bacteriophages dont replicate in human tissue. It would get cleared out quickly

Also Bacteriophages are really really specialized to individual species of bacteria

[obsidian mantle]

So it will be like human-tissue-phage?

Since tissue isnt exactly bacteria

[supple gust]

Those are just normal viruses.

[obsidian mantle]

That's what i said by mentioning common flu earlier

[supple gust]

Oh do you mean a virus that preys on a virus?

[obsidian mantle]

I dont get it

Either

[obsidian mantle]

Whats the problem about defending against it

Just another organism to be killed by immunity system

[supple gust]

So this is something I've been trying to figure out. Viruses that prey on viruses exist. Satellite Viruses and Virophages. But barely any research exists on them

Theres a few papers

[mighty thorn]

The problem is that they don’t target us so, given how evolution works, we shouldn’t have any specialized defenses against them

[obsidian mantle]

Huh.. using viruses to hunt viruses?

[supple gust]

The most famous satellite virus being Hepatitis D

[mighty thorn]

So basically half of the immune system gives up

[supple gust]

Well the whole evolution thing cuts both ways.

Bacteriophages have no mechanisms for inflecting human tissues

[obsidian mantle]

But its just a matter of time to develop the mechanism. Arent human eretrocytes (sorry idk how to spell it) basically omnipotent except dealing with its owners cells?

[mighty thorn]

That’s where “breeding them to target human tissue” and GMO shit comes in

[mighty thorn]

It would be just a matter of time however that time would be, give or take, a few hundred million times longer than it would take the infection to kill someone

[obsidian mantle]

Huuh

Oh nvm the reply

[supple gust]

How to put this... The amount of work it would take to turn a bacteriophage that infects and replications in Prokayotes. To a virus that can infect eukaryotes... You would literally be better off picking up any eukaryotic virus to change them even if that technology existed.

[mighty thorn]

Yes yes

However

It’d be funny

[obsidian mantle]

I thought they can kill anything alien

Since it constantly evolves anyway

[supple gust]

What are you referring to specifically?

[obsidian mantle]

Or these guys are super different thing?

[rigid snow]

what am i reading

i;m so lost

[obsidian mantle]

Bioweapon development

[rigid snow]

[supple gust]

I am trying my best out here. I'm just a broke undergrad student man

[mighty thorn]

Yeah basically boweapon development

[obsidian mantle]

this sounds interesting after what ive said

[rigid snow]

vine boom sfx

[supple gust]

I just wanna research my funny little guys to help fight anti biotic resistant bacteria

[obsidian mantle]

Starman the evil biologist

[supple gust]

Dang talk about out of context

[mighty thorn]

Yeah but it’d be much funnier to research bacteria to help them fight off antibiotics

[supple gust]

Look if you wanna go and do that in the labs then hats off to you. No idea where you are gonna find that lab though

[mighty thorn]

DIY bioweaponary lab is gonna go crazy

[rigid snow]

very little stuff offends me on the internet but this somehow did

[supple gust]

Good luck on the Illumina machine. Not even sure how much those run

or you could go Nanopore I guess

[obsidian mantle]

Wait, bacteriaphages = antibiotics? Makes sense actually

[mighty thorn]

No

[obsidian mantle]

Im confused

[supple gust]

Long and short of it. Bacteriophage Resistance and Anti Biotic Resistance seem to be exclusive traits in Bacteria. Or atleast difficult to co evolve

[supple gust]

So bacteria with high Anti Biotic Resistance seem to have poor Bacteriophage Resistance

[mighty thorn]

This little goober

Is bacteriophage

[supple gust]

Yup siphoviridae morphotype

Long tail with spikes at the end

[obsidian mantle]

It kills it to use it as building material to replicate?

[mighty thorn]

It makes it replicate itself for it and then kills it

So basically but also no

[supple gust]

It injects its genetic material into the host bacterium then replicates inside it

[obsidian mantle]

Virus type shit? Turn the target into factory to make its copies?

Oh right

[supple gust]

via taking over the hosts cellular machines

[mighty thorn]

Virus but with legs and funny movement pattern

Way better

[supple gust]

Well Viruses dont really... Move with intent

[obsidian mantle]

Why did it not evolve to use human cells like viruses

[supple gust]

they just kinda float till they bump into a bacterium that triggers the tail tips proteins

Because Bacteria have a totally different way of storing DNA. Prokayotes store there DNA in a circular formation in the center of the cell

Eukaryotes have organnelles. And a Nuclear Envelope.

Also Bacteriophages have no ability to actually penetrate the cell membrane of Eukaryotes

[mighty thorn]

*yet

Not until I get involved

[obsidian mantle]

A bunch of major issues then

So its very active virus basically?

[supple gust]

Because there are two qualities all viruses need for successful replication. Susceptibility and Permissibility.

[obsidian mantle]

Virus with legs

[supple gust]

Susceptibility refers to a viruses ability to actually enter a cell

Permissibility referes to a viruses ability to actually hijack cellular mechanisms to replicate itself

Viruses can have either, neither, or both depending on the target cell

[obsidian mantle]

This is so dank actually

Hijacking replication thing

[supple gust]

Then you can get into the fun stuff like Defective Interfering Particles, Therauptic Intefering Particles, etc

Also Viruses aren't really alive and dont really possess any kind of what we would call agency, will, or really movement

They just float around till they bump into a cell that they have susceptibility on.

Completely random

[obsidian mantle]

And then some button triggers

[supple gust]

Typically from what I understand. A virus will have a membrane protein that matches with something on the target cell. Which causes the reaction to occur

Though my knowledge fails me in a lot of that area. Most of my knowledge is highly specialized in Bacteriophages. But my general virology is decent. Moving further away from that gets a bit out of my wheelhouse

I am working on shoring up my general microbiology. Especially since there really isnt anybody at my college specialized in Virology. Not much more to learn on that front till I can get into grad school.

[mighty thorn]

Spores

[supple gust]

Fungi are across in another zip code from my wheelhouse. Leave those devil beings alone.

This whole conversation makes me wish there was a really popular biology youtuber. Would love to lurk in that Discord.

[mighty thorn]

Is analogy

[mighty thorn]

Anyway

Enough bio weapons

Ai am I right

[supple gust]

Plateaued I think

[faint sandal]

[hearty notch]

[rigid snow]

o

[dull egret]

https://www.techpowerup.com/343119/samsung-reallocates-nand-production-to-dram-across-korean-fabs

Oh hey, they're gonna slightly fix the problem with RAM they started with their cuts in production, by dedicating the ENTIRE new plant to DRAM.

Then there won't be NAND for SSD's, so THAT will shoot up, but RAM will drop again!

[uneven pulsar]

https://pcpartpicker.com/list/Z4YQFZ (heres the link)@stray dragon

[stray dragon]

was going through list "yeah looks fine" and then i see the psu LMAOOOOO

[uneven pulsar]

is it bad? 😭

cause i clicked add from filter

[stray dragon]

that thing might cook off.

let me find something

https://pcpartpicker.com/list/fGtKyW

unsure if the mid tower case will fit the gpu

don't know how big 9070 is

[uneven pulsar]

according to new egg its length can be up to 249 mm - 290 mm

[stray dragon]

ram prices are pain right now if you can wait like 6 months but if you don't want to wait it should be fine to get that much

i do recall intel's new cpu being a good deal though

even with the dead-end socket

let me check price on that

$150 cheaper than the 9900x for same performance

i recall you using linux, not sure if it being intel will cause any problems

build like this would work though https://pcpartpicker.com/list/sk7pqH

[uneven pulsar]

I do not have any problems with intel on linux, but i do plan on going on all amd, Idk the performance on amd cpus on linux is but i do know amd gpus on linux is really good

oh that price is lovely though

[stray dragon]

something to note: the intel sockets tend to be "dead-end" meaning they don't upgrade well

if you want better cpu you probably need new mobo

though i don't think there would be much upgrade path from 9900x to whatever am5 caps off with

if you do think amd is going to do better then i would still recommend that tbh

i welcome fellow believers in the ryzen 9 9900x

sad to see the sk hynix platinum go up in price so much though

was like "$170 for a 2tb? i'm about to hit him with the $120 hynix platinum p41" and then it's just $200+

$220

what happened 😭

[uneven pulsar]

alright, i have decided, im going for the intel build! Cheapter and more powerful than my amd build. Im keeping the amd gpu though, thank you very much T your a real one

[stray dragon]

np

glad i could help

[uneven pulsar]

Plus im calling it the "Oreo Build" (black and white)

[stray dragon]

lol

i just do all black with no leds because i'm boring

i will say that the water cooling is probably unnecessary but it should be quieter if you care about that

and it's better if you like overclocking

[stray dragon]

good news finally

[uneven pulsar]

i do like quiet pcs, you think this one can run games pretty good?

[stray dragon]

absolutely

9070 should put in work

i don't think there's any games that you can't run with a 9070

you'd only run into issues with higher graphics settings

or with a 4k monitor

[uneven pulsar]

https://tenor.com/view/giggi-the-giggi-mina-mania-gif-4484770607757756595

me getting the specs

[stray dragon]

we're going to "epic embed fail" you.

[faint sandal]

https://tenor.com/6xM8w7wLGO.gif

[true hemlock]

also you sure going with such expensive CPU?

[true hemlock]

with intel, you can go with cheaper dram alternative like ddr4 btw

the difference is pretty negligible overall

[stray dragon]

don't think there's any motherboards that can support the 265k and ddr4 ram

wait does pcpartpicker just not show the compatibility

it doesn't show the compatibility but it's still only there for up to 14000 series

ah it DOES show the compatibility and i just forgot to remove the ram from the list before checking

huh, that doesn't actually look that bad https://pcpartpicker.com/list/bd7pqH

i have heard many bad things about 13000 and 14000 series cpus but i suppose it's an option

[mighty thorn]

Casually assumes that prices will go down in the 21st century

[stray dragon]

i feel safe in my assumption that ram prices will not remain tripled for more than 6 months

[mighty thorn]

Yeah it’ll be quadrupled by then at least

[stray dragon]

nah it'll be quadrupled by the end of the month

on its way to being quintupled

industry should move within 6 months to resolve the price issue though i think

[vernal current]

Here's some screenshots of my app, if anyone is interested. It's a media center for watching anime on the gabecube and steam deck.
https://imgur.com/a/tsundere-media-center-xngY6oP

[sick owl]

https://www.pcgamer.com/hardware/the-ram-crisis-is-just-getting-started-micron-makes-the-difficult-decision-to-abandon-the-consumer-memory-business-to-focus-on-supplying-ai-data-centers/

We are so fucked

[jagged turtle]

great... just great...

[vernal current]

they'll come crawling back when the bubble bursts

[mighty thorn]

I’ll be the one to say it

The new Minecraft version numbering system is only there to prevent the inevitable release of version 2.0

[amber fractal]

I want 1.31.4 to be real 😭

[tender river]

@safe path @rigid snow yesterday me make advent of code day 1 gnu apl
z←0◊{z←z+50=100|⍵◊⍺+⍵}/0,⌽{(⍎1↓⍵)×1-2×⍵[1]="L"}¨(x≠"\n")⊂x←⎕FIO[26]"01.txt"◊z

[faint sandal]

[clear sedge]

gotta wait for 31.4.1

[mighty thorn]

Yo neurosama reference

https://huggingface.co/papers/2512.02817

[glad path]

where can i find aoc prompts

[tender river]

problems not prompts

[glad path]

prompts like an essay prompt

ykwim

[tender river]

i know what you mean but these are problems not prompts

[glad path]

like it tells you what ur supposed to do???

[glad path]

ok same thing

but where can i find the problems

[tender river]

https://adventofcode.com/

[tender river]

essay prompt is not that

essay prompt is, quoting wiktionary, "A suggestion for inspiration given to an author"

[glad path]

like

prompt like inktober has prompts

[tender river]

ye prompt is either for llms or for creative open ended problems

[glad path]

isnt aoc somwhat open ended

[tender river]

you have many ways you can solve it but solution doesnt change

[glad path]

yeah so open ended

[tender river]

you would maybe call something like "a program that draws pictures" a prompt, but "make a program that finds an answer to 2+2" is not a prompt

[tender river]

the end is

literally

closed

[glad path]

ok

you get what i meant at least

[tender river]

true me just think you're not an llm

[glad path]

ill prolly start on aoc tomorrow

im tired rn

[olive sable]

https://tenor.com/view/azumanga-daioh-gif-2910224911933025207

Goodmorning

[glad path]

owrhbpvbsrhgpbsfrg

ok im doing day 3s

i dont knwo what im doing wrong since my code works on the example

wait

[warm river]

good morning

[glad path]

nvm i fix

yay

[glad path]

finished day 3 :D

[gentle haven]

https://tenor.com/view/scaler-create-impact-any-key-programmer-coding-gif-25012205

[olive sable]

https://tenor.com/view/i-use-arch-btw-arch-linux-linux-20-century-fox-yes-i-use-arch-btw-gif-8383058472649882025

[amber fractal]

This is facts

[tender river]

haskell version (not code golf language cant really be clever with it

snd.foldl(\(a,b)x->(a+x,b+fromEnum(mod(a+x)100==50)))(0,0).map(\(x:n)->read n*if x>'L'then 1 else-1).filter(not.null).lines<$>readFile"01.txt"

[humble tapir]

I haven't done the advent in a long time. I don't even know what I'd do it in this year I hadn't thought about it.

[tender river]

i like it for learning new languages doing aoc in prolog was very fun

it doesnt ramp up that quickly so if you're a good programmer you can follow along even if you have 0 experience with a language

wont teach it to you that deeply but enough to understand the language at least

[trim valve]

ooh shiny new colours

surely I can manage to setup a nixos dual boot without exploding my pc

[tender river]

i triple booted nixos at some point

[trim valve]

[tender river]

your pc will explode

its expected tho just accept it

[dry charm]

just make sure your bootloader is safe :D

[trim valve]

I have two drives

so surely that mean I can just install nixos on one drive

and nothing will break

[dry charm]

if you can install on completly separate drive, yeah

and even if something breaks, 99% its just the bootloader, which can be easier fixed by reinstalling via Windows recovery

[umbral thorn]

[jagged turtle]

who approved this design bro

why is it so squished

[sage crag]

sourdough

hungry

[faint sandal]

how hungry...

[sage crag]

i could eat a horse

[rigid snow]

sane people that don’t have 9 neurillion tabs open

[jagged turtle]

:CAUGHT:

[rigid snow]

looks like edge

why don’t you use vertical tabs

[jagged turtle]

idk I'm used to horizontal tabs

[faint sandal]

hungry enough to eat food pics?

[rigid snow]

she will print all of them out and consume

[sage crag]

no, too much

not go to hopital

[rigid snow]

i am alive still

are you sure the chip is supposed to rot

[olive sable]

Why must you hurt me like this

[true hemlock]

printing this on a paper and eating it rq

[olive sable]

Its worth it tho

Id def go to the hospital for some good wings

My foxbox package arrived

Only took 4 months

[rigid snow]

what the fuck do you mean 4 months

we were discussing the tablet cover like

yesterday

[olive sable]

I ordered it in august

@amber fractal

[jagged turtle]

why did docker desktop installer just freeze

[dense cosmos]

steins gate PogW

https://tenor.com/view/tuturu-mayuri-tootooroo-steinsgate-gif-9868521318091589737

[gentle haven]

Damn now I’m hungry

[trim valve]

pros of new mousepad: mouse go brr
cons: not used to mouse go brr

[gentle haven]

Then get used

[faint sandal]

[trim valve]

[ivory plinth]

welcome to stupid idiot does stupid stuff
episode 1: hex editing a dll to return to sanity (windows 10)

[rigid snow]

ios bugs

apple is fascinating how do they have any competent people working for them with their culture

i guess being consolidated to one small team and not ever meeting anyone outside of it helps

[sick owl]

https://www.theguardian.com/technology/2025/dec/02/sam-altman-issues-code-red-at-openai-as-chatgpt-contends-with-rivals

Damn, the situations flipped

That was Google a year or two ago

[sage crag]

ye it has

you have compost in your head now

not go to hopital still

[rigid snow]

does EOL imply

you're dead

[sage crag]

end of life ye

it over

[olive sable]

https://tenor.com/view/makise-kirusu-steins-gate-gif-21154463

[silent cloak]

https://tenor.com/view/steins-gate-my-honest-reaction-okabe-okabe-rintaro-outer-wilds-gif-8278718716540290526

[true hemlock]

hahaha the more they're pushed to burn more money for chatbots the faster the bubble pops (and bigger)

pretty much inevitable but i want it ASAP

openai can go, idgaf

[maiden geyser]

openai mindbroke bro skull

[sick owl]

I don't want that to happen simply because it leaves Google and Anthropic with no real competition in the west (other than xAI, who I'd also really rather didn't dominate the market)

[true hemlock]

all would burst lmao

[sick owl]

Nah Google isn't going bust, Anthropic has a lucrative US military contract and xAI is funded by the worlds richest man as a vanity project

[true hemlock]

google has deepmind that research for something like alphafold

[sick owl]

Deepmind are also the ones making the LLMs

[true hemlock]

the llm side would burst

and i won't give a fuck

its pretty obvious though

openai admitted it themself lol

[sick owl]

I think that's something you want to happen more than it is something that would happen

[amber fractal]

[true hemlock]

they need trillions to keep up with their old "pace"

[true hemlock]

its already happening

[true hemlock]

.

this is obviously unsustainable

[sick owl]

Gemini added two hundred million users in the last two months

I'd hardly call that on track to go bust especially when they have Google money

[true hemlock]

im not talking about the userbase

lol

[sick owl]

But the userbase is what matters to Google

[true hemlock]

im talking about purely on tech advancement

which these companies are so obsessed with

they're burning through more than they earn

[maiden geyser]

are they the paying ones?

[sick owl]

Google are not burning through more than they earn

[true hemlock]

i didn't say google

[sick owl]

, Google's not interested in it for traditional revenue streams

[true hemlock]

i said openai

besides, openai gives 0 research value in the field rn

[sick owl]

Exactly, and I said if OpenAI went bust that leaves us with Google, Anthropic and xAI at the frontier of the western market

[true hemlock]

so they're not even important lmao

[sick owl]

Which is grim

We're talking megacorp, never released a non closed source model US military contract havers and Elon Musk

[true hemlock]

good then, faster the bubble burst when more companies are tight with the need to gamble on unsustainable brute force

openai already fucked up the dram market

im not happy with that especially when im still offering services with custom pc building in my place

i HAD to stock $9K worth of dimms

and that is NOT fun

thanks, openai. go to fucking hell with your slop

[sick owl]

The dram companies fucked up the dram market

They're constantly price fixing and Micron just straight up stopped selling to consumers

The demand doesn't anywhere near justify the insane price surges

[true hemlock]

yeah if openai is gone they had to go back begging to us consumers

[sick owl]

No they don't, they go sell to Google, Anthropic and xAI

[true hemlock]

genuinely, these bubble are uh

[sick owl]

Failing that they go to China and sell to Deepseek, Moonshot and Alibaba

[true hemlock]

nah

google aren't dumb

openai is just fucking desperate lmao

[sick owl]

Google aren't dumb, but there's also no alternative

They don't have in house memory fabs

[true hemlock]

and they don't need 900000 fucking wafers of dram per month

because they damn well know these ain't helping

[sick owl]

Realistically I still don't think OpenAI is going bust anyway

[true hemlock]

unlike certain managers in openai

why so? lmao

[sick owl]

When the bubble bursts we'll probably see the smaller startups go under leaving the giants untouched similar to the dot com bubble

The US sees AI competitiveness with China as a national security imperative

[true hemlock]

except the dot com bubble isn't an unsustainable rug pull

but this one is straight up one

[sick owl]

Those companies also didn't have unlimited government money to buoy them though

[true hemlock]

and we're getting nowhere. in terms of advancement llm are pretty much near dead end

talked about this one, maybe not here

[sick owl]

It doesn't feel like a dead end right now at least, though architectural innovation would be nice to see since its getting hella expensive to scale

[true hemlock]

sorry, but agi is not a real thing for the reason that conventional ANN are purely uncertainty emulation machines

[maiden geyser]

half life 3 announced or openai go bankrupt, which one goes first?

[true hemlock]

all openai think is that they can brute force their way of scale

[sick owl]

I'm in the Yann camp too, but that doesn't mean governments don't see it as a national security imperative

[true hemlock]

and its pretty much obviously just stupid

[sick owl]

The US would do anything not to fall behind China in this field

And that anything includes lucrative government bailouts

[true hemlock]

no one would win

the side that spent less effort on this wins, actually

investors didn't learn from the deepseek incident...

openai gonna bust sooner or later because they're pretty much stuck with pure scalability brute force mindset

there's just no more denying it

[true hemlock]

eh, when they burst i hope corpos will then understand this

[sick owl]

To be clear I do think there's a bubble and I agree there's a need for architectural innovation

[obsidian mantle]

holy fuck the x axis isnt linear

[true hemlock]

if not then guess phase 2 of rhe bubble

[sick owl]

I just don't think the big guys are going down with that bubble

[obsidian mantle]

gemini is x10 times more expensive?

[sick owl]

Deepthink is

[true hemlock]

openai over-relied on investors power lol

[sick owl]

Its basically unlimited test time compute

[true hemlock]

that by itself is already unsustainable

they pretty much aren't keeping up with their "exponential" prediction for profit

because apparently their experts forgot that logistic functions are more accurate of a representation

basic statistics

ah man

im pretty hella sure that these ANN tech will exist forever and not being abandoned

[sick owl]

Realistically if they really run into trouble I still think we see a 2008 style bailout and regular taxpayers footing the bill while the small players go under

If you get a $10,000 loan, the bank owns you. If you get a $10 million loan, you own the bank

[true hemlock]

its just that they're so data hungry, horrible scaling wise due to how the data sparsity works

[true hemlock]

after the bailout?

the fact that they're giants, just means that it'll quite straight up fucks up the economy surrounding it

including themself

i feel like this is genuinely different bubble unlike last time with dot com

where for some reason

every side is desperate

throwing everything they could

for the gamble

only later on for investors to realize

its all pretty much diminishing returns

not even realizing about architectural limitations

just straight up pretty much seem diminishing return overall considering how much goes into the R&D and what comes out

the logarithmic side of the logistic function... we're all there rn, just that open ai tries to hide it by throwing in exponentially more funds to cover it

somewhat simple calculus do really represent the situation, lmao

yippie, i guess

i still personally haven't finished being salty about ruining the whole dram industry

my business for next year is fucked, thanks

if not due to the HPC research job i had, i would've literally just die

and here they are, here he is, sam altman, expecting me to happy with him going "2027 AGI" or something something sora prompt puke

they could've stopped rn, focusing on dataset quality instead, and go on with slowly increasing profits

but no, had to order 900000 fucking wafers of dram every. single. fucking. month.

each wafers for enterprise settings are above $30K fyi

they're just gonna die atp

investors money are limited resources

but i'd really, really like to see a whole corporation miscalculated and gets fucked by the one and only logistic function graph that no one magically noticed

[true hemlock]

also guess we're chasing synthetic benchmark scores now on very very specific tasks that these models won't even see in real world usage i guess

man

[sick owl]

I assume after the bailout they actually start innovating again

[true hemlock]

that's... even worse

that would just mean they don't learn shit and would gamble again

[sick owl]

What do you think happened in 2008

I originally said this wouldn't be as bad as 2008 but not so confident about that now admittedly

[true hemlock]

it would be worse, honestly

[jovial remnant]

Why is the assumption that none of the big players are trying to innovate, while also doing the usual fast expansion?

[true hemlock]

dot com at least comes out as things that are actually crucial for even modern systems

[true hemlock]

the opposite

they're trying too hard

to the point of resorting to unsustainable brute forcing

expenses are well over 150 billions iirc

while they couldn't even make more than 20 yet

this is one of those case with big promises of exponential increases

[jovial remnant]

I mean you try what works, if bruteforcing has been working, then no reaon to stop until it also stops yelding results reasonably

[true hemlock]

nope

started to not work

scales are tipping to logarithmic hand side

you know how log function works

all they trying to do now

is just throwing even more to make it seem linear

any person that learned enough calculus knows that only exponential chain straighten logarithmic curves

which is what exactly happening rn, openai just straight up admitting they need trillions to "keep up with same rate"

and begging the government to help them

:)))))))

[jovial remnant]

So why would they be trying to do something that doesn't work? Surely people working on it are competent enough to know, otherwise they wouldn't be in the position they are in. Unless they they alredy stopped doing so, and are working on other things that keep yielding results

[true hemlock]

they're pretty much desperate lmao

its all about the investors

its double sided problem though. once investors realized all are merely diminishing returns

its just gonna go down

[jovial remnant]

Well to get investors you need results, and clearly the results are there among the few top companies

[true hemlock]

nah

investors aren't engineers nor nerds

they're business people, all they need is compelling promises

[olive sable]

in the bottom left corner of the print

[true hemlock]

and guess what seem "promising"

ai, i guess

what a big lie to even call modern ANN as "AI"... for how these models lack even the most basic cognitive capabilities

but i guess emulating language convinces a lot of people

[jovial remnant]

Assuming all investors are blind and can't see into the future is also wrong. Ain't no way most investors only work on promises, it's not that shallow.

[true hemlock]

you underestimated how much they gamble on shit lol

im not saying that they're blind either.

but its pretty much just obvious that they're being mislead with all the technical illiteracy

but hey, "99% of investors quit before we hit AGI"

[jovial remnant]

There is nuance to gambling on investnents. Seems unreasonable to believe every major tech company "fell for the hype" to that level

[true hemlock]

unfortunately

they also fell for the hype

its a bubble after all, and for a reason

[noble zodiac]

It doesn't matter whether the promised future is real or not. Right now enough people believe it to make the stocks go up and that's all that matters. Even if you know that it's not real you still go along with it and ride the wave to milk as much money as possible

[true hemlock]

exactly

[noble zodiac]

even if 50% were to know that its bullshit they would still go with it and why shouldnt they

[true hemlock]

i've sold my nvidia stocks when they're well above 200 lol

made good bucks at least :)

[jovial remnant]

"Make stock go up" is not everyones strategy, believe it or not, big companies plan to continue functioning for a long time. And idea of only constantly playing for current stock value goes against that long term. They can still play the stock market when an occasion happens, but not at the cost of the business

[true hemlock]

may i introduce you to the
https://fixupx.com/naranciagaming/status/1986495552276537441?t=a7n9UDsvJxNDDZiawV3vxw&s=19

narancia gaming (@naranciagaming)

oh my god it's actually happening

**💬 136 🔁 1.1K ❤️ 17.9K 👁️ 5.87M **

FixupX · Nov 6, 2025, 6:07 PM

[noble zodiac]

Microsoft, google, facebook, they sat on a historic level of cash savings. They can burn that and their core business stays untouched. They aren't really betting their existence on it

openAI, oracle and co, those will be the ones to fall

[jovial remnant]

To my knowledge, nvidia still has a solid business model and plan for the future. Nvidia wouldn't just be gone if current "hype cycle" didn't happen. And as far as I know, OpenAI is not the only current "top AI company", is similar situation as the one in OpenAI happening for all other US, Euro, China, exc. based companies? Also why wouldn't OpenAI try to do everything possible to get funding and see what they get away with? It hurts them not to try

[true hemlock]

nice

nvidia and microsoft would survive easily yeah

nvidia is still being relied on HPC

[true hemlock]

what happens exactly when they burn through the fundings and gets mediocre diminishing returns?

they lose trust.

[jovial remnant]

well, it first needs to happen

[true hemlock]

i think them going "we need trillions to keep up with the old pace" should already convince you its just happening.

if you knew basic calculus for statistics, you know what it takes to linearize a logarithmic curve

[true hemlock]

reference ^

[jovial remnant]

and for that to happen, other companies need to deny investments. Also "keep up growth pace" is not quite the "live or die at any moment". Besides there is only so much growth left possible for such a large company.

but doesn't seem like they are being denied investments yet. It's still growing year to year

[true hemlock]

which loops back to what tsurai said

[noble zodiac]

it only needs one of them to fall for everything else to tumble down along with it

[true hemlock]

investors including companies only cared about the overinflated stocks

once they're not satisfied with it, everyone just simply sells

[jovial remnant]

that "only" is doing a ton of semantic heavy lifting in that sentence

[true hemlock]

and for open ai, that is currently operating at a loss under investors funding

they'll straight up die lmao

[true hemlock]

because its true?

why would investor invest in the first place?

[noble zodiac]

and openai is invested by private funding, they dont even have a stock that can go up. They want to see actual returns

[jovial remnant]

well because for that "only" to happen, quite a bit needs to change from what the attitude is now

[true hemlock]

nah, current attitude is already making it happen

its simple, let me recap

open ai is spending much more than what they earns, promised exponential increase in profits

[noble zodiac]

the attitude is already shifting. Just listen to the shareholder meetings

[true hemlock]

what happens if that projection of profit didn't happen?

shareholder loses trust, pulls back fundings

[jovial remnant]

well that seems quite contradiction, doesn't it? They all same to "keep falling for the hype" and at the same time they "keep investing more"

[true hemlock]

open ai, being the one operating on loss

just dies

[true hemlock]

i suppose you're on short term memory...

[true hemlock]

.

its really really damn simple

its called as the "bubble"

that's all you need to know really

not the first time happening

there's no contradiction, just constant hype that all of sudden will just crash down

and part of the bubble are the people that wants the short term profit quick lmao

[jovial remnant]

Well trying to simplify the entire topic to a few key words usually doesn't show the true picture and ignores all the semantics of the situation

[true hemlock]

like me as example

[true hemlock]

i dunno what are you even trying to imply...

[jovial remnant]

"its called as the "bubble" that's all you need to know really" is kinda missing a ton of the reason for why stuff is happening, wouldn't you agree?

[noble zodiac]

Just look at tesla. Anyone who took econ 101 can tell you that its valuation is insane compared to it's profits and outlook. There are only a few lunatics that actually believe Musks nonsense. But right now, people are still in, it's still rising, so ppl just play along. Everyone can agree that it's a pipedream and still keep it going

[true hemlock]

did discord only sent you that

or you just couldn't be bothered to read the rest of my messages for the past few minutes

and just hyperfixated on that one bubble message

[true hemlock]

we're mostly just taking advantage of the bubble for short profits :)

[jovial remnant]

I could, but those seem to be different situations summarized and are presented as the exact same?

[true hemlock]

yes because i explained why all of these are a giant bubble

for one, there are the part where investors with absolutely no technical knowledge are promised with huge profits especially for private investors
and for another, we external investors just sort of... take advantage of that for short profit bubble

been fun

once that private investor realized that openai just couldn't keep with the promise and just burning absolutely everything

they pull out

we all pull out

open ai, operating on 13B revenue but 180B loss. where do they get the money from? the same private investor. where is that now? pulled out
dies

[jovial remnant]

but you seem to cherrypick examples and then generalize them to all. Like with the example just now of some investors not having technical knowledge, you seem to present it as "every investor is like that and they all don't know what they are doing"

[true hemlock]

because they do not know what are they doing. or just in for the short profit

either those 2.

[jovial remnant]

And why do they not know?

[true hemlock]

because they're business man?

not some sort of technical engineers?

[jovial remnant]

kinda feels like to be consistently successful with investments you also kinda need to know what you are doing.

[true hemlock]

i don't believe on their "technical degrees". comp sci people are jobless because all they think is that degree would get them a job

[true hemlock]

which is why they can just pull off fundings anytime

its not like the investors are losing anything

if anything, they're profitting off the bubble

and THAT is what they want

[jovial remnant]

ah so the "enemy is strong and weak at the same time" kind of thing. They know what they are doing, but they also don't

[true hemlock]

they don't know shit about the tech

but they know how to profit off short term

now you're overgeneralizing the fields lol

for the sake of argument

[jovial remnant]

am I the one overgeneralizing tho?

[noble zodiac]

And people don't actively have to invest into the bubble. Index funds by their nature weigh high earning stocks more. Everyones 401k, countries retirement funds, they all funnel money into Nvidia and co

you have to try REALLY hard to avoid it

[true hemlock]

^

again, been fun.

open ai is dead

well, will be dead if you cared about the grammar

[jovial remnant]

aigth then, whole conversation was kinda a circular reasoning, but was fun

[true hemlock]

had quite the fun milking off the stock

i get to afford more silicon wafers yay.

this is a wafer full of supposedly dram chips

[rare bramble]

thats worth like 500K in this economy

[obsidian mantle]

is this how they make it? in circles?

[true hemlock]

pretty much

[olive sable]

what happens to the corners? smelted down and back into a new wafer?

would a square or rectangle not be more efficient use? im assuming there is a reason for circles?

[opaque sigil]

more reliable

[obsidian mantle]

yeah pretty weird

[olive sable]

more reliable as in less deformation?

[opaque sigil]

less stress overall cause no sharp edges

iirc the purification process naturally create a cylinder anyway, or something like that at least

which they then slice the wafers from

[olive sable]

pure silicon spike

[amber fractal]

You can also use a turn table

[olive sable]

[olive sable]

this bad boy blocks all the electrical interfernce, has antiviral properties, and is non-reactive to water

perfect for ingestion

[opaque sigil]

yummy

[amber fractal]

[stray dragon]

yum i love taking bites from my silicon boule in the morning

[olive sable]

[obsidian mantle]

is this a monolith material or it has some structures inside?

[olive sable]

jsut dont chew too much cuz silicon dust is bad for you

silicon block: good
silicon dust: bad

[stray dragon]

monolithic silicon crystal

[olive sable]

i think its a big chunk of 100% solid silicon

[obsidian mantle]

oh and the crystallic grid or how its called has to be very well done

[stray dragon]

they're formed like that so the silicon only forms from one crystal

[obsidian mantle]

i get it now

[stray dragon]

they're also worth like $10000 per boule or something

[obsidian mantle]

and they just cut things out of it?

[stray dragon]

ye

[obsidian mantle]

and the more pure crystal molecular structure is, the thinner things they can cut

[stray dragon]

if there's multiple crystal matrices in the boule it fucks things up, as borders between two matrices have different electrical properties

i believe boules can also have some doping agents as well but yeah mostly silicon

[olive sable]

looks kinda like a pile bunker

[stray dragon]

"more pure" it's an entirely pure structure

[olive sable]

except this one would shatter on impact, so maybe best reserved for killing dragons or something

[stray dragon]

if you have less than that, it's a defective boule

they cost $10000 per boule for a good reason

[olive sable]

then how do you end up with imperfections if the crystal is 100% pure?
problems with the cutting? lithography issues?

[obsidian mantle]

10k sounds so little for that cylinder

[stray dragon]

imperfections can result from other means

[obsidian mantle]

or is it smaller than duck's photo

[olive sable]

hard to say, i think the sizes are pretty standart tho no?

since the machines are so expensive, you dont really want to have to deal with different sizes

[obsidian mantle]

so its this wide?

[olive sable]

supposedly sizes are standardised, but there are different sizes

1, 2, 3, 4, and 6 inches

[obsidian mantle]

oh

[olive sable]

no wait, those were the historical sizes

[obsidian mantle]

ducks photo looks much bigger than 6 inches

[olive sable]

the ones used these days are 8 and 12 inch

[stray dragon]

damn so that one in the picture is at least 4 feet long

[olive sable]

depends on which size

[obsidian mantle]

so about 1:1 is lost by cutting

[olive sable]

idk

this is germanium i think

[warm river]

lowkey i'm bored and i want port neuro-sdk to work on postal 2 bcs why not

[olive sable]

i copied it form an article about silicon, but the image itself says germanium

[obsidian mantle]

ai generated article

maybe its the same for silicon

[hearty notch]

gm

[olive sable]

apparently my cpu had a nice long bath before getting lasered and electrocuted into thinking

[obsidian mantle]

is germanium used like that? for circuits

[hearty notch]

[olive sable]

germanium is a semi-conductor, apparently

for high-speed transistors supposedly

and infrared optics

fiber optics too for some reason

high efficiency solar cells?

i guess they just use it for whatever

[fast pagoda]

Germanium was the first used semiconductor

For like the first 15 years

[amber fractal]

I was thinking usb chargers but that is the wrong element

[fast pagoda]

That's gallium nitride

[amber fractal]

Yeah I realised that as I was searching

[fast pagoda]

Germanium has a smaller band gap so it requires less power for its semiconductor properties

And sillycon had impurity issues

[simple sky]

can someone help me understand summation? (∑)

my networking class is starting to use calculus, and I haven't touched calculus in 4 years

[opaque sigil]

sum = 0
for i in range(n):
    sum += i

[simple sky]

oooooohhhhhhhh

see THAT makes sense

[amber fractal]

or even just

>>> sum
<built-in function sum>
>>> sum(range(10))
45
>>>

[simple sky]

haha, I got a C in calc

[sage crag]

[opaque sigil]

took me 3 tries to pass too, sure was fun

[olive sable]

skill issue

[sage crag]

🔪

[olive sable]

i did pass calc, but it did suck

limits are fine, derivatives im okay at, but integrals were a pain

im so glad im doing geometric algebra now instead

[hearty notch]

I LOVE CALCULUS

[olive sable]

why?

this is stockholm syndrome

[obsidian mantle]

integrals are fun but then they introduced matrices and it made no sense

[olive sable]

what

[obsidian mantle]

or are matrices from different subject

[olive sable]

matrices are the part that made the most sense

[obsidian mantle]

it was math related

[olive sable]

but matrices arent calc

[amber fractal]

Graphics programmer moment

[obsidian mantle]

linear algebra is where i died i think

is it matrices

[olive sable]

ye

matrices are algebra

[obsidian mantle]

i never grasped why would you multiply a bunch of numbers in that funny way

[olive sable]

a matrix is like a math operation

instead of doing rotation, translation, and scaling of a vector in 3 seperate steps, you can do it all at the same time. basicly saving you a lot of valuable seconds

[amber fractal]

Surely these days it is entirely useless with no funding to spam those calculations

[obsidian mantle]

yeah this year i looked at neural networks a little and discovered that its just a compact way to depict this operation of multiplying a bunch of shit

ai was considered sci-fi when i was studying so idk

[olive sable]

i guess for LLMs matrices do stuff too, but i dont care about those. im here for the geometric algebra

vectors are my guys

[amber fractal]

LLMs but i don't care about those

[obsidian mantle]

there was something about derivative equations too

[olive sable]

wdym?

[obsidian mantle]

i mean when i studied

[sage crag]

https://tenor.com/view/typing-gif-gif-25597112

[obsidian mantle]

i remember i was overwhelmed by math

[sage crag]

what 1 + 1

[amber fractal]

fi

[obsidian mantle]

i held for 1.5 years

[sage crag]

incorrect answer, the answer is its undefined

[olive sable]

x probably

[sage crag]

as we have not defined any of our axioms yet

[obsidian mantle]

(define + *)

[olive sable]

just define your axioms bruh

[sage crag]

no that still relies on axioms

[sage crag]

yes that is how axioms work

[olive sable]

yes

[topaz tendon]

[olive sable]

i define the number 1 as meaning "one" like the word

[amber fractal]

[sage crag]

[nocturne olive]

Welp the last GTX left is 16 series

[sage crag]

defining symbol is not the same thing as defining axiom

[olive sable]

why no?

tbh i don't know what an axiom even is, i learned all this shit in dutch

[opaque sigil]

x = x

[sage crag]

axiom -> rule that define a system
symbol -> semantic value

[olive sable]

hmm

[obsidian mantle]

so axioms are using symbols

[amber fractal]

konii is

[sage crag]

not necessarily

well, yes

but

not in the way you are thinking

[obsidian mantle]

ok what axiom uses 0 symbols

[olive sable]

if konii is EOL, does that mean she is old?

[sage crag]

one that hasnt been stated

[obsidian mantle]

you need to apply rules to something to form axiom

[sage crag]

wrong ordering

[opaque sigil]

axioms are just things you declare as true

[sage crag]

ye

[opaque sigil]

because obviously they're true

[obsidian mantle]

i mean the rules which are defined in axiom need something to be applied to within axiom

do they not

[sage crag]

not necessarily obvious

[olive sable]

rules are meant to be broken, follow the lead of the imagenary numbers and make some shit up.
im introducing j, the delusional numbers

[sage crag]

trivial or non decomposable are useful

[sage crag]

sorry, j is already used a lot

[olive sable]

dont care

[sage crag]

like in the split complex numbers as the second imaginary unit

or sometimes as the imaginary unit in complex numbers if you are an engineer

[olive sable]

i dont think there are any alpahbet things left anyways

[sage crag]

make your own

define small_red_triangle as something

youre allowed to

[olive sable]

oh

you're so right

🔺

[sage crag]

i used to use smiley face as a variable in my exams

[olive sable]

when are they adding :big_red_triangle: ?

i need capitals too for my delusions

[olive sable]

[amber fractal]

Note to self, print out a photo of enub

[sage crag]

enub is actually trivially easy to draw i recommend it

its good for you

[amber fractal]

after I find a non emote version of it that is

[olive sable]

i should 3D model an enub

[sage crag]

search for benub in this chat

[amber fractal]

[olive sable]

big enub -> benub?

[amber fractal]

✅

[sage crag]

big enub === benub

big enub does not imply benub

[olive sable]

imply would be => no?

double arrow

[sage crag]

ive seen it written both ways

[olive sable]

oh

fair enough

[sage crag]

well, technically x => x is always true

but

[amber fractal]

the urge to crack open an entire poster paper to draw the enub is real, regardless I'll start small

[sage crag]

i was drawing

minub

[olive sable]

iirc we did => for logical implication, and -> for limits and function mapping

anywyas

i need to finish 3D modeling the base of the game characters

[sage crag]

notation is notation

it doesnt matter what notation you use as long as everyone is on the same page

[olive sable]

true

[hearty notch]

calculus is the language of change and all is change

[sage crag]

im a pedant

but

calculus is just maths

infinitesimal calculus

is what you are thinking of

[hearty notch]

ok fair

this is a colloquial English to technical English mapping problem

[sage crag]

yes

[olive sable]

calculus is functions on steroids, and i like to keep it to functions

[sage crag]

wrong

functions are much more evil than infinitesimal calculus

[amber fractal]

It is like if everyone isn't on the same page that is a good emote, we kill them

[olive sable]

explain yourself

[sage crag]

[olive sable]

[sage crag]

a function is well

broad

[olive sable]

input -> math operation -> output

its beautifull

[sage crag]

wrong

allowed input -> ????? -> unique output

functions are not limited to standard arithetic

[nocturne olive]

What's in the middle? Some black magic?

[sage crag]

its not a "math" operation

its just an operation

[sage crag]

yes, whatever the function does

a function transforms an allowed input to a unique output

[hearty notch]

letting time intervals (at first, but also just any dimension) tend towards zero is rly the underlying theme of all (infinitesimal) calculus

[olive sable]

i odnt think it needs to have a unique output, you just need to have only 1 output per input

[nocturne olive]

So it may as well be fun something(some: Any?): Any?

[olive sable]

preferably, that way i dont need to worry about it

[sage crag]

allowed input to unique output is the thing that functions do

[amber fractal]

unique from each other or from the input

[sage crag]

functions are many-to-one or one-to-one, but never one-to-many

[olive sable]

wdym by unique output? cuz x^2 with input -1 will produce the same output as for input 1. thats not a unique output

[sage crag]

two inputs cannot have the same output

wait

uuh

[hearty notch]

koniis just talking about like the vertical line test

or at least thats the first representation of how this property is taught

i forget the formal name of the property

[olive sable]

i thinbk konii crashed, we may need to use a newer version

[sage crag]

injective property

ee

[hearty notch]

ye

[olive sable]

1 input cannot have 2 outputs you mean

[sage crag]

yes

[sage crag]

is when function is one-to-one

[olive sable]

ye

[amber fractal]

we don't have the funding to update got to run with this one

[sage crag]

many-to-one is non-injective

[olive sable]

surely its not that expensive?

[sage crag]

functions are only invertible if they are injective

[olive sable]

jsut dont invert your functions then

[sage crag]

not ask

mhm anyway even some simple sounding functions dont have a closed numeric form expressable with standard algebra

[olive sable]

i said a function does input -> operation -> output. i never said anything about what soem people want to do with the functions. leave the poor function alone

dont invert or integratew them

[sage crag]

[olive sable]

thats black magic math

i dont aprove of inverting and integrating

[sage crag]

example

erf(x)

[hearty notch]

https://www.youtube.com/watch?v=RcVA8Nj6HEo

[olive sable]

oh god

ive seen that video before

never agian

[hearty notch]

yea i cant grok it im not locked in enough

[sage crag]

lambda calculus is a (different*) discussion

(*: technically no but also outside of the scope of this)

[hearty notch]

im too old for this shit

[olive sable]

oldge

[opaque sigil]

lambda calculus is academia propaganda to keep you busy

[sage crag]

lambda calculus is probably older than you

[opaque sigil]

[amber fractal]

the chayleaf license required us to email, so the konii one is probably also expensive.

[hearty notch]

well yes but my brain being too calcified to absorb it isnt about its age relative to mine

[sage crag]

1930s if im not mistaken

[olive sable]

99.99% of math is older than me

[sage crag]

the company is bankrupt and liquidating assets, sorry

[hearty notch]

[sage crag]

statistically 100% of maths is older than you

[hearty notch]

the tech world is so brutal...

[olive sable]

the konii one comes with ableos tho

[sage crag]

not all maths but 100% still

[olive sable]

ye

pretty much

[olive sable]

[sage crag]

shaw

[olive sable]

https://tenor.com/view/silksong-hollow-knight-hornet-shaw-hollow-knight-silksong-gif-13205433055363544945

[sage crag]

let x > 0 such that x(x) = 0

[olive sable]

sure

[sage crag]

good

dual numbers

not useful

not a field or a ring

fun exist though

[hearty notch]

for arbitrarily large values of three

[olive sable]

you lost me

[sage crag]

epsilon

[sage crag]

i think computerphile has a nice introductory video on dual numbers

[hearty notch]

i just like saying malformed expressions sometimes to see if i can throw exceptions in konii's processing

[sage crag]

me natural language processor too advanced