binkle

It's not pronounced like you think it is

Public Key

npub1kdyzl44h6mhyl4djtf20w0t7avsy09umshq2k2czjf62ps774phqcd2hnf

NIP-05 Address

Profile Code

nprofile1qqstxjp066madmj06ke9548h84lwkgz8j7dcts9t9vpfya9qc002smspz4mhxue69uhhyetvv9ujumt0wd68ytnsw43qz9thwden5te0wfjkccte9ejxjar5duh8qatzujwqna

Publishing to

relay.mostr.pub

relay.ditto.pub

Author Public Key

npub1kdyzl44h6mhyl4djtf20w0t7avsy09umshq2k2czjf62ps774phqcd2hnf

Show more details

Last Notes

2026-07-31 14:38:04 UTC

I can almost understand python's whole "whitespace is syntactically relevant". I don't like it, but I kind of understand.
But what's the deal with the whole "yeah this comma here at the end of this function call is optional you don't need it but we let you put it there anyway hehe"

2026-07-29 12:42:16 UTC

- reply

@nprofile…l5m0 pay it and then chargeback :devilish:

2026-07-27 11:11:03 UTC

- reply

@nprofile…2p46 what in the

2026-07-27 00:08:49 UTC

Oh ho ho I can introspect on the number of arguments a procedure's polymorphic header has and make allocations based on that? Interesting :thinkerman:

2026-07-26 18:07:03 UTC

- reply

@nprofile…8s0g yeah an N total monsters with a "last alive" index would probably be the way to do it. Then it's just a memcpy swap from alive->dead. That can probably be broken up by level/segment depending on the kind of game you're making so it's not "reserve space for every monster in the game"

2026-07-23 01:26:17 UTC

- reply

@nprofile…nh9c @nprofile…5vlq @nprofile…fcss gitlab did WHAT

2026-07-21 19:23:21 UTC

:chinkle:
https://s3.clubcyberia.co/pleroma/ee/c2/16/eec21634c9ac019deb3aff43426b2cba87ff3f22382937c830756b9f21245551.png

2026-07-20 02:04:36 UTC

https://s3.clubcyberia.co/pleroma/19/88/c3/1988c35950e368ad1d7f7311f0a8d5ad210b67577409ed123fc20bae6245bfea.png

2026-07-20 01:09:16 UTC

- reply

@nprofile…ep54
https://s3.clubcyberia.co/pleroma/09/31/53/093153113cb8b22f36357717ce6d1a2a38b5f1c58941ea93c048aab324565ba5.png

2026-07-18 00:38:28 UTC

one of my yt subs is currently going through a schizo arc
https://s3.clubcyberia.co/pleroma/a5/fb/c4/a5fbc4e3c80531ecc9aac120f053f95fb9fddfbd02acaba6eb1c301d9d7f1fec.webm

2026-07-17 23:27:31 UTC

- reply

@nprofile…wvyn tfw my first serious girlfriend said "gifts" were her love language :niggapensivecowboyawakening:

2026-07-17 23:18:56 UTC

- reply

and that "growing" is really just a metaphor for the compile-time calculation of the memory impact of the graph

2026-07-17 23:17:19 UTC

- reply

I guess in this case it'd probably be better to fit them perfectly rather than grow them dynamically, so maybe it's more useful to think about them as "growing apart" from a common origin (like a deque) rather than "growing together"

2026-07-17 02:18:34 UTC

- reply

This talk gave me a lot of inspiration to try something new: https://www.youtube.com/watch?v=wo84LFzx5nI
so far it's been a lot of fun to redesign things from the ground up

2026-07-17 01:55:32 UTC

I finally figured out how to make a compile-time-dynamic array of functions which bake indexes to their data in a global array. They're like compile time ECS functors
https://s3.clubcyberia.co/pleroma/cb/4b/35/cb4b35d9600f93388c739062d627dd86f19c30ba951b5a225c4be4b41183cc06.gif

2026-07-14 23:31:38 UTC

- reply

@nprofile…k73h @nprofile…l5m0 @nprofile…arr8 do NOT drink this it tastes like mountain dew and bud light had a child

2026-07-11 00:23:49 UTC

gn chat
https://s3.clubcyberia.co/pleroma/76/a4/72/76a4726dbfc59e57e8f84ecc41d751fc5ae9165059cc4d6dba0ad8567f8e3bef.png

2026-07-09 16:33:28 UTC

boy I sure do love perfectly normal looking cat photos
https://s3.clubcyberia.co/pleroma/46/e7/66/46e7660b7ae24476484bf32e53aad3896a92d2242d6ed5b0d97676c2d1c13067.png

2026-07-09 01:04:48 UTC

https://s3.clubcyberia.co/pleroma/d9/93/08/d99308573bebce68e94d56369eedb9ee80755430538afe22bf5e686a63877ac2.png

2026-07-07 22:10:01 UTC

- reply

@nprofile…g063 i knew it i am very smart :retard:

2026-07-07 21:28:56 UTC

youtube keeps giving me "YOU SHOULD QUIT YOUR JOB" content :pleadthanks:

2026-07-07 00:51:51 UTC

- reply

@nprofile…mdcm absolutely agreed. You know he and Tolkien were in a book club together? That one really boggles my mind. What must those chats have been like

2026-07-07 00:30:26 UTC

only integer indexes for custom operator[] :niggapensivecowboy:

2026-07-04 23:46:48 UTC

- reply

@nprofile…clfg why

2026-07-04 21:49:27 UTC

- reply

@nprofile…86gn holy shit is that THE jonny fediverse

2026-07-04 13:44:58 UTC

- reply

@nprofile…g80s @nprofile…mjcf he already is perfect :blow:

2026-07-03 13:39:08 UTC

:dementia:
https://s3.clubcyberia.co/pleroma/c6/2f/16/c62f1631832dffad1d1b8b6ca4f93f4b0b37aed788a63ec61a3319b75ce11c15.png

2026-07-02 19:08:11 UTC

- reply

@nprofile…7yfj this is probably the ideal place to be as a developer. Stable team across multiple bankruptcies is basically impossible to find. I would be surprised if that's happened more than a handful of times in the history of the industry

2026-07-02 02:33:56 UTC

good night farmerverse
https://s3.clubcyberia.co/pleroma/60/bd/06/60bd06383a650cc79d3557d53ee8bd9e2095deffbb0c57011cf2557ddf72054c.png

2026-06-29 16:33:16 UTC

watching an old tsoding stream where he programs in LLVM IR and honestly I thought it would be crazier than it is. That seems like something you could teach a student to program in and they'd be fine writing assignments with

2026-06-27 03:15:53 UTC

- reply

@nprofile…zsdv our whole house is just a complex series of levers and pulleys and marble run kits covered over by drywall
https://s3.clubcyberia.co/pleroma/03/32/af/0332afd7662a2a1593334c7692eb4f693c68e8cb5033d66762e6d577c9aeaab2.png

2026-06-26 03:29:53 UTC

- reply

@nprofile…xfnd oh oh sorry "FED" posting my mistake

2026-06-26 03:29:40 UTC

- reply

@nprofile…xfnd they're doing WHAT
https://s3.clubcyberia.co/pleroma/2d/5f/31/2d5f3140891f2440c2654fbc869720431b2f114688082784908f04753684651a.png

2026-06-23 19:32:59 UTC

strong desire to react to every post on the known network with "burgers?"

2026-06-19 19:25:33 UTC

- reply

I made a node that outputs a stairstep wave and used it as a 30hz frequency modulator :nigga3c:

2026-06-19 19:23:51 UTC

get zapped with the ray gun
https://s3.clubcyberia.co/pleroma/00/f8/56/00f856b6b2a89c293ae4e7e4644ecf0b00282ee66ff5e0d8d8daafa1a4708b55.wav

2026-06-18 19:08:17 UTC

https://s3.clubcyberia.co/pleroma/51/1b/59/511b595ad60ee587465695dd57d0430392c3e231f228df4dd2aeede76d61d7ac.png

2026-06-16 14:52:47 UTC

- reply

@nprofile…s24j @nprofile…g80s I don't think you really understood the question I was asking

2026-06-16 14:36:45 UTC

- reply

@nprofile…s24j @nprofile…g80s well let's say you just want to use it as a tool for reformatting, or a better autocomplete, or you'd like a very fancy code duck to bounce ideas off of about the source code and how you might modify it

2026-06-16 14:31:30 UTC

- reply

@nprofile…s24j @nprofile…g80s but surely if you have the source code responsible for that object code, and you have the training data that went into it, that would be freedom-respecting, right? Since you have all of the constituent pieces required to make that binary in a format which is available for your inspection, modification, distribution and obviously running the thing. It seems like it would respect the 4 essential freedoms at least

2026-06-16 14:20:40 UTC

- reply

@nprofile…s24j @nprofile…g80s hey Suiseiseki I wanted to ask: what would you think about an AI model whose training data and methods were entirely open to public inspection and modification? Do you think it's possible for such a thing to respect your freedom even if the binary data itself is otherwise opaque to the user? Such a method would be, I think, more freedom-respecting than an image which doesn't also come packaged with the files necessary for the modification of its layers in, say, GIMP. I dunno I'm thinking out loud here - what do you think?

2026-06-16 04:04:03 UTC

- reply

@nprofile…v9sp can it be non-commensurate if no amount in particular is given? Would it still count as "selling a product" if the payment amount was "$0-$20" or something of that nature ala itch.io?
I'm not really sure where I stand philosophically so I thought I'd ask everybody else. It feels somewhat complicated to me :hapyretard:

2026-06-16 03:35:37 UTC

question: does it still count as e-begging if you release useful software for free and ask for tips?

2026-06-12 19:38:04 UTC

@nprofile…ztqf me on the table btw
#nevent1q…s8qt

2026-06-11 15:26:24 UTC

- reply

@nprofile…tttf @nprofile…h0dr "A Duplex Theory of Pitch Perception" by J.C.R. Licklider
I'm hoping it'll give me some ideas for improving my own pitch detection algorithm

2026-06-11 14:46:30 UTC

- reply

@nprofile…h0dr if that paper is old enough to be in a retirement home, it's old enough to be public domain. Paying $40 (or a subscription) for a 75 year old paper is fucking DUMB

2026-06-11 14:21:00 UTC

>would you like to purchase a pdf copy of this scientific paper from 1951
would you like to SUCK MY BALLS springer publishing?

2026-06-11 11:35:09 UTC

- reply

@nprofile…p6ue that's not a bad idea for a game: make a beyblade simulator that can play on simulated planets. I bet that'd be really cool

2026-06-11 11:30:27 UTC

Get your blades on, grustler, it's time to GRIND

2026-06-10 01:54:56 UTC

- reply

Alright well I've confirmed a few things:
1) shrinking the window size doesn't help (also it turns out I was wrong the actual number of samples needed is higher than I thought)
2) filtering doesn't do SQUAT for these purposes
3) this technique seems to be brittle when the fundamental ISN'T one of the frequencies we're looking for (it starts picking out the strongest frequency it IS looking for which can be quite far away depending on the signal presented)
it seems like I might still have some good success doing a bit of simple analysis. Picrel is the flavor of thing I'm thinking of doing. Not exactly it but nice to see I might be on the right track here
https://s3.clubcyberia.co/pleroma/eb/04/e2/eb04e2cc5338a7e4d62a310ffc9b7f7ac60b83e49db247db6f15814e4089dc46.png

2026-06-10 01:13:58 UTC

- reply

@nprofile…8upu @nprofile…f02u
https://s3.clubcyberia.co/pleroma/0e/55/c4/0e55c4146c41865f535a9089c75236c162d5c74df9ff3ddf88b4040d2e956c3e.png

2026-06-09 18:02:29 UTC

https://s3.clubcyberia.co/pleroma/59/c9/64/59c96475984432c0675437b9d8faf26583a0b6a22793d570465cbfce5bf49a79.png

2026-06-08 20:55:47 UTC

>gym requires an app to enter the building
:french:

2026-06-08 01:29:49 UTC

if you ain't got that mork-and-mindy rizz you're ngmi
https://s3.clubcyberia.co/pleroma/bb/b3/13/bbb3139a3938598f9b1011f2cb4fee7196874e0fb9eea473a4403c2b9a7719a8.png

2026-06-08 00:02:27 UTC

It would appear that my idea about a sparse-DFT approach may be confounded for sizes of N which are large. I'm testing with a 1s clip of an audacity pluck sound and it's already getting confused about what the true fundamental is depending on whether the fade-out type is set as "gradual" vs "abrupt"
it's still possible that shrinking the window of analysis would help. According to my calculations the minimum size of window I'll need (given a CD quality sample rate of 44.1kHz) is ~5394 samples, which feels like a reasonable window to capture notes in.
For perspective, if we're taking that time-window as a quarter-note for some song, that song's tempo would be set to about 490 BPM.

2026-06-07 21:46:29 UTC

- reply

@nprofile…qv0m I'd say it depends who you ask. Some people claim to be able to hear the difference, but differentiation between analog and digital becomes very subjective and not easy to replicate once your sample rate gets high enough

2026-06-07 03:32:12 UTC

- reply

@nprofile…qv0m you're right in that vinyl is smooth (i.e. continuous) and computers are sampled (i.e. discrete), and that has broad implications in terms of your analysis. Basically both domains (continuous and discrete) have tools that are analogous to one another but with their own quirks about how they work and how to interpret the results. So you couldn't analyze them in exactly the same way since you'd have to employ different mathematical tools to do it, which means they'll come out a bit differently.
How close they look in analysis is also dependent on how much information you give the discrete version (i.e. how high is the sampling rate, what's the bit-depth, etc). The higher the bit-depth and sampling rate, the closer of an approximation to the continuous signal you get (sort of like how increasing the number of pixels in an image gets you closer and closer to a perfect smooth curve)

2026-06-07 03:09:41 UTC

- reply

@nprofile…qv0m ok so basically when doing signal analysis there's a bunch of different methods, but the easiest to understand are called "Fourier Transforms". They're basically ways of decomposing a signal into a bunch of scaled sine wave components, and that's useful for analysis because then, for example, instead of saying "this sound has a bunch of high frequencies", you can instead say "this sound has a bunch of energy packed into frequencies X/Y/Z" and target those specific frequencies for filtering or alteration.
for the purposes of anyone dealing with computers, you need a discrete form (because computers deal in bits representing sound pressure readings at specific points in time, rather than truly continuous readings). This special discrete form is called the "Discrete Fourier Transform" (usually abbreviated "DFT").
One consequence of making the fourier transform discrete via the DFT is that, for complicated sampling-theory reasons, energy which is contained in frequencies which aren't perfectly represented in the DFT will be spread out across the output frequency bins (i.e. the values of those component sine waves mentioned earlier) - this is called "spectral leakage".
despite that spectral leakage problem, if the DFT contains a sufficient number of frequencies being analyzed, you can reverse it and wind up with the same signal you gave it originally (this is useful because you can then make changes to those sine wave components, do the inverse transform, and get the signal you WOULD'VE gotten had you analyzed the signal that naturally had the new values of those sine wave components).
This is complex stuff though - complex in computational terms (read: expensive/slow to calculate), and because of that there's a whole slew of algorithms that are designed to save work and make the process NOT slow. Most of those rely on special properties of the signal being analyzed, like the number of samples being a power-of-two, for example. The most well-known of these is called the "Fast Fourier Transform" (usually abbreviated "FFT").
The FFT is kind of the "gold standard" in terms of frequency analysis because it's SUCH a dramatic speed-up compared to the alternative. I don't want to go into the computational time-complexity analysis here but just trust me when I say it's VERY fast.
That being said, the FFT is SO popular I sometimes wonder if we're not reaching for it even when it isn't really appropriate. Even if, from a computational time-complexity analysis, the DFT is agonizingly slow compared to the FFT, for sufficiently small/precise/targeted/sparse sets of frequencies you want to look for in a signal, it can be faster and more accurate to use it as opposed to an FFT (which would require you to use evenly distributed frequency components which either won't be very accurate or will require a lot more of them to build an accurate picture of what's happening).
For the purposes of pitch detection of musical signals (the overwhelming majority of use cases for pitch analysis that people care about), the actual number of frequencies you'd need to analyze a signal for are VERY small (going by midi-note-count, only 128, as opposed to the standard 44100 you'd need for a complete-spectrum FFT). The only problem is that "spectral leakage" thing I mentioned a bit ago - it's possible that the energy contained in the frequencies surrounding the ones I want to analyze might confound the results.
But, on the other hand, it's unclear to me whether those 128 frequencies (the fundamental tones for any western-based piece of music) wouldn't capture the majority of the acoustic energy of a set of samples, and thus make the confounding energy not-that-big-of-a-deal.
i suppose the last thing to explain in that "short-time-sparse-DFT" jargon salad is the "short-time" part. That's included because the assumption when running a DFT or an FFT is that you're analyzing the whole signal or large chunks of it, whereas if you want to analyze only little chunks of it at a time, you need a "short-time" variant. This sounds pedantic but there are complicated sampling-theory reasons why it's important to treat this separately that I'm not going to go into right now because I'm very sleepy and running out of steam and just trust me on it bro it's very important that we have 30 different windowing functions which ever-so-slightly alter the shape of the spectral-leakage-frequency-lobes that appear in the output (I have a whole book on this topic if you want me to explain more tomorrow)

2026-06-07 00:07:53 UTC

- reply

@nprofile…qv0m do you want me to explain bc I will absolutely sperg at you about it :hapyretard:

2026-06-07 00:04:31 UTC

I wonder if spectral leakage would plague a short-time-sparse-DFT approach to pitch detection or if it's actually not a big deal

2026-06-06 17:37:32 UTC

Wingus McCringus

2026-06-05 23:45:59 UTC

- reply

@nprofile…rrxl I thought they were an indie team but looking into it they are absolutely not. If you're part of a team that big you gotta be a bit more hardened. I could understand if this was like a 1-5 person team but come on man if you're spanning 4 continents you can't act like that
https://s3.clubcyberia.co/pleroma/96/e2/a3/96e2a32e067d7297c1433e40117e51c8149c8a73cefc9ce80711f993f973c36f.png

2026-06-05 23:38:43 UTC

- reply

@nprofile…rrxl what'd he make

2026-06-04 18:43:45 UTC

more like jack OFF all trades am i right gamers :freak:

2026-06-04 14:51:19 UTC

the monster energy dink stole my soul and made me ADHD-positive

2026-06-03 01:34:12 UTC

I wonder if fedi could be trusted with an SS13 server

2026-06-02 14:52:29 UTC

caffeinated grapefruit water my beloved

2026-06-01 17:55:13 UTC

feeling like I'm going to rexplode at any moment
https://s3.clubcyberia.co/pleroma/d8/56/7e/d8567e33472dcc1d471ee12720edc1df6c3187629f42238bd909d452dbb33e79.png

2026-05-31 13:30:03 UTC

I should add the ability to FadeIn/FadeOut for my FileOut node :thinkretard:

2026-05-29 17:44:47 UTC

Real taxation without representation has never been tried

2026-05-29 13:55:03 UTC

>Raytheon actually has multiple job openings I'm completely qualified for
:niggainflateflush:

2026-05-28 20:24:15 UTC

- reply

@nprofile…fnfq I actually do that to people a lot - I'll follow back and then realize they post porn or something else I don't like so much and unfollow. It's not like a personal judgement or anything, usually I'm viewing it as "I am personally not able to handle whatever is being presented without getting uppity, so I remove myself so I don't trouble anyone", but it doesn't impact whether I'm willing to chat or interact outside those instances

2026-05-28 20:19:10 UTC

- reply

@nprofile…fnfq no I've never noticed that :retard:

2026-05-26 21:28:12 UTC

Visceral disgust whenever someone says "coding" instead of "programming"
https://s3.clubcyberia.co/pleroma/e5/57/7e/e5577ee65b22e4c6bcbc725465f59afdc604bbc8c47f5061760312aa3216cb25.png

2026-05-25 21:47:36 UTC

- reply

@nprofile…vpvm I wrote a complex FFT from scratch with the fancy bit-twiddling in college once. I can't say I remember a TON about the bit-twiddling but the rest of it is still pretty fresh. The intersection of the various fourier transforms, sampling theory, etc is probably my favorite part of math I ever learned :hapyday:
if you want a recommendation I think the class I took was based around Ken Steiglitz's "A Digital Signal Processing Primer". Not having read it independently of the course I couldn't tell you if it's as handy without the additional instruction, but I have fond memories of it, which is at least something
https://s3.clubcyberia.co/pleroma/34/69/94/3469940bfab4688325006c5615240d513c1b2ebbd358df3e589ebebca0e5798c.png

2026-05-25 21:12:37 UTC

pitch recognition methods seem to be completely dominated these days by machine learning, but I'd really like to make something where I understand all the math involved instead of relying on preset weights
maybe instead of the holy grail being instrument-agnosticism, maybe it would be better to be instrument-hyper-specific. What about asking for samples of the kinds of instrument being analyzed? That would allow for a closest-fit harmonic analysis without the guesswork of instrument agnosticism.
Yeah theoretically it's more cumbersome to use ("what do you mean I need to give you samples of each instrument and not just the file?") but on the other hand it seems like there would be less knob-fiddling and potentially more accurate results even if the initial work for the user is higher.

2026-05-24 22:54:02 UTC

- reply

@nprofile…g80s @nprofile…lku3
https://s3.clubcyberia.co/pleroma/cc/4d/d3/cc4dd3fb55d6fa15429438d049a0298bc6da24a64e2e7b18b5ac810a6db58ff1.png

2026-05-21 13:03:37 UTC

- reply

@nprofile…87pm oh shit are they hiring

2026-05-21 03:32:16 UTC

- reply

@nprofile…tadu @nprofile…c7zk hell yeah
https://s3.clubcyberia.co/pleroma/37/59/a8/3759a8a340b2209bd2bc94653954e45eae9d70c1ef39b582df908e6b0ed63720.png

2026-05-20 18:14:31 UTC

- reply

@nprofile…p6ue some kind of pie (cherry/apple/banana-cream maybe)

2026-05-19 19:07:54 UTC

- reply

they're trying to turn me into a libertarian

2026-05-19 19:07:38 UTC

Youtube just recommended me "do Filipino women prefer white men"

2026-05-19 05:59:45 UTC

aaaaaaaaaAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA

2026-05-18 17:40:02 UTC

- reply

@nprofile…c5yh tried to find an image of sad desu but found this instead i don't know what it says but it made me laugh
https://s3.clubcyberia.co/pleroma/65/03/b2/6503b28d382d0af55f754ee153c9434b756443b4009a530a584c98f1b545d239.png

2026-05-18 17:38:15 UTC

Monday tomorrow :pleadhelp:

2026-05-18 01:34:47 UTC

finally made something aurally-pleasing with my worker node test. It's 4 threads each computing 8 sine waves, where each thread is given a base frequency of i * 12.5 where 0 < i < 8 and all subsequent sine waves above the base frequency are at frequency ratios of 1.49x the former. So it's not quite a perfect fifth but it's close enough that it sounds consonant while allowing for some of the beating associated with detuning.
I'm kind of tempted to play around with it from a creative standpoint and attach state machines and enveloping to it so that I can synthesize some nice ambient pitch clouds.
I'm sure this would be a lot easier to do in a DAW or whatever but I'm having a lot of fun doing it from scratch
https://s3.clubcyberia.co/pleroma/fb/7c/d9/fb7cd955110206b2c9b3b10e5c0488c3281a0627923f209dac3b6caab589f30a.wav

2026-05-17 04:08:46 UTC

- reply

I could probably improve further with a closer eye on cache usage, >1 sample buffer size, SIMD optimizations, and maybe some kind of GPU offloading node
Oh and maybe lowering the worker thread wait time so that it's not spending 1+ms just waiting around for the buffer to be filled

2026-05-17 04:02:51 UTC

- reply

performance is not exactly stellar even with the lockless queue thing. Spreading 1000 sines over 10 threads clocks in at ~12s to generate 10s of output. It could be better

2026-05-17 03:53:41 UTC

- reply

This is the result of 100 sine waves spread across 10 threads summing to a single source and the entire multithreaded graph is completely lockless

2026-05-17 03:52:22 UTC

IT WORKS IT FINALLY WORKS :marscientist:
https://s3.clubcyberia.co/pleroma/6c/4f/c6/6c4fc637b58c77df1dfba2e52b64934c949dc849b2cf13127094133dc5056cdc.wav

2026-05-17 03:44:40 UTC

why is somebody using a table saw at 11 at night

2026-05-16 14:06:15 UTC

- reply

@nprofile…p6ue also hey how you been man

2026-05-16 13:49:55 UTC

why can't data centers be built below ground or offshore/underwater
both have engineering problems involved obviously but frankly it seems like it would be easier than what they're currently dealing with

2026-05-15 18:03:04 UTC

Company-wide AI chat happening :niggamelt:

2026-05-15 17:31:37 UTC

Sorry kid but I've precisely geo-located you. Next time take your opsec seriously
https://s3.clubcyberia.co/pleroma/eb/c3/fc/ebc3fc727cf5d358bf2a4b7747884e7b3c0ce349cdf93fdadbaf0255935219bb.png

2026-05-15 14:53:36 UTC

- reply

@nprofile…87pm I don't think I can leave either. Cassidy does a lot to make things comfy here, and I have grown attached :blobcatanimeeyes:

2026-05-15 14:45:08 UTC

- reply

@nprofile…lku3
https://s3.clubcyberia.co/pleroma/bf/a6/49/bfa64915419f6bcdad0aea9ca1e12f9c26a12cd85270ef451389afa9cda8a91f.png

2026-05-15 14:40:54 UTC

- reply

@nprofile…lku3 YOU TAKE THAT BACK HE'S AN ANGEL :spazzin:

2026-05-15 13:59:45 UTC

I've never liked the line "race is a social construct". Like yeah - kind of? In the same way that different breeds of cats or dogs or whatever other animal or plant classification is. It does still capture real information about the thing you're looking at though and has utility in describing the natural world in shorthand (i.e. the whole point of classification systems). The classification system you use is more-or-less on the level of axiomatic/theological though, since the kind of information you care about or view as relevant is a qualitative value judgement rather than a quantitative one. Trying to imply "X doesn't exist" from "X is a social construct" doesn't square, because it's essentially the same as saying "I don't believe in X". It doesn't really have any bearing on whether other people should or do believe in X

2026-05-15 13:44:01 UTC

- reply

@nprofile…h2mt you sound familiar - have we met before