sean_ae

all it tells you is computers can't tell the difference between their own output and a human's, but we already know that
Or to be more nuanced: even with clean data a transformer model won’t be able to discriminate since the resulting weights would be the same
I agree with your core point that this is a total shitshow btw

2026-08-01 12:09:56 GMT

https://post.lurk.org/system/media_attachments/files/117/020/171/642/016/313/original/b028bce65e0722a9.png

2026-07-31 13:43:45 GMT

- reply

right yeah makes sense

2026-07-31 13:37:26 GMT

- reply

how are they getting your email addys

2026-06-05 09:24:02 GMT

dammit
https://post.lurk.org/system/media_attachments/files/116/696/769/047/604/478/original/f14aa2ffe46ac8ef.png

2026-05-06 19:29:13 GMT

- reply

well sure, but complex systems are not just their constituent parts are they

2026-05-06 19:21:48 GMT

- reply

i see a clear difference between curating and designing though
i mean you can either be very careful with curation or not careful at all, depends what it is you're trying to do
and i think @npub1frq…kzu5 is right that most of the big LLM companies are working hard to achieve their objectives via tuning and curation, but i'm not buying the notion of the resulting models being designed
they're extremely black boxy, esp at that kind of scale
you can see this esp with grok because every time they try to rein it in it just becomes a new kind of unhinged

2026-05-06 17:43:29 GMT

- reply

yeah all of those things will affect your observation - but it's still an observation
you can say the same for any kind of analysis, recording, data gathering and whatnot (at least at the macro scale)
but none of that implies design
and the data you observe will exist regardless of whether or not you choose to observe it, or however you choose to
i suppose this is just a difference in philosophical approaches or something, we're prob never gonna agree :D

2026-05-06 17:28:34 GMT

- reply

ok sure, i guess you *could* say that performing an analysis on some data is in some way creating something
i don't see it that way personally, i see it as an act of observation
i think the thing you are observing exists regardless of whether you choose to observe it or not
what you're describing is more akin to editing than writing

2026-05-06 17:14:30 GMT

- reply

no one said it was neutral though, and lack of neutrality (whatever neutrality is supposed to be) isn't in itself an indication of design
i may just be having a hard time understanding what you mean by these terms like 'natural', 'neutral' and 'objective'

2026-05-06 17:04:23 GMT

- reply

yeah, if you carry out some analysis you will get the results of that analysis
idk what you mean by
>natural representation of the data

2026-05-06 16:07:38 GMT

- reply

i mean this is why we repeat experiments isn't it?
i'm not really understanding why you think the results of [some analysis] are themselves designed - i get that the analysis may be - but the results?
and you seem to be disregarding that the data itself was not designed either
and/or you seem to be implying that 'selected' and 'designed' are the same thing, and i don't think they are

2026-05-06 15:41:16 GMT

- reply

well you can tune the weights later on anyway (just ignore the ones you don't want to use)
but yeah you're performing [some analysis] and then working with that data
i mean, what you seem to be saying is that the output of any statistical analysis is designed - and this is kinda like saying the data you get from doing a scientific experiment is designed - and idk if you can really say that
i mean you can say the experiment was designed but the data is just [the data you got from that experiment]

2026-05-06 15:29:13 GMT

- reply

well yeah, i guess there are levels of arbitrary
i can work with some pretty arbitrary sound sets but there are always gonna be questions about how or why files ended up in the set - i get that
i just find this idea that [the model is therefore designed] to be a bit weak, but maybe we can agree on 'weakly designed' or something
and i obvs get where you're coming from wrt the bigger LLMs being horribly skewed and manipulated - but to me there's a big difference between manipulating something and designing it

2026-05-06 15:07:02 GMT

- reply

but what if the data is arbitrary?
(sorry for repeating myself)
i realise 'arbitrary data' is a very spherical cow here, but just hypothetically, if you had arbitrary data - would you still consider the model to have been designed?

2026-05-06 14:31:23 GMT

- reply

the network is the thing you are training, the model is a result of training that network
i suppose what you are saying is that since networks are designed, some of that design is affecting the quality of the model (and it is, but i would not call the model designed - otherwise they would actually work, and they frequently don't)

2026-05-06 14:25:47 GMT

- reply

ah ok, i see
when you say 'model' you mean what i would call a network

2026-05-06 14:17:11 GMT

- reply

can you explain a bit more what you mean there?
i don't have to specify anything when training a network locally, i can just chuck any data in (sounds/gestures/images/text), not label anything and get data out the other end

2026-05-06 14:13:55 GMT

- reply

hmm idk about this - you seem to have ignored the word 'arbitrary' in my question

2026-05-06 13:59:31 GMT

- reply

yeah ok, let me reframe this
if you didn't do any curation at all - and just trained a network on some arbitrary data - could the model be described as having been designed?

2026-05-06 13:50:27 GMT

- reply

if it's entirely designed then why are people having debates about mass scraping of art etc?

2026-05-06 13:43:55 GMT

- reply

right, but what you're saying there is that the design rests entirely in the commercial application of [thing] - but the application of what [thing] exactly?

2026-05-06 13:34:58 GMT

- reply

idk if you can say designed or controlled really, maybe slightly
you could maybe say built
it's not really a set of instructions, more a set of weights (most of which are not being set intentionally unless you only include human-generated data and even that would be a bit of a stretch imo)
not often i end up in a semantic hole, pls forgive me

2026-05-06 13:04:44 GMT

- reply

idk - do they even qualify as technology?
they're weird, statistical models, not designed or controlled, they just kind of *are*

2026-05-06 11:52:53 GMT

- reply

i see people describe LLMs as algorithms quite often but i'm pretty sure that's not the right word
i suppose they can *mimic* algorithms though

2026-04-30 12:24:50 GMT

in celebration of confield's 25 year anniversary we will be releasing 25,000 ball bearings into a gong factory