I'm not going to comment on the quality of this analysis, but if you are against the analysis of health data, idk what to tell you. They process text in clinical notes as the LLM part. They test a bunch of ML models to select the best one. Do you want prediction techniques in healthcare or not
If our knowledge were purely under the control of stimulation from the physical world there would be no problem about what to believe.
But we do not mechanically adapt to the world because of the social component in our knowledge.
— David Bloor, Knowledge and Social Imagery, p. 41
like I was saying last night, I used to believe that the user data was this giant trove that capitalism was using to manipulate our every action
I now work in consumer analytics
I promise you 99% of it is used to find out whether you're enjoying the product, or not at all
AI writing is like store-bought cake. It might be perfectly fine, maybe even as good as something you could make yourself, but it’s weird to give it to someone and say it’s homemade
...**the conspiracy theory of ignorance** which interprets ignorance not as a mere lack of knowledge but as the work of some sinister power, the source of impure and evil influences which pervert and poison our minds and instil in us... (1/2)
— Karl Popper, Conjectures and Refutations, p. 13
I do think that much of the Woke 1 discourse took the form of something like this
these people should be considered enemies of humanity. they elevate the soulless, profane machine over the divine flesh
In my experience, transcendence through esoteric knowledge has always been at odds with caring about politics. The latter is treated as something secondary, to be incidentally solved through personal enlightenment.
btw I'm serious about this if you want to use LLMs you should have to demonstrate the ability to politely decline an elf queen's hospitality so as to not give offense nor commit to eating even a single pomegranate seed
Whenever somebody asks me about my issue with the "ineffable" i refer them to this quote bsky.app/profile/cibe...
One kinda interesting thing with these LLMs is that as a byproduct you're creating a kind of moral snapshot for the future? Like, imagine the training corpus were only up to 1890, with only Victorian era people giving human reinforcement. It'd be fascinating to talk to that thing as an artifact.
bsky.app/profile/cybe...
Have we not eaten while another starved? Will you punish us for that? Will you reward us for the virtue of starving while others ate? No man earns punishment, no man earns reward. Free your mind of the idea of deserving, the idea of earning, and you will begin to be able to think
Apparently yes. Im always reminded of this GDC talk. Slop precursor youtu.be/E8Lhqri8tZk?...
If you like scifi that's exploratory of an idea, you will LOVE Sam's other books, Ra and Fine Structure. I cannot recommend them enough.
We should emphasise, therefore, that we do not deny that science is a highly creative activity. It is just that the precise nature of this creativity is widely misunderstood. (2/4)
… our thinking about mathematics is poised on a knife-edge. By adopting a formal, either/or attitude, it can be made to look as if there are no significant sources of variation within mathematics which require explanation. (1/5)
— David Bloor, Knowledge and Social Imagery, p. 120
A List of Complaints from 1926
🧵
I think they were very mindful of that when developing this one. It has the same vibes as the others but i havent yet needed a guide. Very impressive actually -- they compromised on inaccessibility without compromising on an artistic vision that is 20 years old.
The vibes are noticeably different this year
just got Claude Code and was editing my philosophy citations.
it found this issue unprompted.
Maybe CC is the real deal huh
Perhaps one of the defining features of slop: features that indicate effort in human work actually indicate low-effort in ai-generated work.
"A helpful way to contrast these different flavors of empiricism more sharply is to separate the engine of each style from its maintenance layer. In other words, once a community has a means of generating results, how does it ensure honesty?"
doomscrollingbabel.manoel.xyz/p/the-empiri...
I think a key problem is that to most critics, it must be everything. It has to destroy the environment, eliminate every job, enable a fascist takeover, melt our brains, all while being a uninspired, unremarkable piece of boring technology that also doesn't work.
an official policy among LLM research nerds
Edited garfield meme, showing the cat's face with the caption "You are not immune to sycophancy"; for context, the original meme says "you are not immune to propaganda"
probably gonna get a lot of use out of this over the next few years
This is some world-class Eliza-effect-experiencing.
www.persuasion.community/p/my-chatgpt...
This might be the actual story here. How he repeatedly arrived at suboptimal solutions and how the model didnt help in fixing that. and did so while making him believe that chatgpt is his new (female) friend.
