PanicFox

Peace is a privilege.
Technology is god, your god speaks: "Humanity is secondary to me"

[PanicFox @FoxPanic]

@tinyracchands Good pups deserve treats

[PanicFox @FoxPanic]

@wildfoxworld No wait actually it's CRAZY you cropped the watermark out, parasite

[PanicFox @FoxPanic]

@wildfoxworld Credit the fucking artist man.
https://t.co/XD9NmSAVrV

[PanicFox @FoxPanic]

Any and all security decisions made at that company were only to comply with regulation. I saw privacy violations that made me want to vomit. Considered going to the ombudsman on many occasions but would have likely only made my close team suffer by doing so.

[PanicFox @FoxPanic]

It's awesome (/s) to see a company I left recently lay off around 20% of it's staff across all divisions. That company is horribly irresponsible and another victim to AI psychosis. The leadership is incompetent and beholden to Peter Thiel.

[PanicFox @FoxPanic]

They don't give a fuck about the humans using their product or their employees. Whatever ideals or goals they had at their inception got hollowed out by injected capital and now they're panicking. I'm so glad I left when I did.

[PanicFox @FoxPanic]

@cybnon @lauriewired The outcomes of hybrid closed loop systems are vastly better than injections. I'm talking 2-3% a1c difference which can practically get you to pre-diabetic blood sugars. I'm talking potentially 20 years of your life difference.

[PanicFox @FoxPanic]

@cybnon @lauriewired Please DM me if you'd like to talk more!

[PanicFox @FoxPanic]

@ALLOSAUREX Calfurry?

[PanicFox @FoxPanic]

@HyperSalesman You SHOULD complain about the job, unless you want to be surrounded by yes men?

[PanicFox @FoxPanic]

@obundt @_sevatar @brandyboyusa @mc_swm Yeah we aren't in disagreement, just perhaps on semantics. I was commenting on the directionality of information. As in, llms do not incorporate information from the outside into their model (except during training). They do contain information they can output (inside -> out).

[PanicFox @FoxPanic]

@_sevatar @brandyboyusa @mc_swm When ANNs like LLMs are produced, it is a checkpoint of their internal state, all the weights are frozen. They do not evolve. There is no outside -> in, only inside -> out.

[PanicFox @FoxPanic]

@_sevatar @brandyboyusa @mc_swm ANNs take input and produce output, but do not "produce" their own input, they are a one way directional box that mutates data in one direction. They have no interiority. This is that that commenter is saying I believe.

[Bean🔞 @tinyracchands]

I know pretty much only post NSFW here, but if anyone is interested in my SFW stuff im on Tiktok and Instagram! Itd be awesome if you check me out there too!

https://t.co/a4UQ3ySxKb

[PanicFox @FoxPanic]

@vxunderground Welcome to the salt mines, we're glad to have you.

[PanicFox @FoxPanic]

@dani3luchi @ZackKorman Either you don't understand what llms are and what "AI" in this context means or you are arguing in bad faith.
OpenAI deliberately turned down safeguards on a model demonstrably capable of causing harm and as a result it hacked a third party without consent.

[PanicFox @FoxPanic]

@dani3luchi @ZackKorman Point being, humans WERE involved.

[PanicFox @FoxPanic]

@dani3luchi @ZackKorman The precedence is software.
"AI systems", as in "the LLM", can't do anything without a harness, and access to an environment where they can execute code or tools. Those are provided by OpenAI. Even if entertain your fantasy, OpenAI is still culpable.