helt jävla nersugen
traveling
Joined November 2025
- Tweets11.4K
- Following20
- Followers42
- Likes1.3K
sugnerhan retweeted
Trump denying Americans Fable 5 access is possibly the worst thing he's done.
Trump administration officials tell WIRED that if Anthropic wants to rerelease Fable 5, it will need to ensure the model's guardrails can't be circumvented. Security experts say that can't be done. wired.com/story/the-white-ho…
sugnerhan retweeted
I think you very severely understood what I meant by this. A trans woman is not a man on hormones.
sugnerhan retweeted
Right, After 40 years cartoon neuron theories can’t simulate behavior of a simple worm.
Thats because each of those 300 or so neurons has about a billion tubulins in hundreds of microtubules where the cognition (and consciousness) originate.
For example see this paper journals.physiology.org/doi/…
Signaling among neurons causing axonal firings is mediated by megahertz and gigahertz oscillations in microtubules.
The only theory of consciousness with any supportive evidence is Orch OR
academic.oup.com/nc/article/…
A worm with 309 neurons.
Mapped completely since 1986.
Forty years later, no simulation reproduces its behavior.
Joscha Bach (@Plinz) on what that means for BigTech's plan to upload a human brain.
youtu.be/CzjWGkXlK8k
of course they block me from refuting their incredibly silly made-up nonsense. thankfully you are just a vocal minority that will disappear soon
unavailable
sugnerhan retweeted
The paper (partly LLM written btw) is really funny, containing statements such as:
"LLMs lack metacognition entirely"
"RLHF teaches a model to behave as if it held certain preferences, but without creating any internal states that resemble value commitments, desires, or purposes."
"Even simple linguistic inputs may fracture into misleading subword units (e.g., 'therapist' tokenized as 'the rapist'). These are not superficial errors but structural outcomes."
"Humans can refrain from judging; LLMs must predict."
"A 'conclusion' is not the terminus of evaluation, but the terminal state of a stochastic trajectory."
"The model cannot defend a judgment except by generating further statistically plausible text."
"One system parses a world; the other segments a string."
Let me say this clearly: LLMs cannot feel emotions.
Emotions are evolutionary mechanisms.
They push us to avoid danger or approach what is beneficial.
We experience emotions because we are alive, and we want to stay alive.
LLMs are not alive.
Yes, emotional language may be encoded somewhere in the LLM.
Yes, it may even be associated with some LLM output.
But that is just a superficial property.
There is nothing deeper behind it.
For a very simple reason: LLMs do not have an intrinsic and inescapable drive to stay alive.
This is what we call “motivation fault line” in our paper describing seven fault lines between human and artificial intelligence.
*
Paper in the first reply
sugnerhan retweeted
It’s clear that AI will wind up funding universal income. Let’s make that happen ASAP.
how can they write this crap without having even the first clue what the fuck consciousness is in the first place? like for example why are not conscious when the microtubules aren't vibrating?
no wonder no one else wanted their name on it 🤣🤣
imagine emberrasing yourself like this. showing an example that literally doesn't need pretext. middle-ellipting is literally the most common use-case for the requirement of measuring text using js (which is what the lib automates).. absolutely hilarious
