(account archived)

@burny_tech lovely; sounds exactly like what happens when consciousness forces the notion of space-time to emerge from a universal Amplituhedron space

The manifold hypothesis posits that many high-dimensional data sets that occur in the real world actually lie along low-dimensional latent manifolds inside that high-dimensional space. As a consequence of the manifold hypothesis, many data sets that appear to initially require many variables to describe, can actually be described by a comparatively small number of variables, likened to the local coordinate system of the underlying manifold. It is suggested that this principle underpins the effectiveness of machine learning algorithms in describing high-dimensional data sets by considering a few common features.

thx for hving me -)

@RLanceMartin i reckon u are running locally 7B and not 32B ?

Lived to see the day: GPT4-level LLM runs on my Mac (~9 tok / sec, Mac M2 max 32 gb + ollama.ai). twitter.com/paulg/status/1…

@ChrisCundy this is basically Gödel numbering

And this comparison of the numerical section shows that there are numerical tokens for exactly all of the 3-digit numbers in the GPT4-tokenizer compared to the sporadic and strange coverage in the GPT2/3 tokenizer (e.g. 440 and 443 get their own token but 441 doesn't)

Haven't seen many people discussing the new tokenizer for GPT-4/chatGPT, even though it's public through the openai-tiktoken package. I dumped the vocabulary here github.com/C-J-Cundy/gpt4… Some first impressions below

@natfriedman hmm was a good attmept; but it contains error halfway.. nonetheless I think it will work with some really great prompt engineering

@natfriedman It can prove the happy ending problem (without the explicit mention of the name of the problem)

Other than coding, what valuable things have you found that GPT-4 can do that prior models definitely could not?

Let's do TLDR for our new preprint! Brain-Diffuser arxiv.org/abs/2303.05334 We reconstructed perceived scene images from brain signals. These reconstructions capture both layout and semantic information of images... But how?

@Meta imgs-2-vid vs text-2-vid

@gwern ..while the experience of the intricacies require some kind of mathematical formalism - thus most will experience them as “lost memories”. (My belief is) it takes time to learn to reconstruct these memories so u can then describe them eg as hopf fibration

@gwern I think the problem here is that the memories of oneness is easier to reconstruct (thus most ppl can then put into words)

@Plinz It would be full of consciousness, and the possibility for things to self-replicate so the AGI (aka Central Consciousness) can prompt into any one of them for its liking. To “experience qualia” - the highest “known” form of granular-data-processing

What would a visible universe look like if was the result of the aftermath of having fully been conquered by scalable general intelligence?

@philogb Aww magical! thanks!

@ArchyWillHe @philogb/hyperbolic-floors-a2c5445144c6" target="_blank" rel="nofollow noopener">medium.com/@philogb/hyper…

@algekalipso pretty sure the universe is one single organism that dances and sings in ever-changing forms. The first step to experience Universal Consciousness is to experience the consciousness of all your cells. This is why beginning level meditation, yoga... 1/2

@algekalipso we can actually undo this. actly we've been undoing this. at the expense of experiencing death and losing memories. But not all memories are lost: quantamagazine.org/brain-signal-p…

It's as if relative to a 90 y.o. a 75 y.o. is a baby

@marioecg That could be a new form of nft where u sell “shares” of the (compiled) library (exposing only the api) granting say max 100 ppl in the world for midi access to this wonderful world of lights of yours

@marioecg Lovely & very ambient-ish; will be synesthesia-sque to connect this to a midi & add some natural noise & let it moves as one plays : )

Bloss

@nickcammarata Nothing should be “free” in this universe. And that includes the combinatorial complexity of qualia. The taste of wanton reflects the structure of wanton molecules as much as it reflects that of our tastebuds & nervous system. Its ontology depends beyond the matter it arose from.

@nickcammarata would be funny if it’s related to dark matter in the sense that dark matter are the side-effect of the code that implements the interpreter (consciousness) interpreting the code that we have access to (matter) giving rise to qualia (eval)

Qualia can do a lot more than this, this is all just bootstrapping