mycall

The deeper the layers you try to edit, the more compressed it gets, and the harder it is edit.

> Ah yes, I forgot about part X

Would something like LMCache help here?

As a native speaker, I read this as "This controls enabling or disabling of all analytic features used by AI model loops"

Now there are times when I let the model keep its own docs in its own language, for its own future reference; but, it is not intended for human distribution due to its odd selection of words and phrases (aka slop).

It makes sense as Polyend's competitor, Synthstrom, has been open source for a while already.

Does the board come with a jtag interface to unbrick the device?

Bring Roland VSynth capabilities to it.

> "next token predictor" has always referred to inference

Going back to first principles, wikipedia mentions [0] inference as:

> "Inferences are steps in logical reasoning, moving from premises to logical consequences. Inference is traditionally divided into deduction and induction"

Is that exactly what next token predictor means, or is it correlated?

[0] https://en.wikipedia.org/wiki/Inference

At this point, the working class, who imagine themselves as possible-future CEOs, will try to extract the same wealth generation out of the tech when possible now that they have seen people get ungodly wealthy through tech. Humans are opportunists.

Of course, there are outliers who use tech individually to improve the human/world experience, but it isn't the norm.

How do you define value?

> strict environment with very clearly defined goals and constraints

This defined a huge set of possible work.

> I've spent around $100-$120k in api calls

Have multiple $200 subs and just use Codex as the intermediary, save yourself $90k-$110k

> paired data is actually hard to come by

Similar to zero knowledge proof (ZNP) and its use cases?

Yeah, LTE is good enough for lower bandwidth applications.

IPv6?

I agree, where world models, e.g. JEPA, JSpace, etc, will have qualities that could be extracted to further compress the DIKW pyramid evolution, but separate from automata processes. It reminds me of interplay of the Human brain is two organisms idea [0].

[0] https://med.stanford.edu/news/all-news/2026/09/two-separate-...

Try web.archive.org

You all are tip toeing around HackerMedia.

> What AI can do at most — is to demonstrate our level of thinking.

It just processes data, so change the data and it will produce different aggregate data, as in y=f(x). We just feed it the internet, so it spits out (mimics) internet as its derivatives. You are only personifying the current state of affairs.

FP in 13 years, gratz

When everyone is unique, nobody is?

Faraday cages aren't good enough to handle ULF/ELF waves. You need much more than a cage for these.

Will LTE continue to work when 6G rolls out in 3 years?

Managers or rather inventors and solution providers? We can do a lot more than just manage now.

Gentoo seems like a nice in between to standard distros and unikernels.

So 3 hours in total for the work? That is not a cheap fee.

What are the options available besides using intermediate signing keys? This seems like a weak link in trusting crypto.

Amiga Toaster and AT&T Truevision were great 93-94 era solutions for video processing.

An LLM could equally decompile the binary, clean up the code, and recode it in one /goal.

Now that it has a nice webpage the describes the obfuscation technics, it is basically cracked at this point to next-gen models.

That assumes nothing can detect or analyze it beforehand.

Isn't this where everything is converging? Binary is more efficient than text for many system, including language.