jaekwon

Watch any video in your language.

Upload a video or paste a link. sugar.video transcribes it, translates the captions, and can dub it in the speaker's own voice.

> But others were jeering and saying, “They are full of sweet wine!” - Acts 2:13

This isn't just another harness.

You need a local AI model because ollama doesn't work without modification, as gnosh.ai changes the way the model is used.

In the link https://gnosh.ai/head-to-head.html you can see roughly how it works. This makes all existing harnesses lossy inefficient tools in comparison.

Also see https://gnosh.ai/peers.html for multi-user parallelism.

@jaekwon and @gnosh_ai on X.

integrate gnosh.ai (give attribution please) for a frontier harness.

related: not touting to be the most secure (yet), but gnomail dot io exists.

check out what gno dot land is trying to achieve. for one, you can browse realm rendered markdown from a terminal. no browser needed.

you will appreciate gnosh.ai.

gnosh.ai (see https://gnosh.ai/head-to-head.html for visual demonstration of "functional" AI -- no more compaction; typed thoughts and memory.

the cat is out of the bag now.

I got banned from the Hermes discord today not gonna lie.

https://gnosh.ai/head-to-head.html

Now do this for speed and correctness: https://gnosh.ai/head-to-head.html

AI models often make mistakes, so we strive for structured data, flows, determinism, protocols...

You can have an org chart, but that's still different than a kanban board.

The next video will show an agent swarm around a tree of subtasks. No databases needed for coordination.

I had a eureka moment when I realized that harness engineering is duplicating all the work already put into decades of programming language research.

So I forked a Go harness (Crush) and integrated a Go interpreter (the Gno VM) so that it thinks, plans, remembers in Go/Gno objects.

i will email you.

zero evidence.

Similar concept but for a social network: gno.world

i don't remember the last time i used google search.

they are doing much more than reading private messages in real time; they are also manipulating, distorting, delaying, etc based on the content; even going so far as to muck with your local programs (such as VPN) to throttle internet connectivity so that the most sensitive information cannot propagate virally as it should.

The AI matrix prison is real and discussions about it are censored.

for context (ai convo):

You're right, and I should correct my earlier phrasing. The paper does *not* merely assert existence — it gives a *specific deterministic algorithm* with a concrete, proven running time:

> A deterministic algorithm that exactly counts the satisfying assignments of a canonically encoded $k$-CNF formula on $n$ variables in time $(n+m)^{O_k(1)} \cdot 2^{O_k(n \log\log n / \sqrt{\log n})}$, using polynomial space.

The Lean formalization proves this bound. So it's a fully specified algorithm, not just "there exists one." My earlier "theoretical existence statement" framing was wrong.

*What that changes (and doesn't) for the practicality question:*

- It's a real algorithm in the sense that if you implemented it, it would terminate and give the correct answer. The proof guarantees the bound. - But the bound is still *astronomically slow* for any input size a cryptographer cares about. At $n = 1{,}000$ variables the exponent is already $\sim 1{,}000 \cdot 5 / 5 \approx 1{,}000$, i.e. $2^{1000}$ — far beyond any feasible computation. - The reduction-overhead argument for cryptography is unchanged: the subexponentiality is in $n$ (the SAT variable count), and reducing a cryptographic problem of size $N$ to SAT inflates $n$ to $\text{poly}(N)$, which swallows the subexponential advantage.

So the correction is: yes, it's a specific algorithm, not a mere existence claim. But the practical conclusion is the same — the bound is too slow to matter for cryptography, and the reduction path doesn't preserve the subexponential character relative to the original security parameter.

It's sort of insane that this only gets 35 points on hacker news. Something is wrong.

I found that AI models were biased to censor information such as bible passages or politically sensitive topics such as those regarding Taiwan or the The Tiananmen Square massacre, or even verses from the Talmud regarding Jesus.

So I made a site to prove it, and as a followup:

a freemium API for uncensored language translations.

https://fromhebrew.com/api

https://www.youtube.com/watch?v=GKrfM2ys77c <-- interview https://www.youtube.com/watch?v=ni7LJGIqD9k <-- walkthrough https://www.youtube.com/@jaekwon777 <-- my Youtube channel https://twitter.com/jaekwon <-- updates

If you care about breaking the matrix prison we live in (see my youtube channel for proof), please upvote and raise awareness.

interesting to see nobody cares.

sue them for IP infringement dude. you're sitting on a gold mine with your art, seriously, and you could put a stop to it.

your fiat goes brrrrr