I can add features and make it work just the way I like it. I've been on IRC since the 90s so there are some idiosyncratic things =)
It's not a massively complicated thing, just a single server that says online 24/7, which talks using a custom protocol to different clients (Web, CLI, Linux native and macOS native).
I could've used pre-existing software to do most of it, but there's always jank and things I can't customise without trying to get changes into the mainline.
There are many IRC clients and this one is mine. I'm not expecting anyone else to use it. It won't have a community, a snazzy web page and it won't be taking in any issues or PRs from anyone =)
That is not the same Walt. I'm assuming you found the TrustPilot site for https://walt.io, which is some Telegram crypto thing from the US. What GP linked to is an independent wallet app that currently supports passes and is scheduled to support tap-to-pay in the near future.
Which is only marginally more effecient than the H.264 encoding that many of them are delivered in. And the biggest shiniest Linux ISOs already use H.265 out of the box.
Tokens are the most basic input unit of an LLM. But tokens don't generally correspond to words or letters, rather sub-word sequences. So Strawberry might be broken up into two tokens 'straw' and 'berry'. It has trouble distinguishing features that are "sub-token" like specific letter sequences because it doesn't see letter sequences but just the token as a single atomic unit. 'Straw' and 'r' are two tokens but an LLM is entirely blind to the fact that 'straw' has one 'r' in it.
As an analogy, I might ask you to identify the relative activations of each of the three cone types on your retina as I present some solid color image to your eyes. But of course you can't do this, you simply do not have cognitive access to that information. Individual color experiences are your basic vision tokens.
LLMs see tokens, not words spelled out with letters.
Imagine verbally asking someone who has never seen written text the same question: unless they memorized the answer for the specific word you're asking about, they'd have to guess.
People assume that's the reason because it's intuitive and "strawberry" is one token. But that doesn't explain why those models would also often get it wrong for "StRaWbErRy" or even "s-t-r-a-w-b-e-r-r-y", where the r's are not combined into one token.
We don't know what was going on inside the closed source GPT models, but this paper investigated on some of the open-weight models and found it's not due to tokenization: https://arxiv.org/abs/2604.00778
And circling back around to AGI, tokenization or some other underlying cause should pose no issue. A competent human would think to write a program (ie create a tool) to do the job. It's routine for a carpenter to make a jig.
i asked opus 4.5 what the problem was and it said it pattern matched too much. i asked it how it should do it, it wrote a file that told itself to stop pattern matching. it wrote a file that started with the following and then had an english language procedure for how to count letters. so it knew the algorithm already, but the "instinct" was to pattern match rather than running the algorithm.
CRITICAL: Do Not Skip Steps
Your instinct will be to "just know" the answer. This is how you get it wrong.
You don't see characters. You see tokens. Your "intuition" about character counts is pattern-matching, not counting. It is unreliable.
You MUST execute this procedure step-by-step, writing out each step visibly.
reply