My pc’s not really feeling it today. Second bluescreen already 🥴

GM 🫡

First person to make me laugh out loud gets 100k sats.

Entirely subjective.

Just gonna read through the comments and the first irl lol gets the sats.

The cool thing about this spam filter is that the LLM provides an explanation of why it thinks something is spam, which i can include as a local reply to them email after my workflow marks it as spam.

The workflow so far is:

1. Feed in raw email -> generate analysis of if its spam or not

2. Feed in analysis -> get LLM to generate a single SPAM or NOTSPAM token by reading the analysis and making a decision

I then check for this SPAM token and mark the email as +spam and/or +analyzed with notmuch.

the tagging potential of this beyond just spam is huge. best of all its all local AI, so no need to feed my person emails to the openai machine.

nostr:note13edqfznfkud7t48crkhu8y4h938jwftcekg7nk6hwk9366djta7s8lr7tr

I was at my local bakery the other day, and I saw an electrical heater in the shop… It bothered me because I was thinking to myself that they could use a miner as a spaceheater too…

I kinda want to do the talk with them and try to convince them. But how on earth would you convince someone who has no clue about Bitcoin to buy a #bitcoin miner in order to heat up their shop? Buying Bitcoin already seems crazy to some people, let alone buying a (expensive) machine that mines bitcoin and actually heats up a room too.

What’s your take on this?

#asknostr

i do this all the time on CLN. it's fun

I'm at 60s now and it feels amazing. I get an endorphin rush every morning for free.

big parser doesn't want you to know that building parsers is actually incredibly easy. for some reason parsers are scary to programmers.

here's a parser I just built that takes a list of text tokens out of visa pdf statements and converts them to ledger-cli (plain text accounting) format... because apparently csv exports from my bank is too much to ask for.

the basic idea is that you have one large array of strings extracted from the pdf. this is an unreadable stream of unstructured text chunks called tokens.

for each token you see if you can parse the next N tokens into a transaction line item. if it fails you just advance the parser by one and try again. if it succeeds then the parser would have eaten all of the tokens it needed so it can continue from there.

you can build this kind of parser in *any* language, it would look identical. it's just a few simple functions and data.

after doing these kinds of things at a record label for 6 years I'm convinced that accounting is more of a software engineering job, can't imagine going through pdfs manually each month.

I'm not sure what the point of this post was other than to say I like parsing things and that banks suck.

carry on.