rpastuszak

We should chat (check bio)! For almost a decade now I’ve been working on a writing tool that might benefit from it. I build occasional forks with procgen on top it. That being said better bidi handling is something I’m most interested in.

Enso.sonnet.io - what I’m talking about.

(On untested.sonnet.io you’ll find notes about the experiments / toys related)

Speaking as someone who managed to get uncle Murdoch to sponsor writing an adblocker/payment tool similar to yours (while I was working at an adtech business), but then got stalled by the publishers (who on avg. hated ads), specifically the direct supply sales people (who tend to have louder voices in those companies, have mortgages to pay) -- the issue is systemic.

ISPs are already selling your data to a bunch of data brokers, I don't think ad spend is the only goal there.

Thanks for sharing. Love the zstd trick. I suspect it still works with older (e.g. bs/python based scrapers) but with puppeteer/chrome not so much...

Nice, what’s your setup? How much work was involved?

FYI I just tried to read your blog but I’m getting the: “Sorry, you have been blocked” banner from CF.

(Safari w. an adblocker, VPN off )

> I run through a $200 Claude Code subscription in less than 24 hours now just for a single project

How?

Nah you're misreading it, it's just a creative interpretation of the famous Kent Beck comment:

"for each desired change, make the change easy, then make the easy change"

Yes! I picked up a similar trick so to speak from Jackson Galaxy.

The main rule for me is this: show that you’re not a threat, don’t get too close but give them a chance to approach you. Curiosity > fear

Unfortunately my dog has too much of a prey drive to tolerate cats, but the silver lining is that he’s quite cat-like anyway.

https://untested.sonnet.io/notes/dog-mode/

I saw it first on a recent Anthropic post, followed by a similar comment thread. There’s some sloppy techbro poetry to it, like when we used to call marketing people growth hackers a decade ago.

How's the sequel now, is it... playable? Good enough to replace CS1 (which I imagine a lot of folks are still playing)?

Let’s wait and see!

Yeah, although, the word “just” is doing a lot of work here.

Any chance you could share a bit more detail? I’d love to try this myself but could use some proven structure / approach.

Who is them and who is us? People are dying and we’ve been tricked to treat this as if it were a football game.

Years ago I used to get a neat (50%) tax discount in Poland for creative / artistic work as a SWE.

Wiping tears with my Portuguese tax return this year.

Humans adapt and will always break the rules, but that’s not something that can be solved with tech. Think: prohibition vs public anti smoking campaigns/healthcare support/and then increasing prices

Culture/social pressures work better than purely technical solutions

Hetzner + Coolify + BunnyCDN for the entire stack

I'm moving off Vercel / AWS. I kept DO for a Minecraft server for friends/family but that's dead.

> the ability to search my old chats easily.

I’d try:

1 exporting my data (I imagine it’s common outside of GDPR?)

2 asking Claude to convert it to an easily digestible format :)

What I find interesting about this + a random bucket of associations because I’ve fallen under the spell of Satan’s Lettuce:

Just a few days ago someone was talking about a machine - human patois.

This (your project) sits somewhere between Lean and BDD cucumber syntax.

At the same time Claude spits out phrases like “a container paying the price of -42px”.

Recently I was listening to a lecture about metaphor in poetry, the misconception that poems are riddles whereas we use metaphors all the time in our language because they convey the meaning more precisely.

> Lay people might talk about traditional computers "thinking" but there's a reason actual computer scientists developed precise language around what they were actually doing instead. > not latch onto some totally unproven equivalence

Precisely. This is akin to my dad, back in 1995, telling me to stop moving the mouse cursor in MS Paint because the computer was "thinking" (he heard the HDD buzzing). That I can understand.

But surely we can expect more from people here than from a carpenter in a tiny mountain village who bought his son a computer in the 90s. (The second one in town!)

It is tiring to write it as well! But, people make real decisions based on unrealistic expectations that LLMs would act like humans.

> It’s not always consistent, but humans have a higher capability of self-awareness. It’s kind of telling that these Claudes don’t seem to consider this pretty obvious failure mode.

We need better words to describe this than "self-awareness" or "consider". These words mean fundamentally different things when speaking about humans or clankers.

This is brilliant! I wish I had seen this a few years back!

Q: Generating higher quality sounds with a specified duration (e.g. within a 300ms margin) -- is it possible with your approach, is it hard, how would you go about that?

I'm asking because a few years back I had an idea for a breathwork/meditation app that would allow you to connect / control the breath duration using a gong sound. Example pattern: 4s in, 4s wait, 8-10s out. Next: use a feedback loop with the real user breathing pattern, slowly extend towards the goal.

Ended up with scaling it down to a "dot to stare at" + basic gong, because I realised how ridiculously chromatically complex the sound is. (my approach was to synthesise an envelope with a shape determined by time -- pity the fool who thought it was easy or even made sense!)

https://untested.sonnet.io/notes/projects-and-apps-i-built-f...

I got +100k people visiting it and still a couple dozen daily users running a 5+ min session, so I'm happy with it. But, the itch is still there, so I'd love to revisit my old nemesis so to speak :)

Also, if you want to chat about this, hit me up sonnet.io/posts/hi !

I can imagine it makes things easier for you but I find it hard to imagine it teaches anyone anything. Just offloading social effort to an LLM (fair enough but let’s call a spade a spade)

Has it gotten any better in the past 6 months on Mac (UI glitches, password management)?

I really want to use it but basic things like password management kept breaking (I even tried switching between different options, like BitWarden - no luck).

I think "through writing" / "write to think/feel" a lot (usually in a mix of 1st/2nd person, 5-6 pages of text per day on avg.). Think: stream-of-consciousness writing, journaling, but also just keeping a screen with disappearing text notes at hand (append-only) where I ramble.

That being said, not many things can compete with talking at your dog for an hour or so.

Another meaning that is becoming more common, esp. among gen-z/booktok folks: overly verbose/poorly structured/flat writing. The example being Duma's Count Monte Cristo.

Yes, although I don't think this effort is futile as you seem to be implying.

A more interesting question is how much effort we're willing to put into making those choices. My answer was posting here and doing a bit of research. Maybe I'll stick with Mullvad for the time being, maybe I won't and I'll help a few people make the same choice. It's better than shrugging and being defeatist, cynicism is cheap.

(Adobe subs weren't https://untested.sonnet.io/notes/alternatives-to-adobe/)

Asking for a friend, who is in a somewhat similar predicament — it wasn’t Portugal, was it?