Hacker News: charlieflowers

New comment by charlieflowers in "I know why the mathematicians weep"

charlieflowers — Fri, 16 Oct 2026 00:22:49 +0000

Yeah. "Glorified autocomplete" might be all the human brain is too. "Glorified autocomplete" can do an amazing number of things, including perhaps pass the Turing test and perhaps ... "think".

"Glorified autocomplete" is meant to describe the _underlying mechanism_, but unfortunately it also sounds like one of the things LLMs are actually used for, so it gives the mistaken impression that the claim is all they are good for is a component inside an editor application.

New comment by charlieflowers in "I know why the mathematicians weep"

charlieflowers — Fri, 16 Oct 2026 00:18:26 +0000

All my life, when I would think back to what mill towns went through when they all closed (something I did not live through), all my naive outsider brain could arrive at was, "Yeah, but those were just grunge jobs, humans were freed up to do better stuff."

And perhaps the same thing is now happening to me. It feels different, but maybe the same thing is true.

New comment by charlieflowers in "Nix wrote half of my debugger"

charlieflowers — Mon, 12 Oct 2026 22:15:28 +0000

Is the high barrier to entry essential though? Is there something about the nix domain that would mean any solution would have as high a barrier to entry? Or is it something they could have simplified greatly, and just never did? (Asking as an outsider, I've always been curious about this).

New comment by charlieflowers in "Iranian campaign planted fake articles in real U.S. publications using ChatGPT"

charlieflowers — Fri, 09 Oct 2026 16:22:23 +0000

Resolved it was on my end

New comment by charlieflowers in "Iranian campaign planted fake articles in real U.S. publications using ChatGPT"

charlieflowers — Fri, 09 Oct 2026 15:57:05 +0000

Edit: Resolved now it was on my end. > fyi not working for me

New comment by charlieflowers in "US man given prison sentence for bot-farming music streams"

charlieflowers — Thu, 08 Oct 2026 20:24:43 +0000

Surprised it is not "Super Intelligence of America"

New comment by charlieflowers in "I-have-ADHD: A skill to stop coding agents from burying the answer"

charlieflowers — Tue, 08 Sep 2026 18:49:08 +0000

That tension is doing real work.

New comment by charlieflowers in "Dolly Parton has died"

charlieflowers — Tue, 25 Aug 2026 19:32:17 +0000

I still remember how hard I was struck by the scene in The Orville, where someone from another planet hears a Dolly Parton song playing and asks who it is. He goes on to say, "She speaks with the might of a hundred soldiers!"

She did. She is one of our species' great poets.

New comment by charlieflowers in "GitHub is the wrong shape for this new world"

charlieflowers — Thu, 30 Jul 2026 03:45:28 +0000

But do we really need a paradigm shift in how we express the concept of a paradigm shift? It's not like the meaning of the words expired.

New comment by charlieflowers in "Who's afraid of Chinese models?"

charlieflowers — Tue, 21 Jul 2026 14:18:59 +0000

But inference costs scale per task whereas platform services like postgres typically amortize across tasks. If you are selling tasks done by inference, then compute is part of your COGS and it scales per task.

New comment by charlieflowers in "Gossamer: a Rust-flavoured language with real goroutines and pause-free memory"

charlieflowers — Sat, 27 Jun 2026 00:56:16 +0000

Originally I replied here that I thought you were both missing the point (but I was wrong).

I wrote: It has a garbage collector and goroutines, so clearly it is not trying to be a systems programming language.

Then ... I saw that it does indeed pitch itself as a systems programming language. So, I guess you both are right.

If Gossamer were to drop that claim, then I'd say it looks impressive to me. I have often wanted this particular mix of language features.

New comment by charlieflowers in "Gossamer: a Rust-flavoured language with real goroutines and pause-free memory"

charlieflowers — Fri, 26 Jun 2026 20:56:08 +0000

The language choices demonstrate excellent taste though.

New comment by charlieflowers in "I rewrote PostHog's SQL parser, 70x faster, while barely looking at the code"

charlieflowers — Thu, 25 Jun 2026 16:56:37 +0000

I interpreted it as referring to a table named "FROM" and fields named "SELECT", "WHERE" and "AND".

New comment by charlieflowers in "DuckDB Internals Part 1"

charlieflowers — Fri, 19 Jun 2026 17:52:38 +0000

BTW, columnar is very similar to struct of arrays (SOA) and some of the reasons it works well overlap with SOA.

New comment by charlieflowers in "US President says 'I love the inflation'"

charlieflowers — Wed, 10 Jun 2026 22:22:15 +0000

But shhhhh! Iran might read this!

New comment by charlieflowers in "Zig Structs of Arrays (2024)"

charlieflowers — Tue, 09 Jun 2026 14:11:35 +0000

I’m not advocating this, but it is worth observing that it is yet another problem one could attempt to address with dependency injection, similar to io and allocators.

New comment by charlieflowers in "Higher usage limits for Claude and a compute deal with SpaceX"

charlieflowers — Wed, 06 May 2026 18:41:49 +0000

Help me understand why not? I know solar power generation in space, and "beaming" the power back, was a naive idea. But this would actually use the power up there, mostly for training, but also for inference.

That claim seems reasonable. I have zero knowledge of the economics of launching and maintaining satellites though.

New comment by charlieflowers in "Higher usage limits for Claude and a compute deal with SpaceX"

charlieflowers — Wed, 06 May 2026 18:38:28 +0000

Not a merger, right, unless I missed something (admittedly skimming).

New comment by charlieflowers in "The bottleneck was never the code"

charlieflowers — Wed, 06 May 2026 17:34:45 +0000

I have been thinking about this a lot lately. How do you capture key factors succinctly, and even harder, keep it succinct as it evolves?

The shrinking that property based testing does when it finds an issue is kind of what we need for specs/context.

New comment by charlieflowers in "What async promised and what it delivered"

charlieflowers — Sat, 25 Apr 2026 19:20:43 +0000

The simplest example is that you can easily be wasteful in your use of threads. If you just write blocking code, you will block the thread while waiting on io, and threads are a finite resource.

So avoiding that would mean a server can handle more traffic before running into limits based on thread count.