CuriouslyC

5 ms·

31.
by CuriouslyC 24d ago
There will always be communities of people who want to handcraft everything, who will be available to commiserate about the struggle of their craft. We still have people doing crazy shit like collecting/smelting iron by hand and smithi
32.
by CuriouslyC 24d ago
The nuance here is that people aren't good at evaluating their own work/progress, and use recognition as a proxy milestone of their own improvement. No matter how good they get, they don't "feel" good until someone
33.
by CuriouslyC 24d ago
A CNC machine will beat a whittler every time. If a craft is only about technical execution, AI is going to own it. The crafts that are safe are the ones that elicit/depend upon a complex human response that's not easily predictab
34.
by CuriouslyC 24d ago
This article's note about game making 100% wrong, games are highly "complex" and there's a lot of iterative refinement and experimentation that is involved in "finding the fun." You can't just throw a pot
35.
by CuriouslyC 25d ago
I don't care about AI bands and wouldn't listen to the music by itself for pleasure, but for video game music? Or soundtracks to sub-AAA/hollywood productions that can't afford someone who's won an Oscar/Golden
36.
by CuriouslyC 25d ago
Given the witch hunts around AI creative output, nobody who has put in work to avoid the stigma is going to out themselves.
37.
by CuriouslyC 25d ago
Lots of people still make bad music that other people still manage to enjoy (a lot of it has gone multi-platinum!) even though they're not Mozart or Bach.
38.
by CuriouslyC 25d ago
I started Valknut ( https://github.com/sibyllinesoft/valknut ) when I saw the writing on the wall regarding Agent code structure/abstractions/etc being a limiting factor in the ability to autonomously build pro
39.
by CuriouslyC 26d ago
Imagine any Ansel Adams picture was taken by a monkey. Are they suddenly trash? Imagine your favorite haiku was instead created by a parrot mashing together overheard snippets from its owner. Again trash? The argument gets frighteningly lit
40.
by CuriouslyC 26d ago
I was pointing out that your "art requires intent" statement is built upon a house of cards as written.
41.
by CuriouslyC 26d ago
You're baking a lot of philosophy into "creativity requires intent." We can't even prove qualia are universal, which is a prerequisite for "intent" beyond "neurons firing" (what about hypothetical al
42.
by CuriouslyC 26d ago
The issue is discovery and perverse incentives. Search algorithms have needed a major overhaul for a while, and previously the existence of a thing was a proxy for effort, which some people use as a proxy for value (dizzying, I know). Now p
43.
by CuriouslyC 26d ago
Anthropic is shady too my guy. Claude Code will frequently re-ask you to turn on data sharing even after you've said no, it's dark patterned like a task approval, and once you turn it on it's harder to turn off than OAI makes
44.
by CuriouslyC 28d ago
Instagram seems like their most engaged property. FB seems to be devolving into groups + marketplace, and WhatsApp gets a lot of play but I don't think it's the best ad serving/bot surface. I don't think Insta alone can
45.
by CuriouslyC 1mo ago
> 1. It's usually low value and not trustworth - I could have just asked the AI myself. You have no way of knowing what the input to the writing was. You're making the erroneous assumption that the person posting the article wa
46.
by CuriouslyC 1mo ago
You're marginalizing yourself. I don't have hard data for writing, but I do for another area: YouTube Thumbnails. AI generated thumbnails outperform human thumbnails, often with a +2-5% delta in CTR. Yet "so many" people
47.
by CuriouslyC 1mo ago
The part about false positives was telling. The author doesn't want to demonize human trash, that's not fashionable, they're only concerned about virtue signalling.
48.
by CuriouslyC 1mo ago
Some people are still losing their shit over em-dashes, with no other tells, and humans can't use the not x; y construction anymore either, regardless of any other merit to the writing. LLM writing is verbose and meandering, but people
49.
by CuriouslyC 1mo ago
Ending with a spin on "if you didn't bother to write it why should people bother to read it." Way to rail on cliche repetitive slop, with more cliched slop. Will the irony never cease.
50.
by CuriouslyC 1mo ago
Ironically the Chinese are much more science pilled than we are, in part because the party and government have a higher proportion of scientists and engineers. Culturally this biases them to open sharing of information and likely against st
51.
by CuriouslyC 1mo ago
I tell Astra to plan, delegate and check the work of GLM 5.3-flash and Terra medium/high subagents now.
52.
by CuriouslyC 1mo ago
Code review is soon to be an outmoded concept, (un)fortunately. You have to design orthogonal code (e.g. independent modules in a modular monolith, or microservices) and soak test using canaries.
53.
by CuriouslyC 1mo ago
Specialized AI code review software is so pointless though. Back when agents were dumb about git surgery and tool use it might have had a purpose, but now you could replace coderabbit with a skill and I bet the results would be better in so
54.
by CuriouslyC 1mo ago
This is true, but with the newer generation of models you don't want to do this yourself, they're really good at orchestrating and triage. Run Fable or Astra on low/medium, and tell them to come up with a plan then direct sub
55.
by CuriouslyC 1mo ago
The timing is related to the fact that their benchmark was saying it was the same as Sol, and below Opus 5, when anecdotal reports and other benchmarks strongly disagree. It looked bad for them for their benchmark to disagree with people&#x
56.
by CuriouslyC 1mo ago
Might wanna have some coffee then browse my comment history.
57.
by CuriouslyC 1mo ago
I'm having no issues, I think in their case it's just load from people falling back from the broad frontier failure.
58.
by CuriouslyC 1mo ago
Good time to have a backup GLM 5.3 plan.
59.
by CuriouslyC 1mo ago
Ed's right about OpenAI being out over its skis, he's right about the numbers for everything being wildly optimistic and he's right about the circular financing/leverage aspect of a bubble. He hates AI though so he can&#
60.
by CuriouslyC 1mo ago
Claude was really far ahead of GPT in writing from 2-4, but the later models have started to get an overly distinctive style, wheras these days GPT tends to be coherent but fairly concise and dry.

More ›