Hacker News: shmval

New comment by shmval in "Which AI coding tools do AIs recommend? Data across 10 models and 1000 responses"

shmval — Mon, 01 Jun 2026 19:53:40 +0000

I'm running these AI visibility analyses across different niches, the report goes into the methodology details.

AI coding tools shows the disparity between real on-the-ground sentiment and what AI responses suggest much more glaringly compared to my previous reports on observability and sales outreach stack.

Happy to answer any questions.. or take requests for more niches to analyze..

Which AI coding tools do AIs recommend? Data across 10 models and 1000 responses

shmval — Mon, 01 Jun 2026 19:49:06 +0000

Article URL: https://tryrenown.com/research/ai-visibility-ai-coding-tools/full/

Comments URL: https://news.ycombinator.com/item?id=48361716

Points: 2

# Comments: 1

New comment by shmval in "Magnifica Humanitas"

shmval — Mon, 25 May 2026 13:48:23 +0000

The message here obviously comes from a good place. But it can't escape the trap baked into its own theology. Putting humanity one rung below God, with everything else below us is a trick that allowed semitic civilizations to make empires and economies go FOOM faster than the rest of the world. But it also makes things more dangerous when we start building systems smarter than we are.

What happens when the tool outgrows the toolmaker?

Into the Unwritten Dawn

shmval — Mon, 23 Jun 2025 18:24:56 +0000

Article URL: https://dayafter.substack.com/p/into-the-unwritten-dawn

Comments URL: https://news.ycombinator.com/item?id=44358575

Points: 2

# Comments: 0

New comment by shmval in "The Emperor's New LLM"

shmval — Sat, 14 Jun 2025 21:04:28 +0000

The trouble is there is a clear dichotomy in the scenarios where the LLM pushes back like this, most likely derived from RLHF. So anything health related, and "touchy" subjects in general will face this. But random design choices in your code, product roadmap debates etc do not elicit the same resistance.

New comment by shmval in "The Emperor's New LLM"

shmval — Sat, 14 Jun 2025 21:00:00 +0000

Yes but you have to want that, and most people do not. Therein lies the rub.

New comment by shmval in "The Emperor's New LLM"

shmval — Sat, 14 Jun 2025 20:58:48 +0000

This. I think it's the key.

New comment by shmval in "The Emperor's New LLM"

shmval — Sat, 14 Jun 2025 20:56:01 +0000

I think there's a lot of alpha left in building a better and more intuitive UX for seed/top-p/temperature etc. The vast majority of users don't get that far.

New comment by shmval in "The Emperor's New LLM"

shmval — Sat, 14 Jun 2025 20:54:18 +0000

Better prompting does provide more balanced responses to a certain extent but users looking for validation often subconsciously leave bread crumbs that the more powerful models pick up on.