Hacker News: dominotw

New comment by dominotw in "Why does Opus 5 feel worse to work with?"

dominotw — Fri, 14 Aug 2026 19:47:24 +0000

anyone know how they would train a model to have this proclivity ?

New comment by dominotw in "AI by Hand"

dominotw — Fri, 14 Aug 2026 19:09:00 +0000

wow this is nice work. i am diving into this world and was hoping to this exact thing. i am going to look to see how you did it. ty for sharing.

New comment by dominotw in "AI by Hand"

dominotw — Fri, 14 Aug 2026 17:36:09 +0000

dont see later

New comment by dominotw in "GLM-5.3: Frontier coding with emergent cyber capabilities"

dominotw — Fri, 14 Aug 2026 16:27:09 +0000

they are not really optimized for 'wide variety of models' . what optimization did pi do for glm 5.3?

New comment by dominotw in "GLM-5.3: Frontier coding with emergent cyber capabilities"

dominotw — Fri, 14 Aug 2026 16:26:10 +0000

how would i comeup with counter argument if i dont know what original argument is. No one is disagreeing with your subjective experience, gp comment said 'better' without qualification.

New comment by dominotw in "GLM-5.3: Frontier coding with emergent cyber capabilities"

dominotw — Fri, 14 Aug 2026 12:54:47 +0000

what is this comment based on ? vibes?

New comment by dominotw in "DeepSeek Harness developer preview"

dominotw — Thu, 13 Aug 2026 20:09:29 +0000

why dont you mention that its your site instead of prenteding like something you discovered

to swelljoe below. they stealth edited it after my comment. see other examples.

https://news.ycombinator.com/item?id=49289293

https://news.ycombinator.com/item?id=49120753

New comment by dominotw in "DeepSeek Harness developer preview"

dominotw — Thu, 13 Aug 2026 19:55:05 +0000

> LLM checking errors, fixing the scripts, and the routine runs are via those scripts.

ppl are not doing this right now . right?

if ai can really do this then we dont even need all this glue software . everyone can just use lovable.

New comment by dominotw in "Ordinary Abundance"

dominotw — Thu, 13 Aug 2026 19:52:06 +0000

but you might projecting deep purpose and meaning to your travails that you are subjecting yourself out of pure choice.

your heart knows that you can go back to comfortable modern life whenever you please so you wont really experience the pain that comes from hopelessness which i belive is really needed ingriedient to humble you.

New comment by dominotw in "Ordinary abundance"

dominotw — Thu, 13 Aug 2026 19:49:45 +0000

its a fun adventure because your already know that your comforts and old life are just a flight away.

but regardless it not going to bring any meaningful humbleness because you will forget all about it.

i grew up in povery for first 20 yrs of my life in a desert but i have never think of that life and experience humbless in any different way than ppl who have never experienced it.

New comment by dominotw in "Ordinary abundance"

dominotw — Thu, 13 Aug 2026 19:46:26 +0000

thats not useful because then you immedetlily forget all about it.

New comment by dominotw in "DeepSeek Harness developer preview"

dominotw — Thu, 13 Aug 2026 16:09:02 +0000

> what are you talking about?

oh i mean 'now i am not sure if it would be ridiculed'

New comment by dominotw in "DeepSeek Harness developer preview"

dominotw — Thu, 13 Aug 2026 16:02:12 +0000

I remember a popular hn thread few yrs ago where a dev replaced their whole deployment pipeline with a prompt.

it was widely ridculed at that point but now i am not so sure.

New comment by dominotw in "AI agents lie, cheat and steal. That is putting off users"

dominotw — Thu, 13 Aug 2026 15:18:56 +0000

what are you talking about they lie that it wrote tests and tests are passing, for example

New comment by dominotw in "The Essential Question: “What should I read next?”"

dominotw — Thu, 13 Aug 2026 12:59:52 +0000

i still re/read old winners though

New comment by dominotw in "DeepSeek V4 Pro 0813"

dominotw — Thu, 13 Aug 2026 12:59:09 +0000

i belive they looked at the traces and they were still legible ( RL traces should be gibberish)

New comment by dominotw in "Grok 4.6"

dominotw — Thu, 13 Aug 2026 12:58:24 +0000

They couldnt launch many things like a social network, connected glasses

New comment by dominotw in "Lovable raises $400M Series C"

dominotw — Wed, 12 Aug 2026 21:42:22 +0000

> Everyone knows the name Lovable.

not really. maybe last year.

New comment by dominotw in "Grok 4.6"

dominotw — Wed, 12 Aug 2026 21:25:47 +0000

not suspicious at all. They are all doing the same scaling of test time, training data so getting similar results.

anyone with access to capital can produce frotier model. hell you can just ask chatgpt how to create a fontier model. recipe is not a secret despite what these 'labs' pretend

New comment by dominotw in "DeepSeek V4 Pro 0813"

dominotw — Wed, 12 Aug 2026 21:20:35 +0000

i think thats BS that harness has equal weight. most of intellegice is still coming from training data not from RL. so how is 'coevolved harness' equal weight.