Hacker News: bridgettegraham

New comment by bridgettegraham in "Statement on US government directive to suspend access to Fable 5 and Mythos 5"

bridgettegraham — Sat, 13 Jun 2026 03:47:55 +0000

man the us govt is becomming a proper bully in the truest sense of the world. dictatorship, censorship and obfuscation of truth. 1984 "the truth is what the dictator says it is" and the sheep are too dumb to realize it.

New comment by bridgettegraham in "Statement on US government directive to suspend access to Fable 5 and Mythos 5"

bridgettegraham — Sat, 13 Jun 2026 02:10:53 +0000

"i can has the bestest powers of the world, i is the strongest, i is the badest president ever" what a retard

New comment by bridgettegraham in "Statement on US government directive to suspend access to Fable 5 and Mythos 5"

bridgettegraham — Sat, 13 Jun 2026 02:09:40 +0000

this is just the US government bullying everyone wherever they can because they "are the bestest government that has ever goverend" ugh puke. the US govt and its leader is a typical schoolyard bully and I wish someone could stop that bully. i hate that the govts have so much power.

New comment by bridgettegraham in "Why does it look like LLMs consistently overestimate implementation time?"

bridgettegraham — Fri, 22 May 2026 10:10:10 +0000

yeah this is my suspicion too

New comment by bridgettegraham in "Why does it look like LLMs consistently overestimate implementation time?"

bridgettegraham — Fri, 22 May 2026 10:09:47 +0000

yes exactly - I have never asked them for an estimate

Why does it look like LLMs consistently overestimate implementation time?

bridgettegraham — Fri, 22 May 2026 01:26:22 +0000

I have my suspicion: they estimate how long people would have taken to implement some feature, becasue they were trained on such data. I consistently see estimates of 2 week/3 weeks or 5 days, etc. But then implementation takes a day or 2 max using agents within Claude/GPT. Unless I am missing something? Anybody else notice this?

Comments URL: https://news.ycombinator.com/item?id=48230861

Points: 4

# Comments: 14

The NPCs That Test Themselves

bridgettegraham — Tue, 14 Apr 2026 13:00:55 +0000

Article URL: https://aarils.com/personal/the-npcs-that-test-themselves

Comments URL: https://news.ycombinator.com/item?id=47765112

Points: 3

# Comments: 0

The Expensive Anxiety of AI

bridgettegraham — Mon, 13 Apr 2026 04:25:34 +0000

Article URL: https://aarils.com/personal/the-expensive-anxiety-of-ai

Comments URL: https://news.ycombinator.com/item?id=47747590

Points: 2

# Comments: 0

New comment by bridgettegraham in "Show HN: Yesterday's Bread – a modern MUD with AI NPCs and a personalised novel"

bridgettegraham — Tue, 07 Apr 2026 16:05:19 +0000

Thanks for the comment! I like your idea about changing the voice to first/second/third, and I am also considering full translation - this has not yet been built. The game is built with locales in mind, supporting English and Afrikaans (my native language) out of the box and I plan to add more locales, but I still need to investigate this.

Oh I remember AI Dungeon - I have played it myself :-)

Show HN: Yesterday's Bread – a modern MUD with AI NPCs and a personalised novel

bridgettegraham — Tue, 07 Apr 2026 13:44:47 +0000

Hi HN — I'm Bridgette, a software engineer in South Africa. Over the last year I rebuilt a classic-style MUD from the ground up on a modern stack (Postgres, server-authoritative simulation, generative-AI NPCs that keep schedules and remember conversations, dynamic weather, item decay, the works). The twist is that when you finish the single-player arc, the game compiles your run into a personalised PDF novel — your name, your race, your choices, your ending. I'd love technical feedback on any of it: the simulation, the NPC design, the localisation engine (it's bilingual English/Afrikaans), or just the idea of "the game writes you a book about you" as a finishing reward. Happy to answer questions.

Comments URL: https://news.ycombinator.com/item?id=47675292

Points: 2

# Comments: 2