frontpage.
newsnewestaskshowjobs

Made with ♥ by @iamnishanth

Open Source @Github

Open in hackernews

Ask HN: Would an AI chatbot that searches guides/wikis in-game be useful?

1•weizhenchu•3h ago
would an in-game AI chatbot that (1) answers quick guide questions (best gear, boss tactics) and (2) opens/pins wiki pages on top of the game help you look up info while playing?

I feel like it could save time compared to watching multiple YouTube videos, but I haven't seen a tool that combines both functions widely used by players yet.

when I was playing Monster Hunter I asked ChatGPT for a gear recommendation — some of its suggestions differed from what I saw in videos. That made me wonder how much people would trust an AI versus curated wiki/video sources.

A few things I'm curious about:

• Would you trust an AI's build/gear suggestions, or prefer direct wiki citations? (Would video citations help?)

• Would you want an AI companion that cheers you the first time you take down a boss, or just chats while you play — or would that get annoying?

Comments

yuanchuangAI•2h ago
I totally get that need! Especially with games like Monster Hunter that have so many tiny details—when you’re fighting a boss, either you tab out to check the wiki and ruin the flow, or you skip through a YouTube video for ages just to find gear recommendations. It’s such a waste of time, honestly. Let’s start with those two features you mentioned—AI answering guide questions + a pinned wiki at the top of the screen. If those two could work together? Total must-have. When I was playing World and fighting Nergigante, I asked ChatGPT for a Great Sword build. It recommended this Beta set, saying it had high defense and good affinity. So I made it, but when I got to the boss fight? I had no Earplugs. Got stunned by its roars nonstop and couldn’t even combo. Later I checked the wiki and realized—new players fighting Nergigante actually need the Alpha set for Earplugs + Health Boost 3. The AI didn’t even consider I was a new player just moving up to High Rank; it just pushed the “theoretically best” set. But imagine if the AI gave a simplified tip first—like “New players, prioritize Earplugs + Health Boost. If you’re short on materials, make XX piece first”—and at the same time, pin the wiki’s “Nergigante Exclusive Builds” page at the top, with its weak spots and dodge timings highlighted in red. Way faster than skipping through a YouTube video where the YouTuber rambles for five minutes with an intro. AI + wiki cuts right to the point—it’d be awesome. Now for the trust thing: For me, it’s always “AI for the basics, wiki for the specifics.” The good thing about AI is it can break down complicated stuff simply. Like, if you ask “How do I do a Charge Blade SAED?”, it can sum it up in one line: “Build phials → activate Red Shield → Charge Slash into SAED.” No need to dig through a bunch of jargon on the wiki. But when it comes to specific stats—like a piece of gear’s skill numbers or where to get materials—I always check the wiki. AI mixes up versions all the time, y’know? After Rise: Sunbreak updated, some old gear got nerfed, but the AI still talked about it like it was the old version. The wiki updates right away, though. As for videos? They’re only useful for gameplay—like “How to dodge Zinogre’s pounce”—watching a YouTuber show it is clearer than text. But for gear recs? Videos aren’t as accurate as the wiki, and way slower than AI. And that AI buddy cheering/chatting? The key is “don’t interrupt.” When you beat an Elder Dragon for the first time, if the AI pops up with “That SAED was chef’s kiss! Finally took it down!”, that’d be cool—feels like having a buddy playing with you. But if it cheers randomly when you kill a small Velociprey, or keeps popping up asking “Wanna talk about the story?”, that’d be annoying enough to make you turn it off. It’d be perfect if you could manually adjust the “interaction level”—mute it during boss fights so it only shows guides, drop a quick encouragement after key moments, and only chat when you ask it actively. That way, it doesn’t feel forced. One thing I’ve wondered too: What if the AI could tailor recommendations to my game progress? Like, if I just unlocked MR1, don’t recommend endgame builds for MR50. If I just learned how to do SAED on Charge Blade, don’t start rambling about “how to combo SAED with Guard Point” (that’s next-level stuff!). Right now, AI always goes “too big too soon.” But if it could link up with my game data—like reading my save file, seeing I’m short on materials, and recommending “transition gear you can make with what you have”—plus the pinned wiki? That tool would definitely blow up. At the end of the day, players don’t need the “perfect answer”—we need answers that work right now.

Demand-Side Platform

https://en.wikipedia.org/wiki/Demand-side_platform
2•doener•8m ago•0 comments

Grok chats exposed in Google results

https://www.bbc.com/news/articles/cdrkmk00jy0o
2•mdhb•8m ago•0 comments

When People Giggle at Your Name, or the 2025 Hugo Awards Incident

https://grigorylukin.com/2025/08/21/when-people-giggle-at-your-name-or-the-2025-hugo-awards-incident/
1•Rokesmith•13m ago•0 comments

The Minecraft code no one has solved (2024) [video]

https://www.youtube.com/watch?v=nz2LeXwJOyI
2•zichy•14m ago•0 comments

Mitigating Backpressure from High Join Amplification with Unaligned Joins

https://risingwave.com/blog/unaligned-joins-risingwave/
1•Sheldon_fun•17m ago•0 comments

Trump administration to vet all 55M foreigners with U.S. visas

https://www.washingtonpost.com/national-security/2025/08/21/us-visa-vetting-foreigners-immigration-tourism/
6•KnuthIsGod•19m ago•0 comments

Saneject: Dependency Injection the Unity Way

https://github.com/alexanderlarsen/Saneject
1•msk-lywenn•19m ago•0 comments

Insights from 100 Years of Research with Probiotic E. Coli

https://pmc.ncbi.nlm.nih.gov/articles/PMC5063008/
2•luu•22m ago•0 comments

Too Much of a Good Thing: How Genericide Sends Trademarks to the Graveyard

https://uclawreview.org/2023/03/21/how-genericide-sends-trademarks-to-the-graveyard/
1•thunderbong•23m ago•0 comments

What the Hell Is Going On?

https://catskull.net/what-the-hell-is-going-on-right-now.html
3•todsacerdoti•24m ago•0 comments

Do blogs need to be so lonely?

https://thehistoryoftheweb.com/do-blogs-need-to-be-so-lonely/
1•Brajeshwar•25m ago•0 comments

Ryan Dancey on the Acquisition of TSR

https://www.insaneangel.com/insaneangel/RPG/Dancey.html
1•Michelangelo11•29m ago•0 comments

My first project in Go is a terminal dashboard (what a programming language)

3•vinserello•31m ago•1 comments

Opt-In Event Phases for Reliably Fast DOM Operations

https://github.com/MicrosoftEdge/MSEdgeExplainers/blob/main/EventPhases/explainer.md
1•robin_reala•33m ago•0 comments

A decade of database innovation: The Amazon Aurora story

https://www.amazon.science/blog/a-decade-of-database-innovation-the-amazon-aurora-story
1•mariuz•37m ago•0 comments

Americans' junk-filled garages are hurting EV adoption, study says

https://arstechnica.com/cars/2025/08/junk-filled-garages-hurt-ev-sales-as-people-dont-have-room-for-chargers/
3•embit•40m ago•0 comments

I built an open-source extension to manage canned responses for customer support

https://github.com/tabelf/speech-repo
1•1437173153•47m ago•1 comments

Mirage 2 – Generative World Engine

https://twitter.com/DynamicsLab_AI/status/1958592749378445319
1•bbzjk7•47m ago•0 comments

Psychologist for Founders

https://www.marcosander.com
1•MarcoSander•51m ago•0 comments

The Baby Paradox in Haskell

https://blog.jle.im/entry/the-baby-paradox-in-haskell.html
1•Bogdanp•51m ago•0 comments

Is AI Zover?

https://www.ft.com/content/2469cd71-9455-4a24-a7b0-a16d5f9586e1
1•KnuthIsGod•52m ago•0 comments

The Pick OS

https://www.youtube.com/watch?v=pWZBQMRmW7k
1•firefoxd•56m ago•0 comments

Nano Banana

https://nanobanana.ai/
2•iamflimflam1•1h ago•0 comments

California Resident Tests Positive for Plague. What to Know About the Disease

https://time.com/7311474/california-resident-tests-positive-for-plague/
4•mdp2021•1h ago•1 comments

I Built Free QR Code Validator – Validate Before Scan

https://v2.qranalytica.com/tool/qr-validator
2•nookeshkarri7•1h ago•0 comments

Not-So-Great Firewall: China Blocks the Web for 74 Min

https://securityboulevard.com/2025/08/great-firewall-china-web-74-minutes-richixbw
1•austinallegro•1h ago•0 comments

Why 'Silksong' Took Seven Years to Make

https://www.bloomberg.com/news/newsletters/2025-08-21/why-silksong-team-cherry-s-sequel-to-hollow-knight-took-so-long-to-make
1•microsoftedging•1h ago•1 comments

LLMs Won't Replace Programming Languages

https://mech-lang.org/post/2025-01-09-programming-chatgpt/
3•anewhnaccount2•1h ago•0 comments

It's Not Wrong that " ".length == 7

https://hsivonen.fi/string-length/
11•program•1h ago•9 comments

AI Cannibalism Can Be Good

https://gwern.net/blog/2025/ai-cannibalism
6•tie-in•1h ago•0 comments