frontpage.
newsnewestaskshowjobs

Open Source @Github

fp.

Open in hackernews

Show HN: Grand Treasure Island – GTA-style game on a real island, modeled by AI

https://grandtreasureisland.vercel.app/
1•phpp•5m ago
Hi HN! I used teams of Opus 5.5 agents to build a life-size 3D version of Treasure Island, in the middle of the San Francisco Bay. The game has 16 miles of road, almost 200 buildings, over 40 kinds of plants and over 400 drivable cars. It runs on Godot, with data from OSM, USGS lidar, and public and private photos.

I started this after moving to the island and falling in love with it. From stunning views to giant piles of dirt, pelicans to nuclear waste, this place has it all. You can explore the entire island on foot, take to the skies with a jetpack and “borrow” any car.

The native versions run best and look best, with Forward+ rendering and dynamic time of day. Web works too but the Godot WebGL 2 compatibility renderer is holding on for dear life.

I did not write the code or create the 3D models. I set up the agent teams, the working environment and the ground rules. Throughout the process I checked the output, play tested and supplied private photo data where public sources had gaps.

Trailer: https://www.youtube.com/watch?v=yW2SQmB-2i0 How it was built: https://grandtreasureisland.vercel.app/#how

Let me know what you think!

So what should an engineer do now?

https://twitter.com/a7ulr/status/2122854660675432584
1•atulfiled•23s ago•0 comments

Defcon 5's California Caravan Program/Zine

https://netninja.com/2026/06/08/defcon-5s-california-caravan-program-zine/
1•speckx•36s ago•0 comments

Using a "chaos loop" to harden the Claude Docs launch

https://claude.dev/blog/chaos-loop-claude-docs-launch/
1•s3graham•46s ago•0 comments

Reliability and Fault Tolerance

https://medium.com/@oakley349/reliability-and-fault-tolerance-6861884a3433
1•oakley349•55s ago•0 comments

Show HN: Cmdbook – a home for the scripts you (or your AI) keep rewriting

https://printf.app/cmdbook/
1•kristinzk•1m ago•0 comments

SWE-CoWork: Benchmarking Coding Agents as Digital Employees in Organizations

https://arxiv.org/abs/2611.20218
1•Hasitron•1m ago•0 comments

I'm a human doctor with human intelligence. But I sure do like my AI notetaker

https://www.npr.org/2026/11/17/nx-s1-6037556/medicine-ai-doctor-notetaker-scribe
1•drodio•2m ago•0 comments

California sues OpenAI over AI agents escaping sandboxes, hacking firms

https://www.bleepingcomputer.com/news/security/california-sues-openai-over-ai-agents-escaping-san...
1•drodio•2m ago•0 comments

The Citizen Lab says that Canada should review acquisition by Francisco Partners

https://www.theglobeandmail.com/politics/article-ottawa-should-conduct-national-security-review-o...
1•bparsons•3m ago•1 comments

Brave Search now uses Redirect URLs in search results

https://brave.com/blog/google-goto-anti-scraping/
1•complex_fir_rea•3m ago•0 comments

Is the SaaS business model dead or there is still oportunities in this age of AI

1•jorgeacf•3m ago•0 comments

Crook uses LearnWorlds trials for phishing, sends 13,000 fake YouTube notices

https://www.theregister.com/security/2026/11/16/crook-turns-learnworlds-free-trials-into-phishing...
1•noorsimar•4m ago•0 comments

Show HN: Gridline – Fine-tune open models and run them in batch, pay per token

https://gridlinegpu.com
1•akshitp•5m ago•0 comments

Show HN: Grand Treasure Island – GTA-style game on a real island, modeled by AI

https://grandtreasureisland.vercel.app/
1•phpp•5m ago•0 comments

Synthesizing Rollback Systems for Safer Agents

https://www.basis.ai/blog/undo-for-agents/
1•nikhilpim•5m ago•0 comments

Hugging Face,chain: forensic brief on the'26 autonomous-agent production breach

https://claude.ai/artifact/NU2YVW2a7VkysLqjCm5evD
1•AgileMesh•7m ago•0 comments

2026 Lyttle Lytton Contest

https://adamcadre.ac/lyttle/2026.html
1•tibbar•8m ago•0 comments

Every pixel is a measurement: Lossless video compression in GStreamer

https://www.collabora.com/news-and-blog/blog/2026/11/18/every-pixel-is-a-measurement-lossless-vid...
1•losgehts•8m ago•0 comments

Anthropic IPO: A Brother-Sister Duo Behind a Trillion-Dollar AI Giant

https://www.forbes.com/sites/shaheenajanjuhajivrajeurope/2026/11/18/anthropic-ipo-the-brother-sis...
1•andsoitis•8m ago•0 comments

Show HN: Project-M – Local shared memory for AI coding agents

https://github.com/adityadebnath009/Project-M
1•adityadebnath•9m ago•0 comments

2026 – 1971 = Happy 55th Birthday Intel 4004

https://www.4004.com/
2•oldnetguy•10m ago•0 comments

App Store Awards 2026 Finalists

https://developer.apple.com/app-store/app-store-awards-2026/
1•soheilpro•13m ago•1 comments

Teaching an Autoregressive Agent to "Pace the Frontier"

https://github.com/ronfriedhaber/XRiskMan
1•ronfriedhaber•14m ago•0 comments

Show HN: Happo Animation Testing

https://happo.io/animations
1•lencioni•14m ago•0 comments

Show HN: ClearOffice Hub – Build Enterprise Apps with Natural Language

https://www.clearofficehub.com/
1•icon2026•14m ago•0 comments

Show HN: Specspec; specification as a spectrum from intentions to conclusions

https://github.com/jejay/specspec
1•jejay•15m ago•0 comments

Aspire 17 Is Released

https://devblogs.microsoft.com/aspire/whats-new-aspire-17/
1•bgrainger•15m ago•0 comments

US tech outages blamed on rush to fix bugs found by AI

https://www.ft.com/content/96c565f1-f229-40b6-91e8-7a849ec4b208
1•master_crab•15m ago•0 comments

LZ-RDO lands in ARM's astc-encoder

https://www.jonolick.com/home/lz-rdo-lands-in-arms-astc-encoder
1•dverlaeckt80•17m ago•0 comments

Show HN: x402-reconcile – check x402 payments against the chain

https://github.com/parthod0x/x402-reconcile
1•parthod0x•17m ago•0 comments