Tag
StreetComplete is a mobile app that gamifies contributing to OpenStreetMap by presenting users with simple on-site quests to fill in missing map data, directly updating the map in the user's name.
Someone connected Blender to Fable 5, and in just 20 minutes, it reconstructed the entire New York City building complex based on public building data, with the model to scale. The poster believes this 'check data before acting' approach is smarter than Opus 4.8, reflecting that AI is beginning to have the ability to 'do its homework'.
A collection of cybersecurity datasets for machine learning and model training, covering network traffic, malware, web attacks, phishing, and more, including notable public research datasets like LANL, CTU-13, and UNSW-NB15.
A new tool provides access to 80TB+ of data from 30 astronomical surveys, enabling crossmatching on a laptop with minimal RAM.
Simon Willison created a tool to convert Mozilla's browser compatibility data into a downloadable SQLite database, hosted with open CORS headers for easy exploration via Datasette Lite.
StartupWiki is a free, open startup database for discovering and researching companies without accounts or subscriptions.
A new paper builds and releases a corpus of 2.2 million city and county laws from across the United States, making publicly accessible legal texts that were technically public but hard to find.
Ukraine launches TrophyLab, a secure platform providing global partners with deep technical data, reports, and physical samples from captured Russian weapon technologies to accelerate countermeasure development.
Bellwethr is developing an open methodology for tracking the real USD cost of a single inference token from capable models, with a draft benchmark suite and community contributions underway.
This website provides a free interactive global power grid map, containing data on 120,000+ power plants, 2.7 million transmission lines, etc., revealing information about energy infrastructure.
The Open Repair Data Standard (ORDS) defines a shared approach for collecting and sharing repair data about small electrical and electronics, enabling aggregation and analysis of repair trends across different community repair groups.
A user tested five AI models summarizing immigration news articles and found that all models inherited the framing of the source text, sounding neutral but shaping reader understanding through emphasis and omission. The study is small and exploratory, with open data available.
Learn how to set up and use Common Crawl data locally for web data processing tasks.
Paperclip adds over 220K FDA regulatory documents and 1M clinical trials from multiple registries, enabling AI agents to search and reason over clinical and regulatory data without web search. This update allows users to query FDA documents, ClinicalTrials.gov, and international registries via a unified filesystem interface.
Released a free 9.8 million document multilingual Indic corpus (11 languages, CC0 license) on HuggingFace, containing approximately 8.4 billion tokens, built for multilingual research.
Protovoters is an open-source tool that lets users build local voter files from public data and use them with standard geospatial software, aiming to replace expensive proprietary platforms like VAN or NationBuilder.
Version 2 of the GM-SEUS open dataset now maps 3.4 million U.S. solar panels plus new rooftop arrays, up from 2.9 million in v1.
DR-Venus-4B is a 4B-parameter deep-research agent trained on only 10K open samples via agentic SFT+RL with turn-level rewards, outrunning prior sub-9B agents and rivaling 30B models on research benchmarks while staying deployable on edge devices.
Katzilla is a product designed to provide easy government data access for citizens, optimized for AI consumption.
Bert Hubert beschrijft hoe hij hyperlinks heeft toegevoegd aan Tweede Kamerstukken en motienummers in debatten heeft gekoppeld aan documenten, en reflecteert op de traagheid van innovatie.