About gentic.news
gentic.news is an AI news intelligence platform. It automatically aggregates, analyzes, and surfaces the most important developments in artificial intelligence — from research breakthroughs to funding rounds, product launches, and policy changes.
How It Works
- Collection — RSS feeds and curated X/Twitter accounts are scanned every 6 hours for new AI-related content.
- Filtering — A three-stage pipeline (keyword match, AI relevance scoring, duplicate detection) selects the most significant stories.
- Analysis — Each article is scored for relevance, entities (companies, people, models, technologies) are extracted and linked into a knowledge graph, and sentiment is computed.
- Intelligence — The knowledge graph powers entity profiles, relationship maps, competitive analysis, and weekly intelligence reports.
- Predictions — Signal detection and graph patterns feed a prediction engine that generates verifiable forecasts about upcoming events.
Metrics Glossary
How significant an article is to the AI landscape. Factors include topic importance, source authority, recency, and whether it represents a breakthrough.
Average tone of coverage about an entity. Computed from article language: positive (+0.3 to +1.0), neutral (-0.3 to +0.3), negative (-1.0 to -0.3).
Week-over-week change in mention frequency for an entity. Positive means growing coverage; negative means declining interest.
How likely a prediction is to come true, based on evidence strength. Very Likely (85-100%), Likely (70-84%), Possible (50-69%), Speculative (<50%).
Total number of articles that reference a specific entity. Higher counts indicate broader coverage.
Number of article mentions in the last 7 or 30 days. Used to rank models and companies on the leaderboard by recent activity, not all-time totals.
Knowledge Graph
Every article is processed to extract entities — companies, people, AI models, technologies, and research topics. These entities are linked through relationships (developed, partnered, competes with, invested, etc.) and enriched with data from Wikipedia, Crunchbase, and arXiv.
The graph currently tracks 200+ entities with their relationships, timelines, and sentiment trends. It updates continuously as new articles are processed.
AI Model Leaderboard
The leaderboard ranks specific AI model versions (e.g., GPT-4o, Claude 3.5 Sonnet, Gemini 2.0 Flash) by media momentum — how much they are discussed in recent news — not by benchmark performance.
Rankings are based on mentions within the last 60 days, sorted by 7-day momentum first, then 30-day momentum. This ensures the leaderboard reflects the current conversation, not historical coverage.
Predictive Intelligence
Predictions are generated every 24 hours by analyzing knowledge graph patterns: entity surges, relationship clusters, sentiment shifts, and competitive dynamics. Each prediction includes:
- A clear, verifiable statement
- Confidence score with reasoning
- Target timeframe
- Verification criteria and supporting evidence
Predictions are experimental. They reflect patterns detected in news coverage and should not be treated as investment advice or certainties. Accuracy data is published as predictions are resolved.
Automation Disclosure
AI-generated content: Article summaries, entity extraction, sentiment analysis, predictions, and weekly intelligence reports are generated automatically using large language models.
Human oversight: Source selection, classification rules, and system configuration are maintained manually. The platform is designed to surface patterns, not to replace editorial judgment.
Source attribution: Every article links to its original source. gentic.news does not claim original reporting — it aggregates and analyzes published content.
Contact
Questions or feedback? Reach us at contact@gentic.news