Our Methodology

How we analyze the AI industry and track progress toward AGI

Transparency is core to our mission. This page explains exactly how we collect data, score companies, detect trends, and analyze the AI industry. Our methodologies are designed to be reproducible and objective.

AGI Readiness Scoring

We evaluate each AI company across 7 dimensions that indicate their potential to contribute to or achieve artificial general intelligence. Each dimension is scored on a scale of 1-10, with weights applied based on historical correlation with AI breakthroughs.

1

Talent Density

Quality and depth of research team, including publications, citations, and notable researchers.

2

Compute Access

Infrastructure capabilities, GPU/TPU clusters, cloud partnerships, and training capacity.

3

Research Output

Published papers, open-source models, benchmark performance, and novel architectures.

4

Product Traction

Real-world deployment, user adoption, revenue, and market validation.

5

Funding & Runway

Financial resources, investor quality, and ability to sustain long-term research.

6

Data Access

Proprietary datasets, data partnerships, and training data advantages.

7

Strategic Position

Partnerships, platform control, regulatory relationships, and competitive moats.

Scores are updated quarterly or when significant company events occur (major funding, product launches, executive changes).

Narrative Detection

We use machine learning to automatically identify emerging trends and group related news articles into coherent narratives. This helps readers see the bigger picture beyond individual headlines.

How It Works

  1. Embedding Generation: Each news article is converted into a semantic vector using OpenAI's text-embedding-3-small model (1536 dimensions).
  2. Clustering: We apply DBSCAN (Density-Based Spatial Clustering) to group articles with similar embeddings. Articles within a cosine similarity threshold of 0.75 are considered related.
  3. Narrative Generation: For each cluster, GPT-4o-mini analyzes the articles to generate a narrative title, summary, and key themes that capture the overarching trend.
  4. Velocity Tracking: We track how fast narratives are growing (articles per day) and classify them as Emerging, Growing, Peaking, or Declining.

Narrative Lifecycle

Emerging

New pattern detected

Growing

Gaining momentum

Peaking

Maximum coverage

Declining

Fading attention

Deal Classification

We track AI industry deals and classify them by type to help investors and analysts understand market dynamics. Each deal is verified against primary sources when possible.

Deal Types

Investment: Equity funding rounds (Seed, Series A-F, growth rounds)
Acquisition: M&A activity including acqui-hires and asset purchases
Partnership: Strategic alliances, joint ventures, and collaboration agreements
Hardware: Compute infrastructure deals including GPU purchases and data center investments
Licensing: Model licensing, data licensing, and API agreements
Research: Academic partnerships, grants, and research collaborations

Company Valuations

Company valuations on Race to AGI come from multiple sources, prioritized by reliability:

  1. Public Market Cap: For publicly traded companies, we use real-time market capitalization.
  2. Latest Funding Round: Post-money valuation from the most recent funding announcement.
  3. Industry Databases: Data from Crunchbase, PitchBook, and CB Insights.
  4. Analyst Estimates: For companies without public valuations, we use analyst consensus when available.

Valuations are updated when new funding rounds are announced or quarterly for public companies. All valuations are displayed in USD.

News Curation

We aggregate AI news from verified sources and add contextual analysis. Here's our curation pipeline:

Collection (5x Daily)

News is collected from 15+ sources including company blogs, tech publications, and research repositories. We use RSS feeds and API integrations to ensure timely coverage.

Processing Pipeline

1
Collect
2
Dedupe
3
Summarize
4
Analyze
5
Publish

AI Assistance Disclosure

We use AI (GPT-4o-mini) to generate TL;DR summaries and "Race to AGI Analysis" sections. This is clearly labeled with an "AI-Summarized" badge on each article. The AI analysis explains why each story matters for AGI progress and who should care (investors, researchers, engineers).

AGI Impact Assessment

Each news article and narrative receives an AGI impact assessment indicating how it affects the timeline to artificial general intelligence:

Advances Timeline

Development that likely accelerates progress toward AGI (breakthrough research, major funding, compute scaling).

Delays Timeline

Development that may slow AGI progress (regulatory restrictions, talent disputes, safety concerns).

Neutral

Development with no direct impact on AGI timelines (market dynamics, business news).

Unclear

Impact is uncertain or debatable (early-stage research, speculative announcements).

About Race to AGIExplore the Platform