• AI Spotlight
  • Posts
  • This AI Startup Reached $200 Million in Revenue This March—Without Spending a Dime on Marketing

This AI Startup Reached $200 Million in Revenue This March—Without Spending a Dime on Marketing

Greetings AI enthusiasts. Anysphere, an AI startup known for its code-writing tool Cursor, is shaking up the tech scene with remarkable momentum. In just two months, the company doubled its revenue, making Cursor one of the fastest-growing AI products ever, having largely achieved this without the help of enterprise clients or traditional marketing.

In today’s email:

  • This AI Startup Reached $200 Million in Revenue This March Without Spending a Dime on Marketing

  • What People Are Actually Doing with Gen AI in 2025

  • It’s the first U.S. nuclear plant to use AI. Where does Diablo Canyon go from here?

LATEST NEWS

ANYSPHERE

image credit: Anysphere

AI Spotlight: Anysphere, an AI startup known for its code-writing tool Cursor, is shaking up the tech scene with remarkable momentum. In just two months, the company doubled its revenue, making Cursor one of the fastest-growing AI products ever, having largely achieved this without the help of enterprise clients or traditional marketing.

Key details:

  • Record-breaking growth: Cursor became the fastest-growing software product to hit $100 million in annual recurring revenue and reached $200 million by March.

  • Consumer-driven success: Nearly all revenue comes from 360,000 individual subscribers paying $20 to $40 monthly.

  • Developer-friendly design: Cursor’s interface mirrors Visual Studio Code, attracting developers through familiarity.

  • Power and accessibility: It uses multiple large language models and offers a free tier for up to 2,000 code completions monthly.

Anysphere’s explosive growth has set it apart from other AI startups, many of which are struggling to balance high costs with revenue. Cursor’s individual-focused model, developer appeal, and performance-driven growth make it a standout in the AI landscape.

GEN AI

image credit: Andrea Ucini

AI Spotlight: Generative AI has evolved significantly over the past year, prompting a fresh look at how people are truly using it. An updated report, based on extensive data from Reddit, Quora, and other forums, explores the top 100 use cases of 2025, revealing surprising shifts from technical tools to emotional support systems. From therapy to trip planning, gen AI is now embedded in both our personal and professional lives.

Key details:

  • Emotional and personal applications dominate: People increasingly turn to AI for therapy, life organization, and finding meaning.

  • Wider themes emerge across use cases: Six key categories: personal/professional support, content creation, education, technical help, creativity, and research/decision-making. These reflect AI’s expanding role in everyday support and decision-making.

  • Real-world impact is visible and practical: Users rely on AI for tangible help like meal planning, contesting traffic tickets, and learning skills.

  • Users are savvier—and more skeptical: Prompting skills have improved, but concerns about AI’s limits and ethics are rising.

AI isn’t just helping us think better. It’s helping us feel better. As users grow more sophisticated and emotionally invested, gen AI’s role is shifting from tool to teammate.

NUCLEAR AI

AI Spotlight: The Diablo Canyon nuclear power plant in California is making headlines with its cautious yet innovative use of artificial intelligence. Partnering with local startup Atomic Canyon, PG&E is deploying an AI system called Neutron Enterprise, not to run the plant, but to help workers navigate vast stores of complex regulatory documents. While the AI rollout is being embraced for its efficiency, it has also sparked questions about oversight and the future role of AI in sensitive infrastructure.

Key details:

  • Current use is document-focused, not operational: Neutron Enterprise helps staff search and summarize millions of technical and regulatory documents.

  • Regulatory oversight and concerns persist: There is currently no formal framework governing AI use in nuclear settings.

  • AI tool was custom-built and trained for nuclear jargon: The system was trained on 50 million NRC documents to ensure accurate interpretation.

  • Mixed reactions from public officials: Some leaders support its use, while others express concern about job displacement and safety oversight.

Diablo Canyon’s AI experiment marks a significant but measured step in merging nuclear energy with modern tech. For now, it’s a high-powered search engine, but its future could shape how AI is used in critical infrastructure nationwide.

MORE STORIES IN AI

Isomorphic Labs, the secretive AI life sciences startup spun off from Google DeepMind
Bhagavad Gita: What it can teach us about AI
Microsoft says it’s ‘slowing or pausing’ some AI data center projects
Where AI Is Now: Smaller, Better, Cheaper Models
Amazon CEO Andy Jassy says AI costs will come down over time