Archive — US — techcrunch
Superblocks CEO: How to find a unicorn idea by studying AI system prompts
Brad Menezes, CEO of Superblocks, posits that billion-dollar startup ideas can be discovered by analyzing system prompts from AI startups. His company revealed 19 influential prompts, highlighting the importance of prompt design for AI models and the distinction between core prompts and enriching frameworks. Superblocks has attracted significant attention and funding, aiming to empower non-developers with enterprise coding tools.
Key points
- Brad Menezes believes innovative startup ideas are often embedded in AI system prompts used by successful companies.
- System prompts, which can exceed 5,000 words, direct AI models to perform specific tasks in tailored ways.
- Superblocks recently unveiled 19 system prompts from popular coding tools, contributing to the discourse on prompt engineering.
- Menezes emphasizes that 'prompt enrichment'—the infrastructure surrounding prompts—is critical for successful AI applications.
- The study of system prompts revealed three focal areas: role prompting (setting the persona of the AI), contextual prompting (providing task-specific context), and tool use (executing tasks beyond text generation).
- Superblocks has raised $23 million in Series A funding and has secured clients like Instacart, suggesting strong market interest.
Why it matters
- The rise of AI and automation tools is a prominent trend in 2025, with many companies focusing on enhancing productivity through intelligent systems.
- Understanding AI prompt engineering can provide insights into the evolving landscape of software development and enterprise solutions.
- Superblocks' approach aligns with the growing demand for no-code or low-code platforms aimed at non-technical users—an increasing trend among startups and enterprises.
Menezes' insights into AI system prompts and their role in unlocking startup potential underscore an emerging trend where understanding AI engineering can propel business innovation and growth.