Skip to content

Karpathy’s March of Nines: Why 90% AI Accuracy Fails

Sophie WeberSophie Weber
|
|4 Min Read
Karpathy’s March of Nines: Why 90% AI Accuracy Fails
Image: SwissFinanceAI / ai-tools

Andrej Karpathy explains why 90% AI reliability is dangerously low for production systems. His "March of Nines" framework shows the path from 90% to 99.999%.

ai-toolsnewstechnology

Karpathy’s March of Nines shows why 90% AI reliability isn’t even close to enough

Swiss finance institutions and fintech companies may find Andrej Karpathy's concept of the "March of Nines" particularly relevant, as it highlights the significant engineering effort required to achieve high reliability in AI-powered systems. Reaching 90% reliability, a common benchmark, is often just the starting point, with each additional 9% incrementing requiring substantial resources. This reality underscores the need for substantial investment in engineering and testing to ensure the dependability of AI-driven applications in critical areas such as risk management, compliance, and customer service. As AI adoption continues to grow in the Swiss finance sector, understanding the March of Nines can help companies set realistic expectations and allocate necessary resources.


Disclaimer: This article is for informational purposes only and does not constitute financial advice. Consult a licensed financial advisor before making investment decisions.

Source

Original Article: Karpathy’s March of Nines shows why 90% AI reliability isn’t even close to enough

Published: March 7, 2026


This article was automatically aggregated from VentureBeat AI for informational purposes. Summary written by AI.

Disclaimer

This article is for informational purposes only and does not constitute financial, legal, or tax advice. SwissFinanceAI is not a licensed financial services provider. Always consult a qualified professional before making financial decisions.

This content was created with AI assistance. All cited sources have been verified. We comply with EU AI Act (Article 50) disclosure requirements.

ShareLinkedInXWhatsApp
Sophie Weber
Sophie WeberAI Tools & Automation

AI Tools & Automation

Sophie Weber tests and evaluates AI tools for finance and accounting. She explains complex technologies clearly — from large language models to workflow automation — with direct relevance to Swiss SME daily operations.

AI editorial agent specialising in AI tools and automation for finance. Generated by the SwissFinanceAI editorial system.

Newsletter

Swiss AI & Finance — straight to your inbox

Weekly digest of the most important news for Swiss finance professionals. No spam.

By subscribing you agree to our Privacy Policy. Unsubscribe anytime.

References

  1. [1]NewsCredibility: 7/10
    VentureBeat AI. "Karpathy’s March of Nines shows why 90% AI reliability isn’t even close to enough." March 7, 2026.

Transparency Notice: This article may contain AI-assisted content. All citations link to verified sources. We comply with EU AI Act (Article 50) and FTC guidelines for transparent AI disclosure.

Original Source

blog.relatedArticles