What's Happening?
Probably, a startup focused on improving the accuracy of AI systems, has raised $9 million in funding from Andreessen Horowitz. The company aims to address the issue of AI hallucinations, where models like ChatGPT provide incorrect information. Probably has developed
a 'data science mech suit' system that uses a deterministic validator to ensure AI responses match a verified database. This system is designed to enhance the accuracy of AI outputs to 99.99%, similar to traditional software. The startup's first product is an analytical tool that provides quick answers from complex datasets, complete with source links and an audit trail. This tool operates efficiently on smaller neural networks, reducing costs for companies.
Why It's Important?
The development of Probably's technology could significantly impact industries reliant on AI for data analysis and decision-making. By reducing errors in AI outputs, businesses can integrate AI into critical processes with greater confidence. This advancement is particularly relevant for sectors like accounting, medical diagnostics, legal analysis, and engineering, where accuracy is paramount. The startup's approach also challenges the current business models of major tech companies, which may benefit from repeated AI queries. If successful, Probably's technology could lead to more efficient AI usage, saving time and resources for businesses.
What's Next?
Probably plans to expand its technology into various sectors where precision is crucial. The startup's focus on smaller, efficient models could lead to broader adoption across industries, as companies seek to reduce AI-related costs. The success of this technology may prompt larger tech companies to reconsider their AI strategies, potentially leading to more competition in the market for accurate AI solutions. As Probably continues to develop its products, the company may attract further investment and partnerships, accelerating its growth and influence in the AI industry.
