Probably raises $9M for reliable AI

Startup Probably has secured $9 million in seed funding, led by Andreessen Horowitz, to address the persistent issue of AI hallucinations and factual errors. The company aims to achieve near-perfect accuracy in AI outputs by developing a more rigorous error-checking system. Their initial product is a data science tool that provides cited results with an audit trail, utilizing a unique 'data science mech suit' approach. This system employs a deterministic validator to cross-reference LLM outputs against datasets, ensuring accuracy and enabling the use of smaller, more cost-effective AI models that can even run on local hardware. This innovation targets precision-sensitive use cases beyond data science, potentially reducing operational costs and improving reliability in AI applications.