tech
Probably raises $9M to build a more reliable kind of AI
Probably wants to prevent hallucinations and factual errors from reaching users, and achieve accuracy on par with deterministic systems.
TL;DR
- Probably aims to prevent hallucinations and factual errors in LLMs.
- The company raised $9 million in seed funding from Andreessen Horowitz.
- Their approach uses a "harness system" and a deterministic validator to check LLM outputs.
- This method allows for the use of smaller AI models that can run on local hardware.
- The technology is expected to reduce token costs associated with AI use.
- The engine can be extended to other precision-sensitive use cases beyond data science.