Traditional methods of assessing innovation often overlook the intricate factors influencing an invention’s success and value. ‘Predictive Patentomics: Forecasting Innovation Success and Valuation with ChatGPT’ (Read Here) harnesses OpenAI’s ChatGPT technology for an improved LLM-based patent analysis. It showcases a considerable enhancement in predicting patent value and differentiation of patent applications.

Highlights of the study:

• A notable 24% increase in R-squared scores when predicting patent worth
• Precise segregation of the lowest and highest potential patents
• Revision in valuations compared to traditional methods
• A demonstrated portfolio strategy exhibiting significant annual returns

Highlighting the use of AI in revamping the evaluation of patents and innovation, this study suggests impactful changes for corporate strategies, particularly for startups and small-scale enterprises. The incorporation of textual embedding and deep learning models showcases the potential of AI in transforming intellectual property assessment and informing investment decisions.