OpenAI has developed a new system specifically trained to solve grade school math word problems, achieving nearly twice the accuracy of a fine-tuned GPT-3 model. In a direct comparison, this specialized system demonstrates significant advancements in AI's ability to parse and solve complex text-based reasoning tasks. The performance is benchmarked against human capabilities, showing the AI solves about 90% as many problems as real children.

In a test using a dataset of problems, a small sample of 9-12 year olds scored 60%, while the new OpenAI system scored 55% on the same challenges. This narrows the gap between AI and human-like reasoning in structured problem-solving, highlighting a major step forward in machine learning models designed for specific, logic-driven applications beyond simple text generation. This innovation is crucial for developing more reliable and accurate AI tools for education and data analysis.