Elemental Cognition Sets New Standard for Generative AI, Achieves 100% Accuracy in Complex Reasoning Scenarios

From the Inventor of IBM's Watson, Elemental Cognition's Neuro-Symbolic AI Platform Outperforms LLM Capabilities for Complex Problem Solving, Makes Generative AI Accurate and Transparent

NEW YORK, NY / ACCESSWIRE / February 20, 2024 / Elemental Cognition (EC), an advanced artificial intelligence company, has released the findings from their landmark report, "AI Beyond LLMs: Benchmarking the Complex Reasoning Performance of GPT-4 and EC AI." The findings show a dramatic leap in AI capabilities, demonstrating that EC's neuro-symbolic AI platform significantly outperforms using Large Language Models (LLMs) alone to solve complex reasoning scenarios.

The report emphasizes the crucial gap in the current AI landscape, where LLMs excel in natural language fluency, but fall short in solving complex, high-stakes business problems that require accurate, transparent, and reliable reasoning. EC's neuro-symbolic AI platform bridges this gap by integrating LLMs with formal mathematical algorithms, delivering 100% accuracy in complex reasoning tasks.

Dr. David Ferrucci, founder of EC and inventor of IBM's Watson, commented on the findings: "This research not only marks a significant milestone in AI but also sends a clear message to the industry. We are transitioning from an era where AI's potential is predominantly shown through language fluency, to a new era where AI can, and must, solve complex, real-world problems with accuracy and transparency. Elemental Cognition's success in achieving 100% accuracy in complex reasoning scenarios is a testament to our commitment to advancing AI toward this goal."

Key Findings:

• EC maintained 100% accuracy rate across all tested complex reasoning scenarios. Meanwhile, the accuracy of one of today's most well-known LLMs dropped from 32% to as low as 12% as the complexity of tasks increased.
• EC consistently identified and corrected errors in proposed plans with 100% accuracy, as opposed to 40% accuracy in error detection and only 18% in plan repair of the well-known LLM.

EC is proud to announce the publication of a comprehensive whitepaper on Arxiv accompanying the report. This whitepaper offers an in-depth look into the experimental design, evaluation criteria, and results, providing thorough insight into the methodologies.

For more information and to read the report in full, visit https://ec.ai/performance-benchmarks/

About Elemental Cognition

Elemental Cognition (EC) is the next wave of enterprise AI that delivers the reliability, accuracy, and transparency businesses need. EC combines the fluency of large language models with the rigor of formal reasoning to create a neuro-symbolic AI architecture that can solve complex problems when businesses can't afford to be wrong.

EC was founded in 2015 by Dr. David Ferrucci, the renowned AI researcher and inventor of IBM's breakthrough Watson technology, to build more holistic AI that thinks before it talks.

Contact Information

Holly Hitchcock
holly@gofrontlines.com
702.758.4079

SOURCE: Elemental Cognition

View the original press release on newswire.com.