Search
Close this search box.

Cognizant Leads the Charge in Generative AI Evolution, Advocating for Neuroscience-inspired Techniques & Metacognition

Screenshot (1410)

Cognizant Leads the Charge in Generative AI Evolution, Advocating for Neuroscience-inspired Techniques & Metacognition

In an interesting discussion with Babak Hodjat, Cognizant’s Chief Technical Officer of AI, Risto Miikkulainen, Vice President of AI Research at Cognizant and Professor of Computer Science at the University of Texas at Austin, talked about the evolving landscape of generative AI. Cognizant, a global powerhouse in IT services, is steering the conversation towards innovative AI models that transcend traditional computational boundaries, embodying a new wave of technological evolution.

Miikkulainen articulated the shift from static AI evaluation methods, like the Turing Test, to dynamic assessments tailored to generative AI’s nuanced capabilities.

“It is a black box and we are interacting with it and try to evaluate how well it’s doing,” said Miikkulainen, which highlights the transition from AI that mimics human behavior to AI that performs tangible tasks. This pivot necessitates novel evaluation strategies that accommodate the “black box” nature of these advanced systems.

Addressing the challenges posed by generative AI’s complexity, Miikkulainen pointed out the inadequacy of traditional statistical testing due to the intensive nature of training these models.

“We cannot really do this kind of statistical testing,” he said, advocating for a more qualitative approach reminiscent of social sciences to comprehend and harness these technologies.

Miikkulainen further discussed the potential of neuroscience-inspired techniques and metacognition in enhancing AI’s self-awareness and problem-solving prowess.

“Metacognition is something we can already build,” said Miikkulainen. This statement envisions a future where AI can self-evaluate and adapt, drawing closer to the cognitive depth of the human brain.

For businesses, this paradigm shift signifies a reimagining of how generative AI integrates into workflows. Miikkulainen suggested treating AI as a human expert, requiring guidance and integration rather than mere calculation.

“Generative AI is no longer really a calculator that you can completely trust,” he said, advising enterprises to adapt their strategies to leverage AI’s evolving expertise effectively.

This forward-thinking approach underscores Cognizant’s commitment to pioneering AI solutions that resonate with the intricate demands of the modern digital landscape.

Featured image: Credit: Cognizant