The future of AI: Beyond language models to becoming real world decision makers

Metaverse

The future of AI: Beyond language models to becoming real world decision makers – Crypto News

Published

9 months ago

May 24, 2024

Dripp

Imagine a world where artificial intelligence (AI) doesn’t just process language but truly understands and navigates the complexities of the real world.

Gagandeep Reehal, co-founder and CEO of Minus Zero, articulated this transformative idea as he addressed the Mint Digital Innovation Summit 2024. Reehal emphasized the need to shift from traditional large language models (LLMs) to what he called “real world models,” positioning this pivot as a unique opportunity for India to lead the AI revolution.

Reehal highlighted the current limitations of LLMs, which, despite their advanced capabilities in pattern recognition and language processing, fell short in real-world cognition and reasoning.

He illustrated this with several striking examples, including the notorious issue of AI-generated hallucinations. Unlike humans, who naturally understand and navigate their environment, LLMs often falter, producing erroneous outputs such as confusing handshakes or misinterpreting physical interactions. This lack of intrinsic understanding becomes particularly problematic in high-stakes applications like autonomous driving, where AI’s mis-judgments could lead to catastrophic outcomes.

The crux of the problem lies in AI’s inability to reason—a fundamental human trait that allows for nuanced decision-making and adaptability in complex scenarios.

“Basic organisms navigate their environments better than the best AI because they understand and respond to the world in ways AI currently cannot,” Reehal explained. This gap underscores the urgent need for a new generation of AI models that move beyond mere imitation of patterns.

This new frontier in AI research was being actively explored by tech giants such as Meta, Google, and other leading organizations. Their focus is on developing foundational models that could understand the physics and rules of the real world, enabling them to make reliable, real-world decisions.

These models aim to address current AI’s shortcomings by incorporating a deeper understanding of the environment, which is essential for applications ranging from robotics and stock market investments to drone delivery and navigating challenging terrains like mining sites.

Reehal emphasized that this transition towards embodied AI represented a big opportunity for India. “With LLMs becoming increasingly commoditized, the real value lies in these foundational models, where there is still a significant gap to be filled,” he pointed out.

Milestone Alert!
Livemint tops charts as the fastest growing news website in the world 🌏 Click here to know more.

You are on Mint! India’s #1 news destination (Source: Press Gazette). To learn more about our business coverage and market insights Click Here!

Catch all the Business News, Market News, Breaking News Events and Latest News Updates on Live Mint.
Download The Mint News App to get Daily Market Updates.

More
Less

Published: 24 May 2024, 05:17 PM IST