Meta guru says ChatGPT-style AI is out-of-date | The Business ...
On Tuesday, June 11, 2024, the chief scientist for Facebook owner Meta expressed his views on generative AI, the technology underpinning ChatGPT. He mentioned that this technology had reached a dead end and emphasized the need for new artificial intelligence that mirrors human rationality.
"Today, AI and machine learning really sucks. Humans have common sense, machines don't," stated Yann LeCun during a press briefing at a Meta launch event in Paris. LeCun's statements coincided with Meta's announcement of its latest AI endeavor, the image-based Joint Embedding Predictive Architecture (JEPA).
Advancing Beyond Generative AI
The JEPA project aims to move beyond the limitations of ChatGPT-like generative AI by enabling machines to grasp abstract concepts rather than merely replicate existing online content. LeCun confidently asserted, "Generative models are the past, we will abandon them in favor of joint embedding predictive architecture."
LeCun, a prominent figure in the field of AI, has been critical of the hype surrounding generative AI models like ChatGPT and Dall-E, which emerged last year. He believes that the perceived capabilities of generative AI are exaggerated compared to reality.
.jpg)
Meta's Strategic Approach
Meta CEO Mark Zuckerberg highlighted the open-source nature of the JEPA tool, making it accessible for researchers to experiment with. The ultimate goal is to develop AI systems that better reflect human understanding of the world. Zuckerberg stated, "We need models that perceive the world and make predictions. This research is another step in that direction."
Unlike some of its competitors, Meta has adopted a more understated strategy concerning ChatGPT-style AI in its platforms like Facebook and Instagram. While integrating generative AI into its products, Meta has not showcased it with the same fanfare as companies like Microsoft or Google.
Additionally, Meta has introduced open-source AI models that are less resource-intensive than the technology behind ChatGPT.
Conclusion
In conclusion, the landscape of AI is evolving rapidly, with Meta taking a fresh approach towards developing AI systems that align more closely with human cognitive processes. By moving away from generative AI models and embracing innovative architectures like JEPA, Meta is at the forefront of shaping the future of artificial intelligence.
