AI will never be equal to humans: Yann LeCun

There is no idea about the physical world of the artificial intelligence-based tools that are so much feared and discussed around the world. At the same time, they cannot provide a rational definition of many things. So, AI will never be able to match humans in reasoning or planning.

In an interview given to the British media Financial Times, Yann LeCun, chief AI scientist at Meta, the parent company of Facebook, known as the 'godfather' of AI, opined so.

According to Yann LeCun, the Large Language Model (LLM) for artificial intelligence powers AI chatbots like ChatGPT and Gemini. These models will never be able to match humans in reasoning or planning. Because these models do not have permanent memory. The model is sincere but insecure. These models can only answer prompts (questions or instructions) correctly. These models can give correct answers if the training or data is correct. The evolution or development of different models is actually quite limited. These models learn only when people give them information. They mainly gather knowledge from data collected through many training sessions.

According to Yann LeCun, the development of Artificial General Intelligence (AGI) is not a technological issue but a scientific problem.

Meanwhile, this week Yann LeCun wrote in X (former Twitter), Musk wants to pursue the truth but raises various conspiracy theories on his social media. In response to the post, Musk wanted to know what kind of science Yann LeCun works with. Yann LeCun, Meta's chief AI scientist, cites more than 80 technical studies published on his Google Scholar page. However, Elon showed little interest in LeCun's paper.

In 2022, Yann LeCun published the first paper on World Modeling Vision. Meta has since published two research models based on this approach.