Last week, I attended a session that Yann LeCun was the guest speaker at a CollectiveI community event. The session was attended by leading experts in AI, ML, CEOs, journalists, and company representatives who simply wanted to hear what was on Yann’s mind. The session was jam packed with everyone engaged and cherishing each word.
Yann is is well known for his work on on convolutional neural networks (CNN), and is a founding father of convolutional nets. Yann received the 2018 Turing Award (often referred to as “Nobel Prize of Computing“), together with Yoshua Bengio and Geoffrey Hinton, for their work on deep learning. The three are sometimes referred to as the “Godfathers of AI” and “Godfathers of Deep Learning”.
I was privileged to be asked to ask him his views on AI Ethics in light of ChatGPT3’s announcement. His views were succinct and offered rich perspectives, often lacking in the AI industry, where the depth of thoughtful comparatives are often missing vs all the hoopla.
Yunn made it succinctly clear that it is only a matter of time before AI amplifies human abilities, but he also referenced with some sadness in his voice that he would unfortunately not be here to experience this reality.
He referenced that a cat is much smarter than AI – as a cat is all sensing. I found this comparison very intriguing as I am in the field of AI and dealing with AI every day, predicting futures with advanced AI SaaS toolkits, building complex AI models, or guiding companies on their ethical responsibilities to avoid data bias – really put in perspective how complex full sensory detection is vs analyzing historical data patterns to predict future outcomes, or even working with neural networks as he truly believes we have not invented the right technology methods to get closer to singularity.
What Yann did in just a few minutes was simplify the complex into a very memorable soundbites, helping us as AI leaders to put everything we do in perspective.
In other words, we are a very long way from achieving AI that is rich as a human mind, let alone a cat’s.
On the topic of ChatGPT3, he essentially said that ChatGPT is ‘not particularly innovative,’ and ‘nothing revolutionary’. Yes, it will provide information that over time as it will be incredibly accurate as it will be cleansed more, misinformation will be extracted, but it will never have any common sense in being able to look ahead and easily recognize multiple sensory patterns.
He almost said that ChatGPT3 was almost primitive vs what it could really be, so we are in the infancy stages of this innovative technology from OpenAI. He also stated that OpenAI is not particularly an advanced lab, compared to the other labs. He reinforced that Meta, Google and many start-ups are using technology just like ChatGPT3. He stated: “I don’t want to say it’s not rocket science, but it’s really shared, there’s no secret behind it, if you will.” “You have to realize, ChatGPT uses Transformer architectures that are pre-trained in this self-supervised manner” and that self-supervised-learning is something I’ve been advocating for a long time, even before OpenAI existed,” LeCun said.
He also did not believe that future approaches to achieving generative AI would be based on neural network approaches and that alternative new approaches are needed. So this must be in a follow-up interview with the Godfather of AI, Yann Lecun. I can hardly wait to speak to him, hope I can be fortunate to capture his attention.
Special recognition to the CollectiveI leadership team and founder, Heidi Messer for caring about our AI Ecosystem and bringing together so many thought leaders for all of us to learn and grow from.
To see more about ChatGPT3, see my prior article as it has over 55,000 views a reflection on the interest on this topic.
Yann LeCun’s @ylecun Deep Learning Course is now free & fully online at @NYUData
Videos, slides, notes, and notebooks!