Keep pulling the thread on Dr. Fei-Fei Li.
The ImageNet dataset, spearheaded by Fei-Fei Li, was a key breakthrough that enabled the modern approach to building and scaling AI models.
World Labs has launched Marble, which is described as the world's first large-scale world model.
The ImageNet dataset was created by curating 15 million images from the internet, organized into a taxonomy of 22,000 object concepts based on the WordNet lexical database.
In the 2012 ImageNet challenge, a team of researchers from Toronto led by Professor Jeff Hinton used two NVIDIA gaming GPUs and the ImageNet dataset to create a neural network that achieved a major breakthrough in object recognition.
The combination of big data, neural network algorithms, and GPUs became the foundational recipe for modern AI and remains the core of models like ChatGPT.
WorldLabs has developed what it claims is the world's first generative AI model capable of producing fully 3D, navigable worlds from text or image prompts.
A virtual production company working with Sony reported that using WorldLabs' Marble platform reduced their production time by a factor of 40.
WorldLabs has demonstrated real-time video generation running on a single NVIDIA H100 GPU.
The trend of technology companies marketing themselves as 'AI companies' began around 2017.
Fei-Fei Li co-created Stanford's Human-Centered AI Institute (HAI).
Alex Wang, the founder of Scale AI, has stated that the ImageNet project was an inspiration for starting his company.
The term 'AGI' (Artificial General Intelligence) is considered more of a marketing term than a scientifically defined concept.