Introduction : Davos 2024

Key Takeaways 

• AI experts at Davos 2024 discussed the challenges and opportunities of generative models, which can create realistic text, images, and other data from scratch. 

• The panellists agreed that the world is running out of public data on the internet to feed the generative models, but there is still a lot of untapped data in the real world, such as visual and sensory information. 

• The panellists also shared their views on the commercial value, ethical implications, and future directions of generative AI, which is expected to revolutionize various domains and industries. 

The World Economic Forum at Davos has long been recognised as a platform that brings together influential leaders from various industries, sparking insightful discussions. This year, AI experts have brought attention to a surprising problem: the internet is nearing its data capacity. This finding emphasises the rapid pace of technological advancements and the challenges that lie ahead in the realm of data accessibility. The experts’ insights have sparked a global conversation about the future of data and its pivotal role in the advancement of AI. 

In the current digital age, data is the cornerstone of innovation, especially in fields like AI. However, AI experts at Davos 2024 have flagged a concerning trend – the internet is nearing a saturation point for text-based data. This implies that the volume of data available for training and testing AI models is dwindling, posing a significant challenge for the future of generative AI. 

Visual and Sensory Data For AI 

Despite the saturation of text-based data appearing as a hurdle, AI experts believe it’s merely a sign of evolution. The future of Generative models lies in visual and sensory data, heralding new opportunities in AI development. As Yann LeCun, one of the pioneers of AI, stated, “There is data hidden in silos, we have to make it more interactive.” He also highlighted that a 4-year-old child has seen 50 times more information than the largest LLMs that we have today, indicating the vast potential of visual data. Daphne Koller, founder and CEO of Insitro Inc and a co-founder of Coursera, concurred, stating “We have just scratched the surface of the data that are going to become available. We have agents in place with augmented reality.” 

The discussions at Davos 2024 serve as a reminder of the ever-evolving nature of data and AI. As we approach the saturation point of text-based data, the focus is shifting towards visual and sensory data. This transition promises a future where AI can interact with the world in more nuanced ways. However, this also necessitates new scientific and technological breakthroughs, as well as ethical and social considerations. As Andrew NG, founder of DeepLearning.AI, co-founder of Google Brain and Coursera, stated, “We are used to running LLMs in the Cloud but now with more open source, we will be running a lot of LLMs on our own devices. We have all of these vectors of innovation and I am optimistic about the opportunity.” 

The insights from Davos 2024 are a wake-up call to the world. As we stand on the brink of a data drought, it’s crucial to explore new frontiers of data collection and processing. The future of AI depends on our ability to adapt and evolve, much like the AI models themselves.