Industry executives and experts share their predictions for 2024. Read them in this 16th annual VMblog.com series exclusive.
Bringing purpose and personalization to real-time data in 2024
By Naren Narendren, Chief Scientist at Aerospike
2024 will bring focus to big ideas. 2023 thrust a renaissance of AI upon us, and companies scrambled to implement it at lightning speed. While AI streamlines tasks and optimizes processes, it has introduced many questions and controversies without clear, tested paths forward. The year ahead will represent a period of adjusting to the catalyst that is AI, bringing a deluge of data with it. But with the initial seismic wave rolling back into the ocean, organizations will be able to keep their heads above the water. Rather than simply trying to stay afloat, they’ll become more thoughtful and measured in their approaches to AI. In the next year, to both leverage and manage the vast amounts of data driving AI, we expect to see more cost-effective and accessible large language model (LLM) options, growing adoption of AI and ML in automation-heavy industries, and real-time data-driven hyper-personalized experiences for eCommerce and ad tech.
LLM options to diversify for accessibility
Though LLMs have some amazing capabilities, they’re cost-prohibitive for the overwhelming majority of organizations. Training them requires significant amounts of time and large amounts of compute infrastructure. They’re more like enormous language models – “large” doesn’t even begin to convey their size, working with tens or hundreds of billions of parameters. Only companies with extremely deep resources have the means to access them. Since there needs to be a path forward for making LLMs more economically viable, we should expect to see solutions that decentralize and democratize their use. We should anticipate more numerous and focused models, as well as smaller ones. These custom models for specific domains would have far fewer parameters than today’s LLMs and would consume lower compute power, making them more readily available to a wider range of users. We can also expect them to perform better since they’ll be trained in a more focused manner. This focus and compression leads to smaller server footprints, resulting in energy-efficient infrastructure. Such capabilities will enable organizations to derive value through cost-effectiveness by lowering an organization’s total cost of ownership (TCO).
Additionally, 2024 will bring further advancement in vectors for semantic search and other more “classical” aspects of AI and ML. While LLMs received significant attention in 2023, they cannot do the job alone. LLMs represent just one piece of the AI world. Other critical segments are also needed to complement the whole picture, so we can expect to see more progress there.
Industries reliant on real-time data to expand AI and ML use
Of the industries most likely to grow their adoption of AI and ML, automation-heavy industries with returns inversely proportional to response time will top the list. Though many companies in finance and fraud detection already blazed a trail for AI long before AI made its mark, we can expect to see another wave of companies in these real-time data-heavy industries adopting the technology to make necessary split-second, automated decisions. Both finance and fraud detection industries demand blink-of-an-eye decisions in critical moments and will find new ways to leverage AI as ML training increases over time.
Real-time data-based hyperpersonalization to win out over generalized predictions
Finally, real-time data will drive hyper-personalized user experiences in eCommerce and ad tech. Rather than platforms generating a response or serving content based on behavior from a user’s digital journey over the past six months, for example, they’ll react based on a search from three hours ago – or even a click from two minutes ago. As ML systems are fed more and more data, we’ll see the generalized statistical predictions of yesterday give way to hyper-personalized ones at the individual level for enhanced buyer experiences.
In 2024, organizations will hone their AI strategies, bringing new focus. As organizations adapt to purposefully using data in a personalized way, we anticipate an innovative year ahead.
##
ABOUT THE AUTHOR
Naren has spent three decades in a variety of activities in the science and technology space – fundamental research at Bell Labs, working at startups in the email and advertising space, and leading engineering teams and launching new products and infrastructure in the areas of Search, Advertising, Storage, Networking, Distributed Systems, and AI/ML at Google and Amazon. Throughout his career, Naren has had a particular interest in adapting novel technological and scientific concepts to pragmatic applications in scalable and performant hardware and software. Naren has a B.Tech in Computer Science from the Indian Institute of Technology, Madras, and an M.S. in Math and a Ph.D. in Computer Science from the University of Wisconsin-Madison.
