In this engaging discussion, Bryan Harris, CTO of SAS, shares his expertise in AI and data science. He highlights five transformative trends for 2025, including the rise of synthetic data and the necessity of AI governance frameworks. The conversation delves into the challenges organizations face with regulation and energy costs while optimizing model training for sustainability. Bryan also explores the groundbreaking potential of quantum computing in AI, promising enhanced efficiency for industries like banking and healthcare.
The development of AI models as products is transforming technology by enabling organizations to deploy diverse, pre-trained models for practical applications and revenue growth.
The increasing reliance on synthetic data is revolutionizing industries like healthcare and finance, enhancing privacy, model accuracy, and cost-efficiency while addressing data sensitivity challenges.
Deep dives
Emerging AI Models as Products
The rise of AI models as standalone products is a key trend shaping the landscape of technology in 2025, particularly with an emphasis on predictive models beyond just generative AI. Today, the market is saturated with discussions about large language models (LLMs), but these constitute only a small fraction of the myriad modeling needs for practical applications. Organizations are increasingly looking to utilize deterministic models for various operational scenarios such as fraud detection and optimization, and SAS plans to offer approximately 50 diverse models that leverage their extensive industry experience. By focusing on creating pre-trained models that businesses can deploy via APIs, companies can drive both innovation and revenue growth, tapping into a multi-billion dollar opportunity.
The Importance of Synthetic Data
Synthetic data is becoming essential, especially for industries that are highly regulated, like healthcare and finance, where data sensitivity is paramount. The use of synthetic data enables organizations to lower acquisition costs, enhance privacy during training phases, and improve model accuracy, with notable improvements of 40 to 50% being reported by users. A strategic acquisition of Hazy software underscores SAS's commitment to integrating synthetic data generation into their offerings, facilitating better data handling and performance for clients. With synthetic data, businesses can mitigate challenges associated with rare events, allowing for more robust machine learning models that can effectively detect fraud and make better predictions.
AI Governance and Cloud Costs
As organizations adopt AI at scale, pressure mounts for effective AI governance to ensure transparency and fairness in decision-making processes. Leaders must understand the parameters of their AI deployments, including data lineage and model performance, as AI increasingly influences critical organizational decisions like lending and insurance rates. Additionally, the sustainability of cloud computing has become a pressing concern, as AI's growing popularity contributes significantly to global carbon emissions. Companies need to prioritize cloud efficiency not only to manage costs but also to align with ESG goals, with alternatives like SAS's optimally designed software showing a drastic reduction in training times and associated energy consumption.
Discover the game-changing AI and technology trends that will reshape business in 2025! đ
I'm joined by Bryan Harris, CTO of SAS, to explore five critical developments: models as products, synthetic data revolution, AI governance frameworks, cloud cost optimization, and the quantum computing breakthrough đ¤ đĄ đ ⥠đŽ