“For the majority of businesses, focus on building applications using agentic workflows rather than solely scaling traditional AI. That’s where the greatest opportunity lies.” – Andrew Ng – AI Guru
Andrew Ng is widely recognized as a pioneering figure in artificial intelligence, renowned for his roles as co-founder of Google Brain, former chief scientist at Baidu, and founder of DeepLearning.AI and Landing AI. His work has shaped the trajectory of modern AI, influencing its academic, industrial, and entrepreneurial development on a global scale.
The quote “For the majority of businesses, focus on building applications using agentic workflows rather than solely scaling traditional AI. That’s where the greatest opportunity lies.” captures a key transformation underway in how organizations approach AI adoption. Ng delivered this insight during a Luminary Talk at the Snowflake Summit in June 2024, in a discussion centered on the rise of agentic workflows within AI applications.
Historically, businesses have harnessed AI by leveraging static, rule-based automation or applying large language models to single-step tasks—prompting a system to generate a document or answer a question in one go. Ng argues this paradigm is now giving way to a new era driven by AI agents capable of multi-step reasoning, planning, tool use, and collaboration—what he terms “agentic workflows”.
Agentic workflows differ from traditional approaches by allowing autonomous AI agents to adapt, break down complex projects, and iterate in real time, much as a human team might tackle a multifaceted problem. For example, instead of a single prompt generating a sales report, an AI agent in an agentic workflow could gather the relevant data, perform analysis, adjust its approach based on interim findings, and refine the output after successive rounds of review and self-critique. Ng has highlighted design patterns such as reflection, planning, multi-agent collaboration, and dynamic tool use as central to these workflows.
Ng’s perspective is that businesses stand to gain the most not merely from increasing the size or data intake of AI models, but from designing systems where AI agents can independently coordinate and accomplish sophisticated goals. He likens this shift to the leap from single-threaded to multi-threaded computing, opening up exponential gains in capability and value creation.
For business leaders, Andrew Ng’s vision offers a roadmap: the frontier of competitive advantage lies in reimagining how AI-powered agents are integrated into business processes, unlocking new possibilities for efficiency, innovation, and scalability that go beyond what traditional, “one-shot” AI can deliver.
Ng continues to lead at the intersection of AI innovation and practical business strategy, championing agentic AI as the next great leap for organizations seeking to realize the full promise of artificial intelligence.
