The Future of AI in Education: 4 Key AI Trends from Andrew Ng ASU+GSV Summit Talk

Updated: May 03 2025 10:17

AI Summary: At the ASU+GSV Summit, AI pioneer Andrew Ng discussed the transformative impact of AI on education, emphasizing not only new teaching content but also revolutionary methodologies. 1 A key takeaway was his assertion that coding is becoming a crucial skill for nearly everyone, regardless of their profession, as AI coding assistants enhance productivity across various roles, a point illustrated by the story of a basketball coach who improved his coaching through coding. Ng highlighted four major AI trends shaping education.

In a recent presentation at the ASU+GSV Summit, AI pioneer Andrew Ng shared valuable insights on how artificial intelligence is revolutionizing education. His talk highlighted not just what we should be teaching in the AI era, but also how AI is transforming teaching methodologies. As someone deeply immersed in both AI development and education, Ng offered a unique perspective on the intersection of these fields and what it means for learners and educators alike.



Perhaps the most provocative assertion in Ng's talk was that learning to code is becoming essential for nearly everyone—not just those pursuing careers in software engineering. He boldly stated that advising people not to learn coding because "AI will automate it" might be "some of the worst career advice ever given."


As AI coding assistants make programming more accessible, Ng has observed a widening performance gap between team members who know how to code and those who don't—even in non-technical roles. At AI Fund, where Ng works, professionals across different departments have benefited tremendously from coding skills:

• Marketers who code outperform those who don't
• Recruiters with coding skills are more efficient
• Even the associate general counsel writes code to screen NDAs more effectively
• Finance team members leverage coding for better analysis

I think we've actually reached that tipping point where it makes sense to frankly get pretty much everyone to learn how to code

One of the most inspiring stories Ng shared was about Kyle Creazy, a basketball coach with a bachelor's degree in physical education. Despite never having written a line of Python code two years ago, Kyle now not only codes but teaches computer science.


When asked what excited him most about teaching CS, Kyle showed Ng a chart he had generated with code that analyzed his players' three-point shooting volumes versus percentages. This perfectly illustrates how coding skills can enhance performance in seemingly unrelated fields—Kyle became a better basketball coach by learning to code.

This transformation was made possible through Kira Learning, a company that AI Fund is involved with, which is working to address the critical shortage of computer science teachers.

Ng outlined four key AI technology trends that are shaping the future of education:



The most significant trend according to Ng is "agentic AI workflows." Unlike standard prompting where an AI tries to produce a complete response in one go (like "writing an essay without using backspace"), agentic workflows allow for iterative processes where AI can think, research, draft, revise, and refine—resulting in much higher quality outputs.

For education, this means AI can help create courses by assembling content from trusted sources, personalizing learning experiences, and handling complex workflows that previously required significant human effort.



AI coding assistants are dramatically lowering the barrier to entry for programming. As coding becomes easier, more people can and should learn it—not just for software engineering roles but for enhancing productivity across all professions.



While AI might make production software development 30-50% more efficient, it makes building quick prototypes up to 10 times faster. This allows for rapid innovation through frequent experimentation:

"I see also many large enterprises now starting to think maybe we can systematically build 20 prototypes, and if that's the price of finding two that work really well, that's fine so long as the cost of prototyping is low."

Ng suggests replacing "move fast and break things" with "move fast and be responsible"—encouraging teams to ship to small user groups safely while maintaining rapid development cycles.



Voice applications are becoming much easier to build, opening new opportunities for educational tools. Meanwhile, data engineering is increasingly important as AI can now extract value from previously underutilized unstructured data (text, images, video, audio).


Ng demonstrated tools from Kira Learning that showcase how AI can transform education. Their toolkit allows teachers to:

• Generate personalized lesson plans (in Ng's demo, he created a lesson on functions illustrated with monster truck examples)
• Align content with educational standards like AP Computer Science
• Create coding exercises with built-in debugging challenges
• Receive AI evaluation of content quality
• Get AI assistance for supporting students

Perhaps most impressively, the system includes tools to help teachers who aren't CS experts. When a student writes code with errors, the AI not only identifies the issue but suggests specific questions the teacher can ask to guide the student without giving away answers.

As Ng recounted his conversation with Kyle, the basketball coach turned CS teacher mentioned that he actually needed less prep time for his CS courses than for math, despite being newer to teaching CS—a testament to how these AI tools can empower educators.



Ng emphasized that AI is a general-purpose technology, and because teaching involves such diverse and complex workflows, AI can transform many aspects of education delivery.

Just as the workflow of educators is very diverse, the tools we need to support educators to support students will be very diverse. This is why AI as a general purpose technology will do many, many different tasks just like human teachers and human students do many, many different tasks.

Ng closed his presentation with an optimistic view of AI's impact on education. He believes we're at a point where the next generation should all learn coding and AI, as these will be essential tools throughout their lives.