A New Way of Building with AI

Key takeaways

AI is moving towards automating everyday tasks effectively, aiming for practical applications that work seamlessly with users.
Designing AI systems requires a rethink on traditional software principles, focusing on AI’s preferred abstractions to enhance user interaction.
Implementing guardrails in AI programming is crucial to prevent common errors and enhance the quality of outcomes.
Balancing the cost and quality of AI interactions is essential for maximizing its utility without overextending budgets.
The future of AI involves creating adaptive systems that understand user preferences and improve creative processes across various fields.

Summary

Agents That Actually Do Work

In this section, Jiquan Ngiam discusses the role of AI in automating day-to-day tasks, such as parsing emails and updating spreadsheets. He envisions a future where AI acts as a co-worker that seamlessly integrates with our systems to manage and execute tasks for us. This practical approach emphasizes hands-on AI solutions that move beyond hype.

Designing with AI

The conversation emphasizes the importance of designing software that allows AI to operate effectively. Jiquan highlights a shift in focus from traditional software design to one that is flexible and responsive to how AI understands and interacts with software. This means creating abstractions and interfaces that cater to AI’s preferences, enabling smoother task execution.

Guardrails for Smarter Code

A significant theme in the podcast is the integration of guardrails for AI systems to prevent errors and mishaps in code generation. Jiquan explains how abstractions can help AI models perform their tasks without running into common pitfalls, such as data type mismatches. This approach seeks to balance flexibility and efficiency in software operations.

Understanding Costs vs. Quality in AI

The discussion dives into the trade-offs between cost and quality concerning AI operations. Jiquan points out that while lower costs for using AI models enable more frequent interactions and use cases, it also raises concerns about the complexity and quality of outputs. Organizations need to find a balance that allows for both effective and economical use of AI.

Future Possibilities of AI Integration

The podcast explores the potential future of AI within personal computing, emphasizing the importance of designing systems that adapt to user preferences. Jiquan speculates on how AI could help in refining creative workflows in industries such as video editing by providing personalized suggestions based on user history and preferences.

Show notes

Thanks to MLflow for supporting this episode — the platform helping teams track, manage, and deploy ML and GenAI projects with ease. Try it free at mlflow.org.

What if AI could build and maintain your software—like a co-worker who never forgets state? In this episode, Jiquan Ngiam chats with Demetrios about agents that actually do the work: parsing emails, updating spreadsheets, and reshaping how we design software itself. Less hype, more hands-on AI—tune in for a glimpse at the future of truly personalized computing.

// Bio

Jiquan Ngiam is the Co-Founder and CEO of Lutra AI, with deep expertise in artificial intelligence and machine learning. He was previously at Google Brain, Coursera, and in the Stanford CS Ph.D. program advised by Andrew Ng. He helped develop the first online courses in Machine Learning, and is now building agentic AI systems that can complete tasks for us.

// Related Links

https://www.youtube.com/@LutraAI

#api #llm #lutra #costefficiency #latentspace

~~~~~~~~ ✌️Connect With Us ✌️ ~~~~~~~

Catch all episodes, blogs, newsletters, and more: https://go.mlops.community/TYExplore

MLOps Swag/Merch: [https://shop.mlops.community/]

Connect with Demetrios on LinkedIn: /dpbrinkm

Connect with Jiquan on LinkedIn: /jngiam/

Timestamps:

[00:00] Agents That Actually Do Work

[08:21] Building Tables With AI Help

[12:54] Guardrails for Smarter Code

[16:35 - 18:00] MLFlow Ad[18:30] What’s Next for MCP?

[23:23] AI as Your Data Conductor

[31:13] Rethinking AI + Data Stacks

[32:10] Sandbox Security, Real Risks

[40:48] Smarter Reviews, Powered by Use

[46:08] Cost vs. Quality in AI

[52:00] Podcast Editing Gets Creative

[56:27] Transparent UIs, Powered by AI

[01:00:28] Can AI Learn Good Taste?

[01:04:45] Peeking Into Wild AI Futures