Let’s be honest, most of us are still wrapping our heads around generative AI… and now there’s this new wave called #AgenticAI.

Never heard of it? You’re not alone.

But here’s why it matters: #AgenticAI isn’t just about generating content, it’s about giving AI the ability to act, adapt, and pursue goals on its own.

I was thrilled to host Adit Sheth, Senior Engineer at Microsoft, on Deep Tech Talks for a conversation that dives deep into the future of Agentic AI, where machines go beyond outputs and start making decisions.

In this episode, we explored a powerful shift that’s already reshaping how AI is built, deployed, and trusted in real-world systems.

Here’s are the key highlights:

• What truly makes an AI system agentic and how it differs from #generativeAI?
• The tech stack powering this shift: modular architecture, semantic memory, and plug-and-play adapters.
• Why is building agents the easy part, and designing trust, control, and alignment is the real challenge?
• How agentic AI introduces new layers of risk?
• Real-world use cases.

It’s time to rethink AI. Not as a tool, but as a co-worker helping you move faster, smarter, and with more context.

Are we ready for that?

Have a question about agentic AI? Drop it in the comments, we’ll answer them in upcoming threads.