If 2025 was about breaking through the noise, 2026 is the year we went all in. Fewer picks, more depth — these three podcasts have been on permanent rotation at Carriff Digital this year. No fluff, no hype — just serious content for people building seriously with AI.
1. Latent Space: The AI Engineer Podcast
This is the one we keep coming back to. Latent Space is built for AI engineers by AI engineers — and it shows. The hosts break news, interview the founders and researchers actually shipping Software 3.0, and cut through to what matters: the architecture decisions, the model tradeoffs, and the infrastructure holding it all together. If you want to understand what the next three months of AI development looks like before it happens, this is where you start.
2. Machine Learning Guide
When you want to go back to first principles, this is it. The Machine Learning Guide treats ML as an audio course — covering the maths, the models (shallow and deep), the frameworks, and the mental models you need to actually reason about what's happening under the hood. It's the kind of content that rewards a second or third listen. We've had it on during commutes, workouts, and late-night sessions when we needed the fundamentals to click properly.
3. NVIDIA AI Podcast
Hardware and AI are inseparable in 2026 — and nobody understands that intersection better than NVIDIA. This podcast brings conversations from the people pushing the boundaries of what's computationally possible: researchers, engineers, and enterprise leaders turning GPU power into real-world transformation. It's the strategic layer that sits above the code — and it's essential listening for understanding where the industry is actually heading at scale.
Three podcasts. Three different angles on the same problem: how do you stay technically sharp while the field moves this fast? That's the challenge for 2026 — and these are the shows keeping us honest.