Uploaded October 2025 | Updated September 2026, 1 hour ago
Meta learning flips the script: you don’t win by hoarding the most data - you win by adapting the fastest. In this episode, Eden sits down with engineer and researcher Liran Tam to demystify meta learning for founders. We cover how to get real performance from tiny datasets, when to use adjacent tasks to supercharge learning, and why personalization and low-compute inference are perfect use cases. We also dig into transfer across domains (from video to robotics), opportunities in cybersecurity, and how the moat is shifting from “most data” to “most diverse domains.”
If you’re an early-stage builder asking “Do I need Google-scale data to compete?” this one’s for you.
Please rate this episode 5 stars wherever you stream your podcasts!
Meta learning flips the script: you don’t win by hoarding the most data - you win by adapting the fastest. In this episode, Eden sits down with engineer and researcher Liran Tam to demystify meta learning for founders. We cover how to get real performance from tiny datasets, when to use adjacent tasks to supercharge learning, and why personalization and low-compute inference are perfect use cases. We also dig into transfer across domains (from video to robotics), opportunities in cybersecurity, and how the moat is shifting from “most data” to “most diverse domains.”
If you’re an early-stage builder asking “Do I need Google-scale data to compete?” this one’s for you.
Please rate this episode 5 stars wherever you stream your podcasts!
