viraldi.
Back to discovery
19:35
Technology & Computing / VIDEO

Why Nobody Has Solved Continual Learning

LangChain· 208K subscribers

9d ago · English audio

Charlie O'Neill, Co-Head of Model Training at Baseten, explains why the train step in the own-your-intelligence loop is still unsolved once you update a model continually instead of once. He walks through the trade-offs: SFT can install new knowledge but is off-policy, and repeated SFT causes catastrophic forgetting and degrades general capabilities, while RL forgets less but can only upweight what the model already…

3.3×Score at discovery
14.7KViews
4.4KMedian · 38 videos

Outlier score

This score was measured on Sun, 25 Oct 2026 23:37:53 GMT. Current view counts may differ.

14.6K views ÷ a 4.4K baseline = 3.3×. Compared with 38 earlier videos from the same channel, matching the format recorded at scoring time. Older-upload baseline: reference videos had at least as long to gain views. This comparison is not age-matched.

YouTube snapshot: Mon, 26 Oct 2026 01:53:43 GMT. Counters may have changed.

Read the methodology

Artificial Intelligence & Data

Channel-balanced percentile: 39

Comparison details

70 videos from 47 other channels, with each channel weighted equally. Same format and language, with 100,000–999,999 subscribers. Observed at 226 hours after publication, within ±45 hours.

The comparison uses this broader topic because the specific niche has too few examples. This describes the collected sample, not all of YouTube. Updated Mon, 26 Oct 2026 02:02:08 GMT.

Niches & tags

Transcripts and detailed analytics are available for recent outlier Shorts up to three minutes. Research tools