Generative models, prompting, agents and AI tool capabilities.
Harness² Explained: Self-Improving Agents Without Training
2d ago · English audio
Harness² is a research framework from Google Cloud AI Research and Northwestern University that lets an AI agent rewrite its own harness (its prompts, skills, scripts, subagents and control-loop hooks) for every new task, without retraining the model. In this video I explain how it improves a harness with no answer key (Propose, Probe, Compose), why it has two levels of recursion with an experience library, what the…
Prompt Engineering
👋 I'm Muhammad! I am an AI/ML Expert with a PhD and have led ML teams at startups over the last 8 years. I am also a Google Developer Expert for ML/AI. The goal of this channel is to share my learning without all the "fluff" and "hype". I offer consulting services to startups and https://calendly.com/engineerprompt/consulting-call…
Outlier score
This score was measured on Mon, 26 Oct 2026 15:24:36 GMT. Current view counts may differ.
77.4K views ÷ a 23.4K baseline = 3.3×. Compared with 50 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 21:51:37 GMT. Counters may have changed.
Read the methodologyGenerative AI & Agents
Channel-balanced percentile: 94
Comparison details
20 videos from 17 other channels, with each channel weighted equally. Same format and language, with 100,000–999,999 subscribers. Observed at 57 hours after publication, within ±11 hours.
This describes the collected sample, not all of YouTube. Updated Mon, 26 Oct 2026 22:59:55 GMT.
Niches & tags
Transcripts and detailed analytics are available for recent outlier Shorts up to three minutes. Research tools