Measured channelAI & Machine Learning

IBM Technology

Open on YouTube ↗

55 Gistil Score · median of 10 measured videos

Across 10 measured videos from IBM Technology, the median Gistil Score is 55 out of 100, and 2 of them came out worth watching outright. The median filler share is 44%. In a typical video the value starts at 1:34.

Videos measured
10
Filler · median
44%
Value starts
1:34

Last measured 19 Aug 2026

Start here

The three densest videos we have measured on this channel, each with the minute its value starts.

  1. 01

    Reinforcement Learning from Human Feedback (RLHF) Explained

    Worth skimming 11:29

    Density
    66
    Padding
    42%
    Useful from
    1:45

    A clear, dense technical explainer that actually delivers on RLHF's mechanics without any sales pitch.

    value from 1:45 ↗

  2. 02

    Diffusion Models for AI Image Generation

    Worth it 12:05

    Density
    64
    Padding
    33%
    Useful from
    0:48

    A clear, well-paced conceptual walkthrough of diffusion models with a genuinely useful worked numeric example — solid intro-level teaching, not deep enough for practitioners.

    value from 0:48 ↗

  3. 03

    AI vs Human Thinking: How Large Language Models Really Work

    Worth it 12:11

    Density
    61
    Padding
    33%
    Useful from
    0:49

    A clear, well-structured conceptual comparison of LLM vs human cognition — solid, accessible education with no filler or pitch.

    value from 0:49 ↗

Questions

Is IBM Technology worth watching?

Across the 10 videos we measured, the median Gistil Score is 55 out of 100, and 2 came out worth watching outright. The score measures how much of the runtime carries substance; it is our own measurement, not a YouTube metric.

What are the best IBM Technology videos?

The three densest ones we have measured are listed above, each with the minute its value starts — so you can open one at the second it begins to pay off.

How honest are the titles on IBM Technology?

Median title honesty across those videos is 87 out of 100 — how closely a title matches what the video actually delivers.

The Gistil Score is our own independent measurement of how much of a video's runtime carries substance. It is not a YouTube metric, not a public rating, and not the video's rating on YouTube. Gistil does not alter, download or cut any part of a video — the videos stay on YouTube and play there. Not affiliated with YouTube or Google.

How this is measured

Every number on this page comes from measuring the videos one at a time. How we measure.

All measured shelves