If you want to understand how generative AI is actually built, this Royal Institution lecture goes deeper than most

It is longer and more technical than the animated explainers that work well for introductions. But for people who have moved past curious and into genuinely interested, the additional depth is the point. Understanding the evolution from earlier language models through the transformer architecture to current generative systems changes how you follow developments in the field rather than just consuming news about them.

The specific thing this lecture does well: it shows the generative AI field as a progression of ideas rather than as a sudden invention. Knowing the lineage makes current capabilities less magical and more comprehensible, which is actually more impressive once you understand what the researchers were solving at each step.

This is a good candidate for a weekly learning thread or a deeper dive discussion for members who want to go further than tool tutorials.

For people with a technical background in the forum: where did your intuition about generative AI come from and is it the same place as where you would send a non-technical colleague to get started?