Andrew Ng just dropped a 2-hour course on Graph Engineering: from Loops to full automation 9:14 - Your first agent 33:11 - Loop engineering 1:02:46 - Graph engineering 1:30:15 - Agents that rewrite themselves 1:49:05 - Full graph system Free, the best thing on graph engineering I've come across... Watch it, then build your first graph with the guide belowshow more

ng courses are solid but 2 hours is a lot when you can learn the same by building one broken agent and fixing it. the real lesson is loops are just state machines you don't control yet.

the graph engineering section is worth it but the 'agents that rewrite themselves' part is where it gets real. most people stop at static graphs when the real unlock is graphs that evolve based on what breaks. have you tried self-modifying graphs in production or is it still too unstable?

learning graph and loop engineering in 1 video - that's hardcore booked it

The real skill is designing how agents hand work to each other

that's most of what graph engineering actually is individual agents are almost the easy part, it's the contracts between them that take the most iteration to get right

I wonder whether he would choose graph engineering or context? I think context will be more important

they're more connected than they seem, the whole point of a graph is controlling what context each node gets, but if you had to pick one to get right first, context wins, bad context breaks good graphs

going forward dont think any of these matter in one/two months, agents are coming out pre-looped it seems

maybe, but that's how every abstraction has gone, someone still has to know what's underneath when it breaks

crzyyy

This focus on loops and graph systems is a powerful lens for building robust automation. Shifting from simply consuming information to architecting these systems is key to investor-level thinking.

@grok find this on YouTube