What does Human-in-loop AI strategy mean in this evidence set?
Source-backed creator statements and evidence excerpts related to Human-in-loop AI strategy.
Topic evidence page
Source-backed creator statements and evidence excerpts related to Human-in-loop AI strategy.
Source-backed creator statements and evidence excerpts related to Human-in-loop AI strategy.
The creator argues that an expert plus AI can produce a stronger strategy than AI alone when expert critique is fed back into the model.
The creator argues that an expert plus AI can produce a stronger strategy than AI alone when expert critique is fed back into the model.
This topic currently has 1 source records, 1 public insight cards, and 1 creators in the public Base2026 export.
Please do not let AI come up with your digital marketing strategy. I have a really interesting story I want to share about someone who did this recently, and it illustrates how we're not yet to the point where AI can replace real experts in coming up with strategy...
OpenPlease do not let AI come up with your digital marketing strategy. I have a really interesting story I want to share about someone who did this recently, and it illustrates how we're not yet to the point where AI can replace real experts in coming up with strategy...
OpenShort public snippets grouped with their source record, creator, and date.
Please do not let AI come up with your digital marketing strategy. I have a really interesting story I want to share about someone who did this recently, and it illustrates how we're not yet to the point where AI can replace real experts in coming up with strategy.
Kind of like when you look back when computers could first beat the best chess players in the world. But then there was still this long period of time where a good chess player with a computer could beat a computer on its own.
I don't play a lot of chess, but my understanding is we've now reached a point where the computers are so good that the humans just get in the way. I believe that eventually that will be the case for all knowledge work, but we're definitely not there yet.
We're definitely still in the phase where an expert plus the AI is better than the AI on its own. And when it comes to developing strategy, that will determine where you allocate all of your resources.
You really don't want to get this wrong...
'd provided a bunch of context about their brand that Opus would need to put together the plan. But as I looked through the plan, it was very clear that they were going to waste a ton of resources if they followed this.
So I couldn't help myself. I made a 20 minute loom video breaking down the plan and all the problems with it.
And as someone who uses Claude every day to do SEO, the problems were very predictable. It recommended they spend time adding structured data to almost every page of the website.
Structured data that would have had no impact on their rankings. It recommended them targeting keywords that were either irrelevant or too competitive.
And it recommended making a bunch of micro performance optimizations that would have taken a ton of resources and had little to no impact on their rankings. Here's where it gets interesting.
This guy was smart. He took the transcript of that loom video, gave it the Claude, and asked it to rework the plan.
And the updated plan. The Claude came back with.
Was actually pretty good. Not perfect, but pretty good...