The Looming Art of Gist Management

I’m gonna make a contrarian argument –

In the age of AI, content strategy has never been more important. AI has quietly forced content strategy to the forefront, through a dynamic that a lot of people don’t quite realize yet.

But let’s start way, way back –

I’ve been developing for the web for… 30-ish years? Since 1995, I think (the web was born in 1993, so this is pretty close to the beginning). I began in HTML 1.0 on Internet Explorer 1.0, if that tells you anything.

I remember writing my first HTML page in Notepad, and when I realized that whatever it put in the TITLE tag would appear in the title bar of the browser window, I felt like an elite hacker.

The first websites were intensely static, and built to stay that way. They only came one way. We designed for 640-pixels wide, because that’s basically the only thing we count on. Back then, you didn’t even get fonts – Times New Roman was it. Everyone got that same thing.

But, over time, the web has been in a steady, constant process of relinquishing control to the content consumer.

Browsers eventually got some sense of user styles. You could start modifying the way pages looked, for the sake of accessibility. And then we got true user styles. People could change how websites looked, ever so slightly.

The web was a visual medium. We were determined to control that visual. How things looked was as important as the content being delivered.

Slowly, monitors got bigger. Screen sizes expanded. We couldn’t count on standard widths anymore, so we had to design for the lowest common denominator. Then new browsers came out, and we added all sorts of hacks to figure out what browser you were on and account for it.

Eventually, monitors and graphics cards got better to the then-mind-blowing point where the monitor was sometimes bigger than you wanted your browser to be. So people would browse in a non-maximized window, which lead to all sorts of odd sizing.

And then RSS happened –

This was (and is) a niche protocol (yes, yes, I know you and I love it, but does your grandmother?). However, the key is that it first introduced the basic idea of restructuring content for consumption elsewhere.

Here’s a quote from the W3C release in 1997:

Code-named Aurora, the software will let PC users customize the way they view information on their desktops.

Content creators were suddenly confronted with the idea that their content could be stripped of its visuals and consumed outside their control (albeit with their permission). This was the first step in multi-channel publishing, and introduced us to concepts of channel rigidity – some things just didn’t work in RSS. A lot of us limited the information to a preview and drove visitors back to the website – where we had total control, dammit! – but others developed different templates and renderings for RSS.

And then mobile devices happened –

Suddenly, we couldn’t count on anything. So we created mobile sites (“m-dots”) to hold onto control as long as possible, then eventually conceded to responsive design.

This was a monumental shift, mentally. We were conceding that what something looked like was less important that the content. We made sure the content looked acceptable, but designs slowly became less complicated. We got back to simple single-column vertical scrollers (remember when we were floating everything…?). Since we couldn’t control how things looked exactly, we just made sure the content came through.

And now we have AI, and this is exactly the same thing, weirdly.

We’re not letting go of visual control anymore. We’re letting go of cognitive control. The current batch of AI tools is interpreting everything we create. It’s synthesizing it and translating it.

The effect on delivery is paradigm-shifting –

1. Visuals are basically gone. Content is being delivered in other apps and tools that we have zero control over, so designing something might become extraneous.

2. Exact wording and phrasing is even gone. AI is like a big game of “Telephone” – our exact words are being synthesized and changed and then delivered with the same claimed authority as the original author.

This shift is even more monumental than anything that has happened in the browser or front-end design wars. Not only can’t you control what your content looks like, but you can’t even control the words it’s delivered in.

(And no, video and audio won’t help. Before long, more people will read, watch, or listen to AI-summarized output more than they consume the source.)

So, where does this leave us? We’ve spent a long time designing content, and managing content, and now that’s becoming less and less important. I’ve seen predictions that the web will eventually succumb to “AI agents” that just gather stuff and do tasks for us.

Weirdly, the agent concept was always the idea. Tim Berners-Lee – the guy who invented the web – wrote a landmark article back in 2001 called The Semantic Web. This was the abstract:

A new form of Web content that is meaningful to computers will unleash a revolution of new possibilities

Berners-Lee was pushing RDF – a method of infusing the web with machine-understandable meaning, so that computers could understand it all and do things for us. For the founder of the web, the “perscriptive web” was just a pause on the road to his ultimate goal.

However, AI almost seems to have gone “around” RDF and achieved basically the same thing. Instead of being fed structure, AI is just trying to figure out structure itself.

AI is basically a new channel. Should we “publish” to it like we publish to social and email?

Which brings me back to my to my original point –

When you strip away control of the visuals, which we’ve been doing for 20 years, and you strip away control over the exact content, which AI is steadily doing now, what’s left?

The message.

The core point.

The… gist.

What’s the gist? That’s the basic idea. That’s the change you want to influence in the person who consume your content. It’s the idea you want to put in their head. The action you want them to take. The impression you want them to get.

In the future, all we can really count on is the suggestion… the gist.

In politics, people are constantly told to stay “on message.” Good political communicators know that with the frenzied, fractured, biased media world in which we live, they have to keep hammering away at simple points to be heard. They have to relentlessly push a core idea. They can’t control what happens to their words after they say them, so they have to stay true to the gist, or everything will get lost.

And the gist is where content strategy rules. You have designers and you have developers and you have copywriters, but technology is advancing to the point where the value these people add can be circumvented. Things don’t have to be designed, or programmed, or even written.

But the gist will always be there. The new content strategist is going to understand that they can’t count on anything anymore, they can only try to make sure the the idea gets through. They will ideate campaigns and write words and film videos with the understanding that 95% of people will consume them through some level of AI filtering, summarization, or contextualization. And they’ll make sure the gist keeps coming through.

It’s not visual anymore. It’s not words anymore. Instead, we’re trying to persuade machines to deliver the ideas we want to get across.

The new skill of content strategy is making sure you know how to teach the machine effectively.

In that sense, it’s a whole new world.