AI can generate posts, videos, and avatars from start to finish. But brands need to begin with human strategy, insight, and story.

TikTok Generates the Video. But Who Is Making the Strategic Decisions?

TikTok’s Symphony Creative Studio offers a glimpse of where social media content creation is heading.

Give it a product description, URL, or a few existing assets, and it can help generate a finished TikTok-style video in minutes. It will generate scripts, visuals, produce digital-avatar videos, and support translation and dubbing.

For a small business with limited resources, that could be useful. For a larger brand, it could help test different hooks, create variations, localize content, and speed production.

But it also raises a question: What happens when AI can finish the content before strategic thinking even starts?

Who decided what the audience cares about? Who identified the insight and the brand’s point of view? Who judged whether the content was worth making in the first place?

Used well, tools like Symphony can help execute a strategy. But they shouldn’t replace the thinking behind it.

A Well Produced Commercial Is Not Necessarily an Effective One

The power and risk of AI-generated content remind me of something I learned years ago working in advertising.

A TV commercial set can be built well. The lighting can be right. The details can look convincing. The final edit can be polished. And the production value can be impressive. But it can still fail.

It can look good without connecting. It can communicate a message without meaning. It can be professionally produced, but still not give the audience a reason to care. The set is not the story.

That lesson is supported by research I did with Michael Coolsen. We analyzed 108 Super Bowl commercials and found it wasn’t the highly produced use of celebrities, animals, humor, or sex appeal that predicted likability. The underlying factor was whether the commercial told a story that resonated. Ads with more complete story arcs earned higher ratings.

We found a similar result in another study, “Drama Goes Viral: Effects of Story Development on Shares and Views of Online Advertising Videos.” After analyzing 155 viral ad videos, we found that YouTube videos with fuller story development received significantly more shares and views.

Production value can bring an idea to life, but it can’t replace the idea. AI makes that distinction more important than ever.

Use AI to Save Time. Then Spend the Time Better.

When I first started using a social media marketing simulation in class, I noticed something interesting.

The students who did well were not always the ones with the best post idea. They were often the ones willing to spend time on the grunt work of creating dozens of variations. They tested different headlines, rewrote copy, changed images, adjusted calls to action, and created platform-specific versions.

Through repetition, they learned that social media strategy is not about finding one perfect post. It is a disciplined process of creating, testing, learning, revising, and improving.

That used to be a big part of the lesson. It still is. But the work has changed.

Today, I don’t want students spending hours producing endless minor variations of posts. Generative AI can help with that. It can draft alternate captions, headlines, and calls to action, suggest image directions, and adapt content for Instagram, TikTok, LinkedIn, or X.

The same is true for social media professionals. AI can help teams create more variations, respond faster, localize content, test ideas, and stretch limited resources.

But the time saved should not automatically be used to create even more content. It should be used to think more deeply about the content.

AI Can Improve the Finish

One of the most useful applications of AI is helping people visualize ideas that might otherwise remain abstract.

In the past, a student could describe a campaign concept or create a rough sketch, but it was harder to show what the idea might actually feel like in the feed. A social media strategist faced the same challenge when pitching an idea to a client or internal team.

Now AI can help create sample posts, test visual directions, generate platform-specific variations, and produce rough examples of Reels or short-form videos.

In one of my classes last semester, students used an AI tool to create a full example Reel for Starbucks. That didn’t mean AI developed the strategy. It meant the students could show the idea more clearly. It also doesn’t mean a final Starbucks Reel wouldn’t feature people instead of AI avatars.

A good mockup moves a concept from “Trust me, this could work” to “Let me show you what this could look like.” For students building portfolios and professionals selling ideas, that is a meaningful shift.

It makes me think about my own experience. After college, I took my advertising portfolio around agencies in New York. Creative directors could see I had strategic thinking and creative ideas. But my finish wasn’t there.

A creative director at Cliff Freeman told me I wouldn’t get the job I wanted until I improved the finish of my portfolio. He recommended Portfolio Center. That is what I did.

Today, students and young professionals may face the opposite problem. AI can produce the finish. But the strategic thinking, human insight, and creativity may not be there.

A polished AI-assisted Reel is not automatically a good strategy. AI can improve the finish. You still need to develop the idea.

Marketers May Be More Enthusiastic About AI Than Consumers

Marketers and consumers are not on the same page about AI-generated content.

Research released by the Interactive Advertising Bureau found that while 82% of advertising executives believed Gen Z and Millennial consumers felt positive about AI-generated ads, only 45% of those consumers actually did.

That doesn’t mean audiences reject every use of AI. Context, creative quality, disclosure, platform, and message all matter. But we shouldn’t assume AI feels innovative or appealing to the people they’re trying to reach.

An academic study in the Journal of Retailing and Consumer Services found negative reactions when brands used generative AI to create social media content. People had lower perceptions of brand authenticity. Yet, the negative effects were weaker when AI assisted human creators rather than replaced them.

That distinction matters. AI-assisted is not the same as AI-replaced.

When Content Shock Becomes AI Slop

More than a decade ago, Mark Schaefer warned about “Content Shock,” the growing volume of digital content competing for a fixed amount of human attention. He recently revisited that idea in “How to Overcome Content Shock in a World of AI Slop,” arguing that generative AI accelerates the problem.

I think he is right. AI lowers the cost of creating content at the moment when creating more content becomes less valuable.

If every brand can produce more posts, videos, images, and synthetic creators faster and cheaper, feeds will fill with material that looks polished but doesn’t feel like it came from anyone. It may look professionally produced. It may fill the content calendar. But it may not mean much to anyone.

The brands that stand out now will not necessarily be the ones that generate the most content. They’ll be the ones that bring a real audience insight, a distinctive voice, a surprising concept, and a community that genuinely cares.

Start With Human Strategy and Story

I’m not saying students or social media professionals should avoid AI. It is too useful to ignore. The issue isn’t whether AI should be part of the process. It is whether people remain in control of it.

That means deciding what problem you’re trying to solve, what audience insight matters, and what story is worth telling — before AI generates anything. It means judging what output is worth keeping and what shouldn’t be published at all.

It also means doing the work AI can’t do for you. Listening to real comments and real conversations in a social media audit. Finding the human story. And before publishing, asking whether the content deserves to exist, not just whether it was easy to create.

AI can now finish content before the thinking even starts.

But brands still need to start with human strategy, insight, and story.

This post was created with the assistance of ChatGPT and Claude. The ideas, experiences, and opinions are my own.