AI slop won’t break through the content economy. Austin Roth-Eagle of Cisco shares what actually does.
Austin Roth-Eagle leads AI adoption for global communications at Cisco, where he heads the company’s AI Acceleration Office. A former CEO communications strategist at Cisco and VMware, Roth-Eagle blends executive storytelling with hands-on AI expertise, including building scalable workflows, training teams and shaping how organizations communicate in the AI era.
The comms leader will join as a guest speaker at Ragan’s upcoming Writing Certificate Course. He will share how to break workflows into simple steps, decide where AI should be used and scale that work using tools like playbooks and agents to create authentic content over time.
What was your first role in comms?
I started my career in the Office of the CEO at VMware. My first role was primarily digital and program management — managing CEO social, owning dashboards and metrics.
One thing I realized very early on was that executives kind of have no business listening to a 21-year-old about social strategy or comms strategy — but they will listen when you have data. If you’re super well researched and have numbers to back up your assertions, they’re not listening to you necessarily — they’re listening to the research.
That sent me down a rabbit hole investigating the intersection between comms and analytics and the value of getting more technical in a storytelling role. Eventually, that grew enough that I decided to go back to school and get my master’s in business analytics.
I started in fall 2022, which was also when ChatGPT was released — kind of the starting gun for widespread AI adoption. I got to see a master’s program reinvent itself with AI at the forefront, and AI becoming a larger topic in my full-time role.
It started as finding ways to use AI to increase my leverage, and it grew into something much bigger. Now my entire day is focused on driving AI adoption across the organization.
As that work grew larger, we hit a point where it made sense to establish a true, cross-functional office focused on AI. That was the transition.
What’s something AI does that genuinely surprised you and something it still gets wrong?
What AI does really well — and still surprises me — is making workflows more efficient.
If I have an idea for content and need to get from point A to point B, a well-built workflow broken into small steps — via playbooks or prompting — can produce exactly the output I’m looking for. That still surprises me.
On the other side, the biggest thing AI is still missing is taste. You can ask it to write a compelling blog about technology, and you’ll get one of the worst blogs you’ve ever seen.
That’s where human judgment matters. We’re seeing huge progress in efficiency, but taste, judgment and voice are still critical.I’m still surprised by the amount of efficiency we can get, and I’m still disappointed by the lack of taste that AI has for particularly writing
How do you define your voice in a world where AI can mimic tone?
There’s a difference between using a tool to build content in your voice and having a tool replace your voice. If you have a clear idea and a unique point of view — even an outline — moving that into a well-written, formatted blog is a separate process. AI is almost like formatting, filling in gaps and cleaning up content.
Your job is still to have a unique point of view.
Across the board, AI-generated content without voice or taste is bad. We see it everywhere — what people call “AI slop.” A lot of those stories won’t break through. Your voice is the most important element. You need to understand your voice and use AI to amplify and scale it.
Do you have a checklist for deciding what’s ready to publish versus what needs a human pass?
It’s less about a checklist at the end and more about breaking the workflow into small, specific steps. Pre-AI, you’d research the landscape, identify your unique angle, build an outline, draft and then edit. When you break it down like that, it becomes clear where AI should and shouldn’t play a role.
There needs to be human oversight at every step. If you’re surprised by AI output, you should almost restart the process. You shouldn’t be surprised — you should be guiding the model from step one through step N.
So it’s not just oversight at the end — it’s oversight across everything. The human voice remains the most important step throughout.
How should early-career professionals think about building their voice today?
The best results come from people with deep domain expertise and strong AI fluency.
If you have AI fluency but no domain expertise, it’s difficult to move forward. If you have deep expertise but no AI fluency, it may also become difficult.
Young professionals should find that intersection — become an expert in something like writing or communications and understand what good, mediocre and bad look like. That’s taste and judgment, which models don’t have and may never have.
At the same time, learn how to get the most leverage from AI. People who understand both will thrive.
What’s one tool you use every day that isn’t AI?
Podcasts. I listen to a lot of podcasts to stay up to date. Morning Brew Daily is great. I’ve been reading their newsletter for 10 years. They also have Tech Brew Ride Home, which goes deeper on AI, and I listen to TBPN, a big tech podcast.
Personally, I also use my Oura Ring every day to track sleep and health.
Don’t miss Austin’s insights on writing with AI without losing your voice at Ragan’s Virtual Writing Certificate Course for Communicators on May 6, 13 & 20. Enroll now.
Isis Simpson-Mersha is a conference producer/ reporter for Ragan. Follow her on LinkedIn.
The post Cisco comms leader on the real problem with AI-generated content appeared first on PR Daily.
