AI Marketing Automation

In a candid and technically rich presentation at Tech SEO Connect, Kristin Tynski, founder of data journalism firm Fractl, shared her “obsession” with AI’s applications in content marketing, SEO, and PR, offering both a vision of dramatic change and practical tools for navigating it.

“I’ve been, I’d say, obsessed with AI and its applications in content marketing, SEO, and PR,” Tynski opened, before diving into how she’s overseen more than 5,000 content marketing campaigns that leverage large language models – always, she emphasises, “with the human in the loop.”

The AI Revolution Is Already Here

Drawing on her experience since 2018, when she created one of the first AI-generated blogs called “This Marketing Blog Does Not Exist,” Tynski painted a picture of an industry at a turning point. “It’s coming and it’s transforming everything, and this industry is among the first that will be transformed,” she warned.

Her perspective is backed by data: “GPT-3 had 100 million users in roughly two months, which is the fastest growth of any product ever.” But more significantly, she points to the quality of AI outputs, showing benchmark tests where models are “reaching parity with humans across almost every domain.”

A Challenge to Content Creators

Tynski’s central message cuts to the heart of content creation: “The future is really about the commoditisation of content.” She challenges marketers to ask themselves: “Am I creating something that other people could have created just as easily with a prompt in response from GPT?”

This isn’t just theoretical – she points to current trends showing approximately 30% of web content may already be AI-generated. “If you’re wondering if dead internet theory is true,” she noted, sharing examples of Google image searches dominated by AI-generated content, “just search for ‘baby peacock’ and see all the red ones are generative AI images… if you actually analyse them versus a baby peacock, you’d see there’s a lot of mistakes.”

The Evolution of AI Capabilities

Tynski mapped out the rapid progression of AI capabilities with characteristic precision:

• “Five years ago: Limited, poor quality expert development by humans, almost nobody was using generative AI”
• “Two years ago: Basic automation, large language models getting better… The challenge there was integration”
• “Today: Multitask agents… pipeline management, scaling and process automation”
• “Five years ahead: Autonomous AI teams, strategic direction provided by humans”

Working With Large Language Models

For those ready to dive in, Tynski shared detailed insights into working effectively with AI. “The most powerful thing in my mind,” she explained, “is combining a source of truth… with being able to connect to your AI to search and then having it iterate over and over again.”

She introduced several advanced concepts, including “agentic prompting,” where she initiates “two agents competing to find the real answer and poke holes in each other’s answers until they agree.” This approach, combined with proper management of context windows up to 200,000 tokens, leads to more reliable and nuanced outputs.

Practical Automation Examples

Tynski demonstrated several innovative applications she’s developed:

• One-input, many-outputs systems that can transform a single YouTube video into multiple content formats
• Location-based content generation using Google Places API and GPT analysis
• Advanced SEO tools combining multiple data sources for semantic analysis
• Automated brand visibility analysis in an AI-first search environment

“These sort of things took me like a day to write with GPT’s help,” she noted, emphasising the efficiency gains possible with proper AI implementation.

The Future of Marketing

Looking ahead, Tynski identified several key developments marketers should monitor:

• Agentic frameworks like Autogen and Crew AI
• OpenAI’s real-time voice API capabilities
• Advancing text-to-video AI technologies
• Edge computing developments enabling mobile AI applications
• The potential for “infinite context windows” that could fundamentally change how we search for information

Tools for the Community

In a significant move toward industry transparency, Tynski announced the release of 25 Google Colab Python scripts demonstrating various AI automation techniques. These tools provide practical implementations of her concepts, from content creation pipelines to advanced SEO analysis.

Navigating the Transition

While the changes ahead are dramatic, Tynski’s message wasn’t one of doom but of opportunity – provided marketers adapt appropriately. The key, she emphasised, is focusing on creating genuine value: “Am I creating something that’s actually Information Gain… presenting something entirely new or relevant, or is it just recapitulation of existing information?”

The presentation concluded with a clear message: the future of marketing will belong to those who can effectively combine AI capabilities with human strategic insight. “The end state of AI plus marketing… I think it’s near zero cost content creation,” she predicted, while emphasising that human oversight remains crucial.

For those interested in learning more, Tynski’s full presentation is available on YouTube, and she regularly shares insights and resources through her LinkedIn profile.