Is AI Actually Working? Many Would Say No.
AI promised to revolutionize the workplace, streamline our processes, and reduce our workloads. But has it?
Many would say no.
I’m writing this on a Friday, reflecting on my 20+ years in agencies. My colleagues and my clients both - we’re always right on the edge of burnout, aren’t we? Doing too much, wearing all the hats, trying to juggle it all while keeping up with the competition (and the boss). The demands are unrelenting.
Then came AI and the promise it would change everything. Yet many of us are still juggling, and a lot of folks would say that AI has actually just made things harder.
But what if they’re just using it wrong?
What’s Not Working?
Chatbots
Sure, chatbots can handle quick tasks like using API’s to fetch the weather or browsing Wikipedia pages to answer factual questions. But they’re incredibly limited. Chatbots work by following strict rules—like “If X, then Y.” They are programmed to sound conversational but often just frustrate users. In fact, a recent study from software firm Intellias shows that chatbots actually increase friction in customer journeys, especially during online shopping.
Automation Tools
The promise of AI was that it was supposed to make work easier. Unfortunately it hasn’t yet been the case. According to a 2024 study from Upwork, 77% of employees actually found their workload increased. The problem? People spend more time checking AI-generated content or even just learning how to use the tools. Layering semi-functional automation tools onto outdated workflows doesn’t solve the problem.
Complex AI Systems
Then there’s the issue of complexity and cognitive load—how much mental effort a person can handle in a given time period. Rather than freeing up brainpower, poorly-designed AI systems can do the opposite. Even well-designed systems implemented without the necessary training. People are left having to think even harder and process even more information. A 2023 study from Oxford found that poorly-designed AI systems can significantly increase cognitive load among clinicians, accelerating clinician burnout.
What is Working? AI Agents.
AI agents can go way beyond superficial automation and scripted responses. AI agents have the ability to analyze data, synthesize responses, even take autonomous actions. They are often built on large language models (LLMs) like ChatGPT, Claude, Llama, etc. They can learn, adapt, and collaborate. They understand context and guide a user through deeper, more creative thinking.
In my world, I tend to think of AI agents as creative mentors.
In other words, you’re not prompting them….they’re prompting you. They teach you new processes and vocabularies so you can quickly swim outside your own expertise and training. They ideate with you, and give you tips to push your own ideas further, getting past obvious solutions. They teach-by-doing, introducing new frameworks like design thinking, without the kind of heavy cognitive load a person would experience taking a course. They synthesize reams of research to help people uncover insights they might not have seen on their own.
AI Agents are Catalysts for Co-Creation
I’m a believer that innovation and creativity are massively enhanced through co-creation. When we bring all the right roles and disciplines together, we can solve the entire problem…not just one piece of it…in a matter of hours.
In this environment, AI agents are a game-changer. They open the floodgates of ideas, helping reframe challenges, and enabling breakthroughs. But it’s not the AI agent’s ideas that are typically at the top of the list at the end of the day – it’s the human’s ideas. The AI agent challenges and teaches the person to go further, and great ideas ensue. It’s not about replacing human creativity—it’s about enhancing it, bringing more people to the table, and empowering them to shine.
Quick Wins & Long-Term Gains
In the next few years, I believe the businesses that thrive will be the ones that adopt AI agents as co-creators and thought partners. I believe companies like these will lead the charge, not by automating people out of the process, but by transforming how their people think. If we use AI the right way, we’ll see wins in the near-term that are also sustainable for the long haul.
