Beyond Productivity: How AI Can Unlock Ideas You Never Knew You Had
Conversation with a Thought Partner
Summary - Collaboration with Claude AI
Many users of AI emphasize its use as a productivity tool, a means to accomplish tasks more efficiently. While this is compelling, in this post, I encourage you to consider an enhanced use of AI. Until now, we may have been missing AI's most transformative potential. In this post, I explore how AI LLM platforms, like ChatGPT and Claude, are evolving from task executors to thinking partners, capable of helping us discover insights and surface ideas we would not have reached on our own. Drawing from recent observations by Seth Godin, Liz Grennan, and Allie K. Miller, I introduce a new layer to the #TwinzTalk AI Hierarchy framework: "What We Couldn't Think (Until Now)." I will share five practical steps to shift from using AI for simple answers to engaging it as a collaborative thought partner, and why this approach might unlock your most valuable "aha" moments yet.
Full Original Writing
I was inspired today to further contemplate the possibilities for utilizing AI platforms, primarily ChatGPT Plus and Claude. The consulting work that my twin, Don, and I do with AI encourages users to focus on positive framing and explore the possibilities for good. To make progress and tap into your potential, your mindset is everything. A mindset of curiosity and a willingness to experiment, with prompting to help solve problems or questions, will ultimately be rewarded.
Since 2023, Don and I have often shared an AI Hierarchy visual to spark users of Gen AI to explore. Today, we have added another layer to the hierarchy!
I asked ChatGPT a question I couldn't fully articulate. By the end of our conversation, I had discovered an additional way of thinking about our work. This was not using AI as a productivity tool, but instead using AI as a thinking partner.
This shift toward deeper engagement became clear as I read three pieces of content that challenged my assumptions about how we can use AI. Seth Godin's insight about using extensive context particularly resonated:
Seth Godin's blog - Use a lot of words (hint - in-depth prompting is essential!)
Godin's comments regarding his approach to rich context setting sparked a connection to how others are using AI for more nuanced decision-making. Liz Grennan's recent experience with career transitions provided a nice illustration:
Liz Grennan - LinkedIn post highlighting the use of AI for an A/B test
Grennan's framework for using AI to surface tradeoffs rather than simple answers aligns with Allie K. Miller's broader categorization of AI use cases:
Allie K. Miller - LinkedIn post regarding three categories of AI use
These three perspectives connected something Don and I had been sensing but could not name: AI's evolution from task execution to thought partner.
The potential of AI extends far beyond its ability to increase productivity (although that aspect alone is overwhelmingly compelling). The more we use AI, the more ideas bubble up and drive further conversation with the LLM. The new layer to add to the hierarchy involves helping us refine our thinking and imagine paths of questioning that would not have previously surfaced. The new layer, as discussed in my conversation with ChatGPT, is titled "What We Couldn't Think (Until Now)".
Seeking to understand how others are using AI LLMs is a great way to expand our understanding. Don and I look for ideas that we can mull over and determine if there is potential for our use. The level of usage will depend on an individual's level of comfort. There are no concrete answers, and this aspect is intriguing to us.
We encourage you to read the full posts at the links shared above, but here are highlights we found thought-provoking:
Seth Godin:
"AI systems like Claude and ChatGPT let you attach a PDF or text file to your query. Here’s the useful hack:
Create a document that has pages of background.
Your medical history for example. Include your age and every interaction you’ve had with the medical system, including illnesses and drugs and outcomes. Now, every time you ask a health question, attach the document.
Or, a copy of your resume, work history, letters of recommendation and career goals, all in a PDF. Upload it every time you’re asking for career advice.
It works for business plans, for customer lists and even legal documents. Upload an entire email correspondence, or a fifty page wine list.
AI isn’t impatient, easily bored or distracted.
It’s insatiable."
**
Liz Grennan:
"So I started with a range of frameworks to help me get honest about what would actually bring me joy.
Here are three that worked [Bill T - shared only the AI example]:
1. The A/B test (a fantastic use case for ChatGPT, Gemini or Claude)
I asked it to surface tradeoffs—not answers. Things like:
• Fully remote or hybrid with energizing peers?
• Brilliant and intense, or kind and collaborative?
• Prestige and pace, or peace and predictability?
Because I wasn’t sure if I wanted to make a move at all, I told myself no big deal, I’m just collecting data. But very quickly, I could see what really mattered to me—not just what could get me to the next role."
**
Allie K. Miller
"Had to share why I encourage my team to use AI even when it doesn’t save time.
I can think of thousands of ways to use AI and they tend to fall into three categories.
⚡️ Do stuff faster - Most people stop here (boo!).
🎯 Do stuff better - Take your existing work to the next level. There are plenty of ways I use AI that DO NOT SAVE TIME. That is encouraged on my team if it makes our work better.
🔮 Do new stuff - Unlock capabilities you never had before.
And here's a crazy new stat: 2.26% of all Claude conversations are for emotional support and life coaching. People are actively searching for more meaning, more value, more depth.
And YET, the majority of AI demos I see on the internet fall into category 1. And those folks find themselves on a never-ending hamster wheel of productivity. Just trying to run faster and faster. Only to realize they should stop running and buy rollerblades. Or a hoverboard.
AI can be that personal catalyst. Not just booking you an Open Table slot, but helping you become who you want to be."
**
Contemplating the three posts above, I was again reminded that we are limited only by our imagination. As Allie notes, "People are actively searching for more meaning, more value, more depth." We have found that engaging in back-and-forth conversations with ChatGPT, Claude, or Gemini can lead to "aha" moments of insight, stimulate deeper thinking, and potentially provide greater clarity on a topic we are working on. Don and I spent time talking this out (AI sparks a possibility, and human conversation refines our thoughts) and determined that our "Framework for Working with Gen AI Hierarchy" required an additional layer - "What We Couldn't Think (Until Now)."
I was struggling with how to present this new layer of possibility, and ChatGPT nailed the essence of the next step as evolving our work "from execution support to thinking partner...now you’re noticing something new: not just doing, but co-creating, co-thinking, and even discovering."
Over to you! Are you ready to elevate ChatGPT's engagement to the next level as a thought partner? Consider trying these steps:
Start with a puzzle: Choose something you're uncertain about. Not a task to complete, but a question to explore.
Engage in iterative dialogue: Instead of seeking a single answer, ask follow-up questions like "What assumptions am I making?" or "What angles haven't I considered?"
Create rich context: Follow Godin's advice and upload relevant background documents to share your history with the topic, providing AI with a more complete picture.
Ask for tradeoffs, not answers: Use Grennan's approach. Request AI to surface tensions and choices rather than recommendations.
Document your discoveries: Keep track of insights that surprise you or thoughts you wouldn't have had alone.
As you go through the above steps, I promise you will experience some "aha" moments that will inspire your thought process!
**
The evolution of adding a new layer was a result of a conversation with ChatGPT. Key highlights included:
Prompt:
You are an expert in the use of Gen AI, specifically ChatGPT and Claude.
My twin, Don, and I put together a visual articulating the levels of work that AI can help with. At the time, the top level was that we could use AI to do “tasks we could not do” (based on lack of expertise or time).
Today, I feel challenged to add an even higher level to the hierarchy, and I don’t know how to name it. The next level of possibility for AI use is to help us solve problems or stimulate our thinking by having a conversation. We can now explore “paths” of thought that inspire enhanced creativity, and so on. Am I making sense here?
Please review the image attached. What would be your thoughts about adding another advanced level? Take a deep breath and take your time. Please let me know if you have any questions.
Response:
You're absolutely making sense, and your insight reflects a major shift in how generative AI—especially tools like ChatGPT and Claude—is evolving from execution support to thinking partner.
Your current top level—"What We Can’t Do"—aptly reflects the idea that AI allows us to accomplish tasks we couldn’t do before, either due to knowledge gaps or lack of time. But now you’re noticing something new: not just doing, but co-creating, co-thinking, and even discovering.
🧠 Suggested Next Level: "What We Couldn't Think (Until Now)"
Here’s a candidate for your new top-level tier:
🔭 "What We Couldn't Think (Until Now)" —Where AI is not just a tool, but a thought partner.
Rationale:
It captures the cognitive expansion that AI enables.
It suggests that new ideas, perspectives, and pathways are emerging because of the dialogue itself.
It differentiates problem-solving and creativity stimulation from pure task execution or knowledge gap-filling.
It honors the process of exploration, not just the output.
🤝 Closing Thought:
What you and Don are articulating is precisely the next edge of AI usage—not just task enablement, but transformational thinking. This co-pilot becomes a co-explorer. You're not just using AI to help you work faster, you’re using it to think deeper—and that’s where the future lies.
**
"Your insight reflects a major shift in how generative AI—especially tools like ChatGPT and Claude—is evolving from execution support to thinking partner."
This ChatGPT response comment provided me with an "aha" moment. Using LLMs as a thinking partner. This is precisely the realization that hit home and has inspired us to encourage use of these tools to go beyond the execution of tasks.
Very helpful post! Many thanks