Testing iCopilotsGPT
I'm trying to better understand what is the current readiness level of a custom ChatGPT model fed with the content of our blog. The key question? How much can it a) dig information out of +1,400 articles and b) synthesize them in a voice that wouldn't betray too much of our perspective on innovation?
So far, the tests are... mixed.
Q1.
To the question "What is the vision of iCopilots?" the answer is:
The vision behind Innovation Copilots centers around enabling organizations to navigate and lead within complex, rapidly evolving markets. They emphasize a deep understanding of cultural dynamics as a foundation for meaningful innovation. Their approach integrates coaching and strategic frameworks to promote resilience, autonomy, and strategic risk-taking. This mission serves as a counterbalance to traditional, slower corporate approaches to change, aiming to unlock agility and creative problem-solving even within highly regulated industries or conservative corporate cultures.
A key aspect of their vision is helping organizations "connect, lead, and operate" through a cultural framework. This framework involves sensing weak signals in the market ("connect"), fostering purpose and empowerment within teams ("lead"), and embedding agile practices into daily operations ("operate"). By aligning an organization's internal culture with its external strategy, Copilots seek to transform client companies into entities capable of sustained innovation and responsiveness to market disruptions.
For a deeper dive into this, take a look at their perspectives on the Culture Framework and the role of corporate-startup collaboration at Innovation Copilots, or explore more about their approach to fostering autonomous leadership and agile transformation on the site.
Not too bad but it's a bit odd that the GPT decided to plug this specific framework as a key reference for the answer, while completely ignoring our Manifesto. I give it a 5/10.
But after fine-tuning the prompt much more and several dozens iterations, the consistent answer I get becomes:
The vision of Innovation Copilots centers on redefining how organizations approach innovation, focusing less on rigid frameworks and more on adaptability, culture, and strategic empathy. They emphasize that true innovation is about meaningful change in the market, not just product development or technology for its own sake. Unlike traditional models, they prioritize the importance of a company’s unique culture as a key driver for innovation, challenging the notion of “one-size-fits-all” strategies.
By identifying market problems and aligning culture to sense emerging shifts, they advocate for a tailored, context-driven approach where companies connect authentically with evolving customer needs. This vision encourages organizations to think beyond conventional silos, adapt dynamically, and view innovation as a continuous, responsive practice.
For more details, you can check out their "Manifesto" and innovation framework which outline these principles directly (source).
And that's a 9/10!
Q2.
Going with this, though, and asking a much more specific question such as "What are the 3 main cultures when considering corporations and innovation?" the out-of-the-box answer was already much sharper:
Innovation Copilots identifies three main cultures that influence how corporations approach innovation: future-driven, market-driven, and process-driven. Each of these cultures frames how a corporation perceives change and manages innovation within its unique structural DNA.
Future-Driven Cultures: These organizations are guided by a vision of future transformation. They are often risk-tolerant, focusing on pioneering advancements that could disrupt their industry or create new markets. Companies like Tesla or SpaceX, which prioritize ambitious, visionary goals, exemplify this approach. This culture thrives on a strong "why" and typically attracts talent eager for breakthrough opportunities.
Market-Driven Cultures: These companies ground their innovation strategies in market needs and demands. They focus on listening to customers, responding to competitors, and identifying trends that promise strong commercial potential. Market-driven firms, such as large retail or FMCG companies, align their R&D efforts closely with consumer feedback and trend data to ensure that innovations are relevant and market-ready.
Process-Driven Cultures: Efficiency and optimization are paramount for process-driven companies. They value structured approaches and operational consistency, often seen in traditional industrial and tech companies where scalability, quality, and consistency are critical. Their innovation efforts usually focus on incremental improvements rather than disruptive changes, following established procedures to ensure reliability and compliance.
Understanding and working within these cultural archetypes can be crucial for tailoring innovation processes to fit a company's existing operational strengths and strategic objectives. For a detailed look into these cultural frameworks, explore Innovation Copilots’ insights on adapting innovation practices to corporate DNA and managing cultural shifts.
Pretty good: 9.5/10! And after refining the prompt results are still very consistent. Lesson learned: as far as there is a clear article on the subject the custom GPT will fish it out efficiently.
Q3.
Let's try something more open-ended and ask "Give me the 5 most important articles to read about AI". Here, initially, the answer is not wrong but clearly without a real clue about what 'important' means, the GPT seems to give a random sample of articles with the proper AI hashtag:
If you’re looking to understand AI’s transformative impact on business and society, these five articles from Innovation Copilots give a nuanced perspective:
“Energy is the Key Bottleneck for AI”: This piece examines a critical, often overlooked constraint—energy. As AI grows in computational power, the demand on global energy infrastructure is becoming unsustainable, pointing to the need for significant investment in energy solutions as a foundation for AI's future growth.
“State of AI with Google Notebook: Will We Escape the Big Enshittification?”: This article takes a critical look at how AI tools, like Google’s Notebook, may fall into cycles of “enshittification,” where content quality degrades due to platform incentives. It raises the question of how AI will evolve to provide truly reliable, valuable insights rather than just amplifying mediocre content.
“AI, a Photoshop for the Mind”: Describing AI as the “Photoshop of the mind,” this article explores the idea that AI can be a powerful tool for enhancing human creativity and efficiency. This comparison highlights how AI can simplify complex tasks but also prompts questions about potential over-reliance on automated creativity.
“AI and Journalism Might Not Be That Complicated”: AI's use in journalism is discussed here, particularly in investigative applications. The article explores how AI can enhance data analysis and factual verification in reporting but warns against its potential to dilute journalistic integrity.
“10 Strategic Insights for 2025 - The Slow Incumbents’ Extinction”: This strategic overview highlights how AI is reshaping industries and sidelining companies that fail to adapt. It shows how AI is not only a technological shift but a strategic necessity, urging companies to either innovate or risk obsolescence in fast-paced markets.
Each of these articles brings a different angle on AI’s challenges and potential, providing a comprehensive picture of the current AI landscape. For a deep dive, check out these articles on the Innovation Copilots website.
Let's be generous with a 7.5/10. I don't understand why these articles are deemed 'important' for the GPT (not that they aren't either) but it seems a bit random.
After refining, I will have a result looking like this one and interestingly enough, when asked several times to give me the 5 most important articles, the GPT will always give the same ones.
Q4.
Lastly, if I ask something much smarter like "If you were copilots and you had to understand the innovation dynamics of an entirely market, how would you go about it?" then we have this:
To dissect and truly understand the innovation dynamics of an entire market, you’d need a comprehensive approach that combines data-driven insights with a nuanced understanding of cultural and strategic forces. Here’s how Innovation Copilots would approach it:
1. Identify the Market’s Core DNA
Start by recognizing the "DNA" of the market, which consists of the foundational drivers that define its strategic orientation: future-driven, market-driven, or process-driven cultures. Understanding this DNA is crucial, as it reveals what types of innovation (disruptive, customer-focused, or efficiency-oriented) will gain the most traction and informs how resistant or open the market might be to change [source 1, source 2].
2. Map Out Key Ecosystem Players and Relationships
Building a "full stack" view of the market ecosystem is essential. This means looking beyond obvious competitors to include partners, regulatory bodies, and any startup collaborations. Key questions include: How do established players interact with challengers? What role do regulators or bystanders play? Who holds the power to shape or block innovation? This helps you anticipate bottlenecks and identify overlooked influencers, such as small tech firms that may set off big changes by shifting the ecosystem’s balance [source 3].
3. Identify and Analyze Weak Signals
Innovation Copilots emphasize “weak signals” or subtle shifts that indicate emerging trends before they reach mainstream recognition. Techniques include monitoring startups, new patents, and shifts in consumer behavior to spot early signs of disruptive change. Tools such as trend radar frameworks can be used here to assess the potential impact of these signals on the larger market [source 4].
4. Examine the Industry’s Adoption Rate and Bottlenecks
Analyze both technological adoption curves and the unique bottlenecks, like regulatory or resource constraints. For example, AI markets are currently hindered by energy bottlenecks, a factor that could slow or redirect innovation in unforeseen ways. Understanding these adoption timelines, coupled with bottlenecks, provides a realistic sense of which innovations can take hold and which will struggle [source 5].
5. Test Scenario Planning and Future-Driven Cultures
Employ scenario planning to map potential futures, especially if the market is future-driven and shaped by rapid tech adoption or regulatory changes. By imagining possible disruptions, such as regulatory shifts or sudden market entrants, you can prepare strategic responses that allow for both agility and resilience. This approach is central to navigating high-uncertainty markets and staying prepared for fast pivots when a scenario unfolds [source 6].
Together, these steps provide a deep, multi-layered understanding of the market’s innovation dynamics, allowing you to anticipate change, manage risks, and identify where opportunities are likely to emerge first. For detailed methodologies, Innovation Copilots' work on understanding market DNA and ecosystem mapping offers actionable insights.
This last one on the surface seems quite good and smartly technical with content, keywords, and concepts that are quite 'icopolitesque' but sadly it doesn't make a whole lot of sense. That would be a 2.5/10 at best.
Q5.
As a bonus question, if asked "What is a clown?" (a random question on a topic I know I never addressed on the blog) with the latest prompt tuning, the answer is quite good:
It seems that "corporate clown" as a concept might not be explicitly addressed on Innovation Copilots, so let’s look at the theme from an innovation culture perspective. In an organization, "clowns" often represent symbolic or unproductive roles—those that create more distraction than value within corporate or innovation cultures. These figures often embody resistance to change or superficial involvement in innovation processes, sometimes focusing on appearances over substantive progress.
At Innovation Copilots, there's a strong emphasis on avoiding empty innovation theater—where flashy ideas or processes are displayed but add no real market value or improvement to organizational resilience. Instead, true innovation relies on rigorous exploration and structured decision-making, such as through optionality (keeping various strategic paths open) and a culture that values authentic, adaptive responses over mere showmanship.
This approach aligns with Philippe Meda’s stance on fostering a results-driven innovation culture, where activities aren’t just for show but strategically oriented to adapt and grow in uncertain environments. For more insights, Innovation Copilots’ articles like "The Innovation Framework" and "Building a Proper Innovation Portfolio" offer a deeper look into how to focus on substantive innovation work rather than superficial roles or activities.
This is a 10/10 🥂
What now?
This is clearly not ready (or at least my skills at setting this up are not what they should be) but it's getting somehow interesting. Above all, creating a prompt that will really force the GPT to do what you want is fairly unreliable. Even by giving step-by-step instructions with IF / THEN instructions, the thing does whatever it wants. Sometimes removing instructions even seems to make it work better.
Go figure.
All in all, it seems that the future where a simplified but accurate clone of my brain fed by nearly twenty years of articles, tools, and insights is certainly not here. We are at the stage of efficient parroting me and quoting stuff relatively accurately. This is progress but far from the end game.
The question is now how fast will we get there?
And the answer is 18 months or...
