Every day seems to bring yet another “AI-enhanced” product with it. My text editor, photo editor, music player, code editor… heck, even my calendar app is now “ChatGPT-enabled.”
One would be forgiven to think we are high on Gardner’s “Peak of Inflated Expectations” – I believe we haven’t seen anything yet. With open-source LLMs emerging and thus costs going down, entrepreneurs will slap AI onto everything – if it makes sense or not.
Brace yourself for a cacophony of noise and finely tune your (strong) signal-spotting capabilities.
Read on to learn why “fail fast” just isn’t enough.
In our work with clients on using a prototyping approach to drive innovation in their companies, we always emphasize that simply learning to fail fast (i.e., the sheer number of experiments) isn’t enough. The key that really unlocks the innovation power of a rapid prototyping process is the design of a larger system that maximizes the learning value of each experiment. A system that makes sense of “failure” to extract insights and then directs/shares those knowledge gains appropriately to best and most quickly support the next round of experimentation.
What holds true for prototyping, not surprisingly, holds true for pilot projects looking to leverage emerging technologies. Systems that make sense of and share learning matter. Connective tissue built between dispersed initiatives and innovators matters — particularly in situations where the business case for adoption may be evolving in real time along with the value / performance / ease-of-use of the developing technology itself.
As the authors of a recent article (MIT Sloan Review) put it: “Budget-centered business case approaches are biased against novel technologies, partly because they don’t factor in the value of learning gains and spillover effects.” These researchers looked specifically at how IKEA evolved a framework for coordinating and evaluating pilot projects aimed at using drones to monitor warehouse inventory – all during a several-year window that also saw drone tech become more autonomous, reliable, fit-to-purpose, and cost-effective.
After a few isolated pilots in 2018-19 had shown the proposed solution to be expensive, impractical, and potentially dangerous, the use case could have simply been written off as hopelessly removed from a sensible business case. Instead, IKEA chose to connect and manage further drone initiatives through a recently established global innovation governance unit. That decision facilitated better coordination and knowledge sharing between pilots across the global org and allowed the potential business case to be “iteratively developed” as the company learned through experience and the technology itself evolved. By the spring of 2023, IKEA’s coordinated approach had autonomous drones deployed for inventory management in 16 warehouses in seven countries – specific contexts identified where the solution and the business case both make sense.
As the pressure to innovate, the pace of change, and the availability of new technologies all increase in tandem – and swiftly, the value of systems and structures designed to support purposeful experimentation and learning at scale is surely going to compound.
So the questions for org leaders: What structures are you building today to maximize the value of experiments tomorrow? How are you designing systems to make sense of and facilitate learning across the organization? (via Jeffrey)
💼 How Tech Is Transforming Entry-Level, Customer-Facing Jobs A survey of almost 900 customer service reps revealed that while some employees outperformed others, it wasn’t because of their level of experience. Instead, what mattered was if their technology connected them to the information needed to complete their tasks. The author argues technology must sit at the heart of future talent strategies. Jane ⇢ Read
🧨 Innovation Doesn’t Have to Be Disruptive We have an obsession with disruption, although market-creating innovation isn’t always disruptive. Disruption may be what people talk about, but it’s only one end of the spectrum; the other end is what they call nondisruptive creation, through which new industries, new jobs, and profitable growth come into being without destroying existing companies or jobs. Mafe ⇢ Read
🕵️♂️ America Forgot About IBM Watson. Is ChatGPT Next? The history of IBM’s Watson suggests a possible path that today’s hyped AI systems could take toward rather disappointing, productized futures. Jeffrey ⇢ Read
⌛ The business trend that unites Walmart and Tiffany & Co We talked about the hourglass of prices before, and recent market data confirms that picture, with companies looking to leave the middle. Julian ⇢ Read
🍏 What I Learned as a Product Designer at Apple Despite being counterintuitive, these takeaways from working at the world’s most valuable company, make sense. Pedro ⇢ Read
🧮 Think of language models like ChatGPT as a “calculator for words” One of the best analogies for LLMs – worth a read and think. Pascal ⇢ Read
🙀 Hinton on AI: Society Might Not Be Prepared for What Is Coming.
👪 This playbook teaches tech companies how to hire and support formerly incarcerated people (be radical’s Learning Partner Ryan Merkley co-authored the paper!)
🧑💼 A rare look into what actually happens to productivity when companies adopt AI.
🧑🍳 Here we go again – cooking robots are back again.
Vinyl record sales are breaking record after record. Turns out – many people buying vinyl, don’t actually own a record player… 🤔
|🏴☠️ The Heretic: Give a Damn||🎧 Listen||📺 Watch|
🧨 Disrupt Disruption - The Podcast: We brought Christina Nesheva, CEO at Officinae Bio, back to talk about leadership. Listen now.
📕 Disrupt Disruption – The Book: Get your copy of our bestselling book and learn how to decode the future, disrupt your industry, and transform your business here.