Jordan’s argument is that AI is not the real problem, though. The real problem is how organizations think. AI can create more answers, more charts, and more work…but if teams do not have clear ways to think and decide, all that speed just creates more noise.
You use the metaphor that AI is like a loudspeaker, making whatever is there. Good thinking gets stronger. Messy thinking gets messier. Some teams feel calmer and clearer with AI, while others feel rushed and confused. But the difference is not the tool, as you highlight, it’s the thinking system around it.
Your argument is that most teams treat AI like a feature to install. You like to think of it more like plumbing. If the pipes are twisted, adding more water does not help. It floods the room. Better decisions come from better thinking systems…not faster output.
Here’s my overarching question: What would change if you focused on fixing how your team thinks, not just how fast it works?
Excited to dig in. Jordan Dalladay work connects closely to how we think about systems in Glare.
Great to chat with you today Jordan. And excited to be joined by your partner @craig_2. Let’s start off with perhaps a basic question, why even lean into AI initiatives in the first place? I’ve heard people suggesting it’s just a bubble.
Thanks Bryan, really happy to be here with you all today! Great question, and there definitely is a lot of hype around AI. For me, the thinking changed when I started not looking at AI as a set of tools, but a profound capability shift for individuals and organisations (like an amplifier), but only if we can distill the signal from the noise. The question I ask now is, will organisations be actually able to capture sustained value from this new capability, rather than just buying licenses to the tool.
That’s a great question @Bryan, my take is that there has been an AI Tool bubble as major tech organisations have pushed tools to justify huge investments. This had an unintended consequence that is only now materialising. That is that when there is cognitive abundance, most thinking systems in organisations are not fit for that reality. This is why we see big stats about AI value failure. The reason why people should lean in, is that when those thinking systems are redesigned for this new age the speed, breadth and depth of possibility increases - used correctly this can create strategic advantage.
Shaping perception is a big part of where this falls down, and the feedback loops people get with their team, I believe, are getting weaker- many people are just relying on thought patterns generated by AI and pushing slop. You touch on this.
This is great Craig. I think many teams are getting frustrated as they still can’t see the strategic advantage. How should they be looking at these investments? Are they looking for the wrong returns?
Ohhh… love this. So efficiency is fool’s gold? How might a product leader think about optionality when building a roadmap for a corporate strategy that may be stuck in old thinking? If speed isn’t the problem, how do they sell into the org?
That is a really great observation, and definitely one of the new risks these kind of thinking tools can introduce into the world at large, but also even in a small team working on a project or idea. The challenge is that these AI tools can act like amplifiers/scalers of both good and bad thinking. This unfortunately opens a new the door to the various cognitive biases and default behaviours we all have (confirmation bias, taking shortcuts, etc.) especially when AI tools can often respond with complete confidence, without any real knowledge of context. Within an organisational context, a common gap we’ve seen is the lack of established effective ways of working with these ‘thinking tools’ (like regularly challenging the machine, explicitly stating assumptions, rerunning analysis), and/or designed in specific boundary/refusal conditions into the way the tools themselves respond or interact (like understanding where presenting an answer with confidence may lead to a negative outcome). But tackling these points can lead to significantly sharper, and faster thinking!
In my opinion, if you lack strong curiosity, you are most at risk of running into problems with these tools.
Or is this not the way to think about this… is there cases where linear, task based thinkers can benefit? Are there new methods for teaching people how to think with these tools?
I wouldn’t say efficiency is fool’s gold. It just isn’t strategy.
Efficiency often optimises the current game. Optionality gives you a greater chance of playing the next one.
With the right mindset and approach, AI is more than a productivity tool, it’s an optionality engine that can compound learning, unlock new business models, streamline activities and reduce the costs of changing direction when market or customer conditions change.
This is great. I love Alex Smith’s perspective on Strategy. He suggests that strategy is all the things you choose not to do. It feels like AI is really good at producing outputs, but those same outputs can start to confuse people about what is really worth pursuing.
The idea of an optionality engine is cool. Do you provide ways to think about judging these ideas? How does critical thinking layer back into this idea of reducing the noise that might be generated by these tools?
Curiosity is definitely a key part to engagement beyond adoption; and unfortunately it is an area often under-invested in many organisations today. The consequence of this is often a small group of curious (and courageous individuals) that steam ahead, learning through ‘trial-and-error’, leaving the wider group behind until the point they reach a bottle neck. Which means we can get localised spots of productivity gains, but the overall time to outcome remains the same. Alongside a clear leadership mandate for curiosity and experimentation, one way this can be addressed is actually taking a leaf from lean management thinking, and developing a continuous improvement community - where the curious pioneers can share their learnings/findings with those less inclined, or even help their collegues find new ways of working.
It’s interesting how companies are trying to reorg, and many have never had to make such large shifts across the ENTIRE organization. Teams have different perspectives on how they want to work. I had an interesting dialog with @stephanie_muxfeld yesterday on the different futures teams have.
Are you seeing any specific patterns from teams that are implementing AI successfully across their companies?
That’s a great tension. AI dramatically expands what organisations could pursue, this makes strategic discipline even more important.
Optionality shouldn’t mean chasing everything. It means building capabilities that accelerate the emergence of multiple future paths, while being ruthless about which ones align with a chosen path.
AI should widen the lens. Strategy should narrow the focus. However, in places where cognitive ambience, scaffolding etc are sub-optimal and you have unharnessed abundance, there will always be chaos.
This is why a holistic and intentional approach from organisations is essential. Quicker stronger signal, not just more noise.
Yes, love this. I’ve been highlighting that teams really need to rethink their customer feedback loops. Feedback matters more than ever, but the tools on this side of the equation are still falling short.
When writing the article, I was asking myself “have any companies have managed to make this kind of scale/depth of changes before successfully?” The first example that came to my mind was the creation of Toyota Production System (think the origins of lean management theroy) - effectively designing and rewiring how the organisation sense-maked and acted as a whole - in order to overcome the 1950’s post-war market forces faced by Toyota (needed to completely rethink/evolve to compete). I think the challenge today is the number of companies having to make these changes all at once, and update their thinking systems from ones based on slow information+analysis scarcity, to ones designed to take advantage of having the close to the sum of the worlds knowledge, pre-digested and one well-crafted question away.
@Bryan this is something we often encounter, and in all honesty are yet to get fully formed answer to. What we know is most signals come from lagging insight (surveys, quarterly research etc).
When AI accelerates idea generation but feedback remains lagging or slower, then strategy drift occurs. You are making faster decisions with slower truths.
In one industry, we’ve looked at this challenge and experimented with Digital Twins, that would cut the time from signal to synthesis to as close to real-time as possible. The goal being to be able to leverage AI at any moment to respond to real-world conditions, demographic mixes etc and simulate the top possibilities.
The goal in the example cited is constant asset optimisation where traditionally signals could have taken weeks to emerge and an intervention made subsequently misses it’s moment.
Love all the insights @jordan_dalladay and @craig_2 today. Thank you for sharing your ideas on this super exciting topic. We’ll keep the thread open so others can jump in. Great stuff!
