The intrapreneurs are already here. The trouble is the handoff: AI lets everyone explore, but most companies have no way to turn that exploration into decisions the organisation will act on.
Vanessa, I’m wondering what happens six months after everyone has the freedom to experiment.
You could have a dozen people discovering better ways to work, but if nobody has the authority to retire the old processes, you’ve just added another layer of activity.
I’d want to know how many experiments actually replaced something the organization was doing before. Otherwise, innovation becomes additional work for people who were already stretched thin.
The speed of learning will matter more than ever. Essentially, the companies who can’t leverage their explorers fast enough and integrate that learning will lose both the talent and their competitiveness.
The people who have been unlocked by ai won’t stick around long. And there will be so many new entrants in every industry to pick that talent.
Vanessa, I think you’re right about losing good people when an organization can’t keep up with what they’re learning.
But I’m wondering about the employees who aren’t experimenting with AI. Some may be excellent at keeping the business running, yet suddenly find their experience carries less weight than someone’s ability to produce something new.
How do we keep those people engaged and learning without making them feel that everything they’ve spent years mastering has become obsolete? I suspect how leaders handle that tension will determine whether these changes actually last.
Oh I see. You know, to me these are the individuals whose judgment will count the most. This is the engine room managing the teams of AI agents, collecting the market responses, optimising. They will be incredibly valuable to the explorers as sounding boards for ideas. They are not just standing on the other side of the fence waiting to exploit the innovation, they are the ones at the heart of turning the innovative ideas into coordinated action and returning the learning back into the organisation.
The explorers are likely more generalist and less concerned with the details. The exploiters/engine room are where the experts who bring their deep domain knowledge to the table to make crazy ideas more likely to make sense in the real world.
Vanessa, your description of the engine room has me thinking about something else. What happens when those experienced people recognize that an exciting AI-driven idea simply won’t work in practice?
I’ve seen organizations where people were responsible for delivering decisions they had little authority to challenge. I wonder whether AI will make that problem worse. Are we prepared to give these experienced people the authority to stop a bad idea, or will we just expect them to make it work?
This is very informative. I particularly enjoyed how it gives me a new angle to understand the probing framework and identify how to move things from exploration mode to planned mode, and the key role of managers under these two distinct phases.
The shared decision record is the piece I keep seeing missing, and it needs a revisit trigger attached to the assumption it rested on. Without that line, a confirmed decision gets treated as settled for quarters, even after the pricing or staffing assumption underneath it has already moved.
Totally agree. I are working on a solution (Afoot), but I find the element of unstated assumptions rather fascinating. How rarely we make our assumptions explicit when making decisions. But knowing them and having them sit alongside the decison itself is rather crucial for the exact reason you state.
I’m not sure this is something AI can do for us. Reading it and routing it maybe, but the people making decisions have to understand them themselves first otherwise we go in circles.
My assumption is that we so often reward the illusion of certainty that few want to acknowledge that their decisions are made based on assumptions that may prove false down the line.
The groan zone point lands for me. AI multiplied the drafts, but the decision meeting still has the same hour and the same five people. My view is that a prototype should arrive with its plot attached: who has to act differently on Monday, and what happens next in finance or sales if it works. Without that, the receiver does all the sense-making alone. Have you seen a team make that "plot first" format a real entry rule before an idea reaches the planning cycle?
Not as often as I’d like. The times I have is when product marketing is involved earlier in the innovation process. But that brings only the GTM perspective.
I’ve been in tech too long. Places where everyone falls in love with their solution and idea far more than whether it goes anywhere meaningful.
The handoff is the part that rings true. Everyone can explore now, but exploration that never turns into a decision just becomes more half-finished work for someone else to sort through.
Exactly. We are just a little too excited right now. And with some new tools and more advanced models I’m not sure our excitement will ever subside. And we don’t really want it to. We just need a way to channel it all into something real.
Great point on innovation labs not working. Isolated AI efforts cannot work, i.e they won't spread to the company. You have to start with the experts where the jobs get done. And you need those expert to run the "innovation lab" inside their own department or team.
The main challenge is that you can't race a formula one when you just got your permit. You'll hit the wall on your first turn. Yet, everyone uses AI nowadays...
I teach AI to business experts and SMEs. The gap I observe between "I use AI everyday" and "I use AI to drive better outcomes" is abyssal.
It takes 3-6 months to build real AI fluency that can be applied in the job.
It assumes daily practice (few hours day on the tool) and accepting a whole new way of working. That transformation is hard with help, impossible on your own.
Vanessa, the volunteers in the network still having their day jobs raises a practical concern. Managers need to make room for that work, too. If exploring only happens on top of a full workload, the people expected to carry an idea into everyday use may already be stretched too thin.
It’s not on top of. There are a few studies (I’ll find them) that show that that performance wise it’s far better to allow teams to focus on either explore or exploit in a given sprint. So it’s either or for a fixed period. Either we recognize exploration as a legitimate and equal value creation activity or we get increasingly well executed sameness.
Give definition of a new term, I’m not gonna read the article to find out whether u explain what an intrapreneur is, nor am I not bothered to try decide what it is for you and BATTLE YOUR OPINION? Nope. Definition first please if u invent a word, then text.
Vanessa, I’m wondering what happens six months after everyone has the freedom to experiment.
You could have a dozen people discovering better ways to work, but if nobody has the authority to retire the old processes, you’ve just added another layer of activity.
I’d want to know how many experiments actually replaced something the organization was doing before. Otherwise, innovation becomes additional work for people who were already stretched thin.
The speed of learning will matter more than ever. Essentially, the companies who can’t leverage their explorers fast enough and integrate that learning will lose both the talent and their competitiveness.
The people who have been unlocked by ai won’t stick around long. And there will be so many new entrants in every industry to pick that talent.
Vanessa, I think you’re right about losing good people when an organization can’t keep up with what they’re learning.
But I’m wondering about the employees who aren’t experimenting with AI. Some may be excellent at keeping the business running, yet suddenly find their experience carries less weight than someone’s ability to produce something new.
How do we keep those people engaged and learning without making them feel that everything they’ve spent years mastering has become obsolete? I suspect how leaders handle that tension will determine whether these changes actually last.
Oh I see. You know, to me these are the individuals whose judgment will count the most. This is the engine room managing the teams of AI agents, collecting the market responses, optimising. They will be incredibly valuable to the explorers as sounding boards for ideas. They are not just standing on the other side of the fence waiting to exploit the innovation, they are the ones at the heart of turning the innovative ideas into coordinated action and returning the learning back into the organisation.
The explorers are likely more generalist and less concerned with the details. The exploiters/engine room are where the experts who bring their deep domain knowledge to the table to make crazy ideas more likely to make sense in the real world.
Vanessa, your description of the engine room has me thinking about something else. What happens when those experienced people recognize that an exciting AI-driven idea simply won’t work in practice?
I’ve seen organizations where people were responsible for delivering decisions they had little authority to challenge. I wonder whether AI will make that problem worse. Are we prepared to give these experienced people the authority to stop a bad idea, or will we just expect them to make it work?
This is a fascinating take - really great read, thanks Vanessa. Also love the design language!
thank you. 10 years ago this was out there maybe but now not so much.
The temporary chaos is inevitable as everyone adjust. I found this read interesting. https://x.com/pmarca/status/2114751107293749731?s=12
This is very informative. I particularly enjoyed how it gives me a new angle to understand the probing framework and identify how to move things from exploration mode to planned mode, and the key role of managers under these two distinct phases.
The shared decision record is the piece I keep seeing missing, and it needs a revisit trigger attached to the assumption it rested on. Without that line, a confirmed decision gets treated as settled for quarters, even after the pricing or staffing assumption underneath it has already moved.
Totally agree. I are working on a solution (Afoot), but I find the element of unstated assumptions rather fascinating. How rarely we make our assumptions explicit when making decisions. But knowing them and having them sit alongside the decison itself is rather crucial for the exact reason you state.
I’m not sure this is something AI can do for us. Reading it and routing it maybe, but the people making decisions have to understand them themselves first otherwise we go in circles.
My assumption is that we so often reward the illusion of certainty that few want to acknowledge that their decisions are made based on assumptions that may prove false down the line.
Idk 🤷🏽♀️ WIP thinking
The groan zone point lands for me. AI multiplied the drafts, but the decision meeting still has the same hour and the same five people. My view is that a prototype should arrive with its plot attached: who has to act differently on Monday, and what happens next in finance or sales if it works. Without that, the receiver does all the sense-making alone. Have you seen a team make that "plot first" format a real entry rule before an idea reaches the planning cycle?
Not as often as I’d like. The times I have is when product marketing is involved earlier in the innovation process. But that brings only the GTM perspective.
I’ve been in tech too long. Places where everyone falls in love with their solution and idea far more than whether it goes anywhere meaningful.
The handoff is the part that rings true. Everyone can explore now, but exploration that never turns into a decision just becomes more half-finished work for someone else to sort through.
Exactly. We are just a little too excited right now. And with some new tools and more advanced models I’m not sure our excitement will ever subside. And we don’t really want it to. We just need a way to channel it all into something real.
Great point on innovation labs not working. Isolated AI efforts cannot work, i.e they won't spread to the company. You have to start with the experts where the jobs get done. And you need those expert to run the "innovation lab" inside their own department or team.
The main challenge is that you can't race a formula one when you just got your permit. You'll hit the wall on your first turn. Yet, everyone uses AI nowadays...
I teach AI to business experts and SMEs. The gap I observe between "I use AI everyday" and "I use AI to drive better outcomes" is abyssal.
It takes 3-6 months to build real AI fluency that can be applied in the job.
It assumes daily practice (few hours day on the tool) and accepting a whole new way of working. That transformation is hard with help, impossible on your own.
I can only imagine. What analogy or framing do you find helps people understand what’s possible?
How do you help them not send slop grenades around because they got a little too excited?
Vanessa, the volunteers in the network still having their day jobs raises a practical concern. Managers need to make room for that work, too. If exploring only happens on top of a full workload, the people expected to carry an idea into everyday use may already be stretched too thin.
It’s not on top of. There are a few studies (I’ll find them) that show that that performance wise it’s far better to allow teams to focus on either explore or exploit in a given sprint. So it’s either or for a fixed period. Either we recognize exploration as a legitimate and equal value creation activity or we get increasingly well executed sameness.
Give definition of a new term, I’m not gonna read the article to find out whether u explain what an intrapreneur is, nor am I not bothered to try decide what it is for you and BATTLE YOUR OPINION? Nope. Definition first please if u invent a word, then text.
Ya. English is hard. Intrapreneurship was first used in 1978 and has been cited in nearly 700 academic papers.
You don’t need to read it. Lots of big words in there.
Get a husband pls