Project Two: People make change - part 1
Individuals are using AI and deriving value from AI in their lives and their work. This will be the biggest factor in determining the nature and shape of how AI changes the world.
This is part 1 of my post on the first major theme of Project Two - People make change. This post is my own thoughts developing on this theme. In part 2 I will share the podcast and the article generated by Notebook LM on this topic. Subscribe to receive all Project Two updates in future.
Individuals are using AI and deriving value from AI in their lives and their work. This will be the biggest factor in determining the nature and shape of how AI changes the world.
The responses to Project Two confirm this trend. In this post, I will set out what the Project Two data shows and explain what that might mean for you and your business.
This is quite a long post so a short table of contents:
The AI Invisibility Cloak
Change is already underway
Making change stick
A surfeit of Cs
Summary
The AI Invisibility Cloak
A group of US economists recently put a number on the value already derived from AI - $97 billion of consumer surplus in the US in 2024 alone.
And our numbers actually line up neatly with the recent MIT State of AI in Business 2025 report.
“While only 40% of companies say they purchased an official LLM subscription, workers from over 90% of the companies we surveyed reported regular use of personal AI tools for work tasks.”
Only 14 out of the 150 respondents to Project Two were not using AI. Almost exactly the same percentage.
Yet businesses are struggling to realise these gains. MIT talks about a 95% failure rate in AI projects. The benefits are real but they are hidden beneath the AI Invisibility Cloak.
MIT describes this as a “shadow AI economy.” So it would be easy to see this as something on the margins, beginners playing with new toys. I think it’s much more important than that.
Change is already underway
First this trend once set in motion will be very hard to dislodge.
The second and third most common reasons given in Project Two for not using AI were “I don’t know what’s out there” (17 mentions) and I don’t have time to learn” (11 mentions).
This is an opportunity. At the same time, it creates an inbuilt barrier to any new or different system. People are using the things they have already found and learned. ChatGPT has the strongest foothold - 38 mentions. Way ahead of second placed Microsoft Co-pilot tools with 11 mentions.
And they have learned how to use these tools in ways that help them. They are learning more about the value of AI every day.
So you will find it very hard to persuade people to change to another tool set. The people who are using ChatGPT, Microsoft and Claude are already likely to be your change leaders. They are halfway down the road. Attempting to implement a different set of tools is asking them to come back and take a different road. That will feel like pain no matter how good your alternative is.
If you follow AI developments, you may be thinking there is another alternative. Plenty of vendors will be offering you AI with specific workflows for your industry. If you are big enough or early enough, they may even build agentic AI specifically for your company.
Why would those people out there using ChatGPT not switch to the bespoke version? Well if it is good enough perhaps they will. Or I should say, if it’s good enough they may at least choose to adopt your shiny new software. That doesn’t mean they have switched and stopped using the public models. Just that they are also using the corporate software for now.
At best they will gradually let the public models go. This is optimistic though. The big public AI platforms will keep getting better. If your people are using both, you are in a never ending race to keep up with the best big tech has to offer.
Note: Not the topic of this post but this is often described as a structural shift from the last wave of tech change. In practice, the advent of mobile and cloud computing turned into the value being concentrated in a handful of giant companies. Perhaps this is a continuation of an existing trend?
Making change stick
About 20 years ago we ran a campaign in PwC Consulting on the theme of “Making Change Stick.” It was all about how leaders could ensure success from change programmes, especially those involving new technology.
I still love some of the materials we used in that campaign but we missed the most important thing. The change that sticks is the change people choose for themselves. People in your team and in every business up and down the country are finding value by using AI. They ways they do this will be the changes that stick
“AI has no energy”
Part of this is a fundamental misunderstanding of the nature of AI. The machines are not going to take over our lives. We are in control.
So Rebecca is right. AI has no energy. The people using it supply the energy. That determines the direction in which the technology will go.
It’s not just raw energy. People care about what they use AI for and how they use it. Yes there is plenty of “slop” and low grade marketing bs out there. So what?
Project Two showed me a whole bunch of people who are using these tools because they think they can do better, or move faster, or get their great stuff to more people.
This is already translating into market activity. Amongst Project Two respondents 5 are starting new businesses based on AI. Two more are working in organisations where the whole product range is being realigned to be AI led.
EXAMPLE
An executive coach friend of mine, Kirsty Bathgate, recently launched an AI product called Bravyn. Here is how she described her reasons for taking this path:
People ask me this all the time. I still coach. I love coaching. It’s where I’m at my best and my happy place.
But here’s the reality: Executive coaching has traditionally been reserved for senior layers of organisations. It’s expensive, time-intensive, and simply not scalable.
Meanwhile, I’m watching first-time managers struggle with the same challenges I see in C-suite executives, just earlier in their careers, before self-doubt becomes cemented and poor habits harden.
When I coach senior executives, there’s often a lot to unpack. Layers of accumulated doubt, buried stress, learned behaviours that no longer serve them.
Imagine if we could intervene earlier. What if we gave people a place to work through their thinking and build emotional intelligence BEFORE they accumulate all those layers?
My vision in building Bravyn is to make leadership more competent, emotionally intelligent and easily accessible to all.
An even more striking example from Project Two. Writing the eulogy for a friend’s funeral came up as a use case. Tell me the people doing this don’t care about the end product.
Doing things we care about has a big impact on change. Both determining which changes happen and which changes stick. Much of the AI “debate” is driven by rationalist, logical, linear thinking. X gives the greatest benefit so this must be where to focus.
You can see this in the MIT report:
Investment bias: Budgets favor visible, top-line functions over high-ROI back office
People don’t care about cutting costs - of course we all know people who do care about this but I am generalising here.
Superficially, the Project Two data confirms this. There are lots of mentions of using AI for marketing, bids, pitches or other forms of business development.
No-one talks about using AI to cut costs. And no-one mentioned ROI as a motivator. Look a little deeper though and you can see that working efficiently is a source of value for many people. Productivity (12 mentions) features quite high on the list of use cases. There is a long tail of other mundane or admin type tasks that feature - form filling, finance analysis, summarising long documents and so on.
This is a classic pattern, familiar for many change programmes over the years. People are happy to cut costs when you attack little jobs they see as wasteful or a pain to do. You can even make people feel good about sacrificing small luxuries for the greater good.
Once the change becomes structural - systems, processes, goals, teams and so on are changing - things are very different. I will talk about how to marry this kind of wider change with the changes people are driving for themselves in a later post.
A surfeit of Cs
You may think this is all very well but it is just a large number of people doing things for themselves and making life easier. Change on a larger scale needs more than that. Absolutely true.
That misses another inbuilt aspect of human nature, the collective mind. Collective action gives this process a momentum of its own. Over and above the drive or lack of it shown by individuals.
Our collective mind is a whole field of academic study. It has a simple practical application. We don’t do stuff alone. We communicate, collaborate, compete and some other words beginning with C.
(Note: compete might seem counter intuitive in that paragraph. Underneath the adversarial surface, competition is one of the most effective ways the human race works together and builds bonds that drive great achievement.)
So your people are not operating in isolation. They are talking to colleagues, sharing with friends, discussing AI in the pub or on the golf course.
This is how human progress works. We are individually quite smart but our collective mind is exponentially more powerful.
(Incidentally, this is why comparing AI to the intelligence of individuals is pointless and irrelevant.)
Today, many AI use cases are personal. They don’t show in the productivity statistics or in your company’s bottom line. All that value is hidden by a sort of invisibility cloak. In the next phase people will start sharing and learning from each other. Flocking to what works
Looking closely at the Project Two data reveals lots of overlaps and connections. Yes, a small number of common uses stand out. But there is also a long tail of examples that are not quite the same but relate to solving the same problem.
So people talk about inspiring, drafting, creating, editing, improving and polishing content for example. Eventually this could evolve into a new and powerful way of communicating.
There are plenty of examples. The things that individuals are trying out today will emerge as new products that disrupt work in all sorts of unexpected ways.
This has massive implications for teams and organisations. Sharing and collaborating happens across organisational boundaries and beyond the edges of organisations.
You will see this well beyond your own organisation. AI will have unpredictable effects across our societies. Those effects will be created by the same people behaving in the same ways that you see inside your business.
So think of the uses people are finding for AI at work as a microcosm of the effects of AI. Both individuals and groups will be using AI in ways that matter more to them than work. Another significant barrier to change.
People are changing the way they do their own work. They are also changing how they work with others. Over time that will be a powerful force that reshapes how teams and organisational structures work.
Collaboration also spills over organisational boundaries. So in some cases this will end up altering the shape of companies, including large enterprises.
What is in-house v outsourced? Which activities give us a real competitive advantage? How people choose to use AI to work together will result in new answers to fundamental strategic questions like these.
Summary
Your people are already using AI. They are doing things they think are important and they are seeing the value.
The value and the use will continue to grow as people move up this curve.
In time, they will start to share that value with others. People will flock around the use cases that work.
That will extend to collaboration across tasks, workflows and value chains.
Eventually this will break down internal barriers within organisations and even go beyond organisational boundaries.
Collaboration and competition will lead to true disruption. New jobs, new business models and new industries will emerge.
So in simple terms go out and listen to what your people are doing. Observe flocking in action and follow the value. Encourage them and be open - don’t try to control and constrain. Focus on your corporate immune system instead.