That top-down AI strategy is never coming
But that gives you freedom to play.
What is your organization’s approach to AI? How does AI enhance your products, improve operations, and level-up your customer service? What AI tools are your people using each day to drive productivity?
If your answers are: ‘Don’t know’, ‘It doesn’t’, and ‘I think we have access to the free version of Co-Pilot’, then don’t worry. You’re not alone.
Among the organizations I’m speaking to each week, the greatest AI experimentation is coming from the grassroots. From people who are interested and see an opportunity.
And according to Dr Markus Bernhardt, an AI strategy who joined us on The Mindtools L&D Podcast last week, that’s the norm. Says Markus:
'We seem to be waiting for this moment in a lot of organizations where some leader will suddenly say: "This is where the future lies, here's the AI strategy. BOOM! Off we go, now you know what to do.”
'This is not going to happen.
'It's your responsibility to start thinking about this and to start moving forward.'
The top-down strategy is never coming because:
🔥 AI is a foundational technology. As I wrote earlier this year, it’s not ‘one thing’. It’s infrastructure, which can then act as your colleague, assistant or sparring partner.
🦣 Creating an AI strategy for a foundational technology is a mammoth undertaking. Does your organization have an electricity or internet strategy?
⌛ Supposing such a strategy could be created, it would quickly be out of date. The pace of change is just too fast. What wasn’t possible last week can suddenly become possible today.
It’s up to you then to drive AI adoption for yourself, your team and, if you’re in an L&D role, for your entire organization!
🤨 How do you do this?
Some of our clients are already doing this well.
Mel Cooley, Head of L&D at Newsquest Media Group, runs regular clinics to share AI best practice within her organization. These cover things like how to use AI in meetings, or how to use it to quickly find answers to questions.
Speaking to my colleagues Gemma and Anna, Mel said:
‘The issue is less about the willingness to adopt [AI tools], but more about helping people see where it fits in their world. And there's still uncertainty and in some cases fear.
‘You know, once someone experiences speeding up a task they find frustrating, for instance, like generating a draft presentation or surfacing key actions from a meeting, then they can start looking at AI as support.’
Another of our clients asked us to design games that managers can play with their teams to help them explore what’s possible.
For example, using AI to create a script that pitches a random product. Participants have to play with prompts to improve their scripts, then perform to one another.
Or competing against AI in a series of challenges to discover where humans are better than machines.
💎 Rolling this out at scale
Mel’s approach is to act as an advocate, sharing best practice across her organization and inviting others to share theirs.
Our second client had us create facilitation packs, PowerPoint decks that can be shared with managers to help them build capability.
And last week we launched ‘Mastering AI for Managers’, one of our series of Skill Bites courses, this time presented by Dr Markus Bernhardt.
Each week, Mindtools users who subscribe to ‘Mastering AI for Managers’ receive a 10-minute session where Markus appears via video to share a particular tool or tip. Users are then encouraged to record how they’ll practice over the coming seven days, and in this way build their AI skills and commitment over time.
We hope with ‘Mastering AI for Managers’ to act as a catalyst for grassroots experimentation, and make AI a part of our users’ everyday practices. Try it out!
Current Mindtools consumer users and enterprise clients have access to ‘Mastering AI for Managers’ now. Just click ‘Skill Bites’ from the top navigation! Get in touch via custom@mindtools.com if you’d like to know more.
🎧 On the podcast
To hear that interview with Dr Markus Bernhardt, check out last week’s episode of The Mindtools L&D Podcast. Markus joined myself and Claire to share his insights into building AI capability over time, with more detail on:
Why a 'big top-down AI strategy' is never coming
How to help managers build AI skills
What to do if your organization bans AI tools.
Check out the episode below. 👇
You can subscribe to the podcast on iTunes, Spotify or the podcast page of our website.
📖 Deep dive — Manager Skills Framework (12 of 12)
We’ve reached the final week in our series exploring the 12 core skills that make up our Manager Skills Framework: social sensitivity.
Often this skill is grouped under emotional intelligence, but only recently have we begun to understand its impact on managers and teams.
👀 So, what is social sensitivity?
Social sensitivity is the ability to accurately read and interpret the emotions of others, using cues like body language, facial expressions, and tone of voice. It’s what we might call the step before empathy – before you can understand someone’s perspective, you need to understand how they’re feeling.
At Mindtools, we see social sensitivity as one of the most essential skills a manager can have – and just as importantly, one of the most valuable for managers to develop in their teams. Why? Because it improves the quality of decision-making, strengthens relationships, and enhances how people respond to the needs and dynamics around them.
It’s also a proven driver of team performance. In fact, research shows it can be a more reliable predictor of team success than intelligence, thanks to its impact on collaboration, communication, and psychological safety.
🧑🚀 Does it matter in hybrid or remote teams?
You might think that social sensitivity is a face-to-face skill only, but research tells a different story.
A study by Lacher and Biehl (2019) explored the role of social sensitivity in virtual teams. Participants first completed the ‘Reading the Mind in the Eyes Test’ – a popular measure of social sensitivity where people identify emotions based solely on the eye region of the face. Then, in teams, they collaborated on tasks and were evaluated on both their performance and communication.
Here’s what they found:
📈 Teams with higher levels of social sensitivity outperformed teams with lower levels of social sensitivity.
📞 These teams also communicated more often – not because individual members with high social sensitivity talked more, but because their sensitivity created space for others to contribute.
💪🏽 How do we strenghten this skill?
1. Model it: Call attention to non-verbal cues and gently check in to confirm your interpretation.
2. Build habits that deepen understanding: Encourage practices like summarising what someone has said before responding and reflecting on emotional cues.
3. Reward emotional insight: In performance conversations, recognise how people treat each other, not just what they achieve.
Lacher, L. L., & Biehl, C. (2019). Does social sensitivity impact virtual teams? In Proceedings of the 50th ACM Technical Symposium on Computer Science Education (SIGCSE ’19) (pp. 36–42). Association for Computing Machinery.
Quick personal note 👏🏽👏🏽👏🏽
Regular readers will have been impressed over the past 12 weeks by the rigor of our analysis and the many astute observations we’ve made about the skills in our Manager Skills Framework.
In the interests of improving my transparent communication (my own biggest skill gap), I want to acknowledge the contributions here of our Head of Research and Insights, Dr Anna Barnett!
Dr Anna wrote our ‘Building Better Managers’ report, and has been feeding us insights and papers these past 12 weeks as we’ve gone into each of the skills in detail.
The sharp insights are hers, any errors are ours.
Thanks Anna!
👹 Missing links
🕰️ The infinite workday: Microsoft reveals our preferences
There’s a difference between ‘stated’ and ‘revealed’ preferences. We state that we want work-life balance and to prioritize time with friends and family. But data from Microsoft reveals that many of us are reviewing emails at 6am in order to get them out of the way for the 8am pivot to Teams. The times when humans are most productive (9–11 am and 1–3 pm) get flooded with meetings, so then we spend our evenings sprinting through all those commitments we made during the day. We might say we hate it, but we’re doing it to ourselves. (via Martin Couzins)
👩🏼💼 Who wants to be a manager?
I’m not a huge Reddit guy, but Kenny Temowo spent some time pulling the best responses to the Reddit question: ‘What makes you not want to be a manager?’ Since we spend our days at Mindtools trying to build better managers, the answers are illuminating! My favourite of the bunch:
‘I used to love my job. Now I just manage the fallout of everyone else hating theirs.’
The short answer here is ‘no’. Given full autonomy by the team at Anthropic, an AI minifridge called Claudius offered ridiculous discounts, ordered bizarre stock choices and had an identity crisis. I’m not quite sure this is a fair test. Models like this are designed to give helpful answers, not run a business. But therein lies the key insight. If you’re asking Large Language Models (LLMs) to make decisions for you, it might be worth remembering that their objective is usually to give a pleasant response—not necessarily one that’s business-savvy. (Thanks Marc Steven Ramos for sharing this!)
👋 And finally…
The evolution of working from home. It’s been nuanced, thoughtful and focused on things that matter!
👍 Thanks!
Thanks for reading The L&D Dispatch from Mind Tools! If you’d like to speak to us, work with us, or make a suggestion, you can email custom@mindtools.com.
Or just hit reply to this email!
Hey here’s a thing! If you’ve reached all the way to the end of this newsletter, then you must really love it!
Why not share that love by hitting the button below, or just forward it to a friend?
I disagree on the AI strategy at the company level not coming. I also don't think the 2 situations are mutually exclusive. You can have experimenters throughout the company with or without an AI strategy. Having no AI Strategy could also be a strategy in itself and it would really depend on the business. Many companies strategies right now seem to be the same as a year ago - efficiency. I don't think this helps anyone as it is too vague.
I am pushing our company from both ends - we need a top down "Strategy" from the board and C-suite to focus the bottom up experimentation on the right areas. I have also introduced our C-Suite to the "Leadership, Lab, Crowd" framework that Ethan Mollick suggests. This covers all the bases
One other challenge for me in the framing of AI as foundational, which it is - "Does your organization have an electricity or internet strategy?"
No, nobody does as these things existed when the majority of the companies out there today didn't exist. 30 years ago, EVERY company had an internet strategy. I remember - I was there. I can't comment on the electricity strategy - I'm not quite that old.