I’ve never seen such a massive, rapid, and optimistic investment in technology as we’ve seen with AI. 2025 capital spending by Google, Nvidia, Meta, Microsoft, OpenAI, and Amazon are over $900 billion, reaching almost 3% of US GDP.
Why the optimism? Businesses are convinced that AI is the productivity technology of the future, and the sooner we get our companies to embrace it the better.
But so far the real results have been mixed.
While people are dazzled by AI tools, US GDP has been dropping (from 2.9% in 2023 to 1.8% estimated in 2025, with -1.6% growth in the first half) and profits are slowing. IT sector profits are up 34% and financial services are up 10%, while profits in all other industries have plummeted, indicating that AI has primarily increased the profitability of AI companies themselves!
So this is a story about promises, expectations, and lots of good ideas.
I’m not arguing that AI is not amazing: it is. We revolutionized our research and advisory business with AI (Galileo®), and we’re able to grow with almost no new headcount.
But for large organizations, which have decades of bureaucracy, job titles, and layers, the path to productivity is tricky.
We have carefully studied this (we interview hundreds of companies and have a major study underway) and see dozens of good ideas. In HR alone there are more than 100 use cases and they improve hiring, employee support, development and productivity. Companies like Standard Chartered now use AI to evaluate performance and write reviews.
But as a recent MIT study shows, the real “re-engineering” has yet to come.
 |
Here’s the issue. As we explain in our four-stage framework, it’s easy for one person to use an agent to speed up their work. OpenAI finds that 41% of all usage is “information retrieval” and then there’s the benefit in writing, analyzing data, and locating answers to arcane questions. These personal productivity hacks fairly quickly.
 |
But at best this improves productivity by 10-20%. How do you get to “multi-process” automation and redesign how work get done?
This is the job of management, and that’s what I want to discuss.
I’ve been studying management for 30 years, and it’s quite a winding journey. For me it goes back to Peter Drucker’s book The Effective Executive, a book I re-read every year. Since then we’ve seen Jack Welch management (fire the bureaucrats), Howard Schulz (care for the employees), and trends like servant leadership, courageous leadership (Brene Brown), conscious capitalism (John Mackey), and then agile management (IBM) and holacracy (Zappos, do away with managers, an experiment that failed).
Management is a fertile ground for ideas, and many models are well understood. (Read “Irresistible: The Seven Secrets of the World’s Most Enduring, Employee-Centric Organizations” for ours.)
But I’m not going to give you another one: rather I want to discuss something else.
How is effective management changing in the world of AI?
This, to me, is the fundamental question, and I think we have an answer. Let me preview our findings.
Two Fundamental Things Have Changed
Over the last decade I’d suggest that two big things have changed: empowerment and experimentation.
- Empowerment: employees today feel more empowered than ever before, so managers must lead and coach people with enormous amounts of autonomy, information access, and powerful tools. The internet and pandemic released the chains we have on employees, and they’re never going back again.
- Experimentation: thanks to the democratization of tech, companies no longer transform from the top down, they change from bottom up. This means front-line and operations teams don’t only execute: they innovate and transform.
These two ideas, which have been debated in leadership for decades, are playing out all over business today. It’s a topsy-turvy new world, and companies ignore it at their own risk.
Look at companies like Microsoft and Meta, who shifted quickly toward AI with a culture of project-based leadership. Leaders at Bayer, Unilever, HSBC, Mastercard, Spotify, and Phillips pride their performance on small, empowered teams, each held accountable for improvement.
Today, unlike in the past, Supermanagers innovate without waiting for a senior committee to make a decision. They experiment, iterate, and drive change.
Here is the dynamic in action (see below).
 |
As technology evolves at a breakneck pace, individuals and job titles hold the organization back. (“It’s not my job.”)
Supermanagers break this pattern and take responsibility for growth. They experiment, empower people, and drive incremental change.
Today I’d argue that companies with traditional managers face risk: risk of stagnation, lagging productivity, and falling behind.
Supermanagers, as we discover in our research, are people who embrace new ideas, share the pioneering efforts of others, and bring practical applications of AI into our businesses without waiting for an “initiative” to come down from on high.
Why this shift toward the front line? It’s the nature of AI itself.
Prior technologies like ERP, the cloud, and even mobile required heavy investments from IT and software engineers to get built. A businessperson or manager had to wait for a new CRM or ERP system to arrive and it took years of investment to make these technologies work.
AI, by contrast, is the ultimate democratizing technology. Everyone can learn to use it and the most innovative and creative may actually be the youngest or least tenured. Why? They simply “learn by doing,” unencumbered by our IT practices of old.
This Superworker effect makes everyone a potential high-performer, and actually reduces the value of experience (Reid Hoffman calls this SuperAgency).
New ideas can come from anywhere, and those closest to the customer or process may innovate the best.
We still need IT governance, data quality, and security standards, but once that’s in place, people create in spectacular ways.
Our company now has more than 30 pages of detailed use-cases our clients have “discovered” in Galileo. The more we share them, the faster others can learn. This is the type of “Supermanager” behavior we need in this modern age.
What happens to the traditional ideas of supervision, work management, feedback, and performance evaluation?
Well AI makes this work easier by tracking what people do, enabling leaders to think more strategically and optimize the “system” rather than monitor each individual. It frees time from “supervision” to let managers focus on coaching, mentorship, collaboration, and the redesign of work.
This is why I’d argue that managers will never fully go away.
Performance management and supervision become the “table stakes” of management, and it’s the re-engineering, experimentation, and growth that differentiate the best.
How do we become experts and super literate about AI? How do we keep up as these things keep getting smarter? Well Supermanagers are not only strong leaders, they create a learning culture and embrace expertise wherever they can find it.
What about the hardline manager behaviors of evaluation, competition, and pushing people to do more?
These values remains important but now they’re reframed in the context of learning and growth, not just “doing your job” as fast and as hard as you can. “What have you learned today” vs. “what have you done for me today,” for example.
And remember, Superworkers aren’t robots, they’re people. They need tools, support, and a growth mindset to thrive. Supermanagers bring trust, advocacy, and empathy to this journey, enabling us to learn, reinvent, and grow.
We May Not Need Fewer Managers, Just Different Ones
This then leads to the question of the day: will AI eliminate the need for managers at all?
If all you’re doing is supervising work, then the answer may be yes. This kind of activity will be coordinated by agents and some of this “dead wood” (or what we always called “empty suits”) will go away.
But that’s not anything new, we’ve had the “empty suit” problem for decades. In the new world, where individuals operate as Superworkers, we differentiate leaders through their change and innovation, not only their performance. Empty suits disappear overnight.
Supermanagers also coordinate innovation across teams. They place calculated bets on productivity projects and halt work on ideas that aren’t panning out. And they bring people together to force knowledge sharing, alignment, and priorities that fit the rest of the company. (Galileo for Managers is designed to teach this skill.)
I’m not convinced we’re losing middle managers at all: rather we’re redefining what they do. Companies that drive Supermanager behaviors will rocket ahead in this new world.
We will be launching this research in October. Stay tuned for more on this topic, and if you’d like to share your experience or story, please let us know.
Additional Information
The Rise of the Superworker: A New World of Work for Everyone
Employee Engagement and Happiness Crisis. What Should We Do?
Introducing Galileo™ for Managers, The Leadership Guru At Your Fingertips
Source link
