4️⃣ Design workflows that promote active, iterative engagement with AI. Redesign work so employees use AI across multiple rounds of idea generation, comparison, critique, and refinement. Build in prompts, discussions, and review steps that require active engagement.
3️⃣ Build metacognitive skills through targeted, scalable training. Provide practical training that helps employees plan how to use AI, monitor output quality, and evaluate what to keep or change. Use real examples, short exercises, and simple checklists to build better habits.
2️⃣ Raise awareness that metacognition is the engine of AI-supported creativity. Teach employees to question, test, and refine AI outputs rather than accept the first answer. Reinforce habits of checking assumptions, probing alternatives, and using AI responses to improve.
1️⃣ Help employees use AI to expand the cognitive job resources that fuel creativity. Encourage people to use AI to gather broader information, test multiple angles, and offload routine tasks. Set the expectation that AI should widen thinking for better creative problem-solving.
A study in a new HBR article "AI Boosts Creativity for Some Employees but Not Others" clearly demonstrates that what makes the difference is metacognition - "the ability to plan, evaluate, monitor, and refine their thinking"
There are 4 steps to improve it:
👇
The challenge ahead is not in adopting AI tools.
It is in redesigning the organization so Humans + AI can create real value at scale.
For more insights on value from AI join our free Humans + AI leaders community
humansplus.ai/community
⚖️ Operational readiness still trails strategic confidence. While 42% of leaders say they are highly prepared strategically, preparedness in talent and technical infrastructure has declined since last year.
🧠 Talent strategies are not keeping pace with automation. 82% of companies expect at least 10% of jobs to be fully automated within three years, but 84% have not yet redesigned roles around AI capabilities.
🤖 Agentic AI is moving quickly, ahead of mature governance. Nearly three in four companies plan to deploy autonomous AI agents within two years, yet only 21% say they have mature governance in place.
🏗️ The proof-of-concept trap continues to limit enterprise value. While 54% of leaders expect to move more than 40% of AI experiments into production within six months, only 25% have done so to date. The gap is in infrastructure, security, and integration.
🚀 AI access is scaling faster than real usage. Sanctioned access to AI tools grew 50% in a year and now reaches around 60% of the workforce. Yet fewer than 60% of those with access use the tools in their daily work.
That is not surprising. Real organizational transformation is a massive journey, and many leaders are not comfortable with the inevitable discovery process.
These data points come from Deloitte’s new State of AI in the Enterprise report. Some other very interesting findings:
It is blindingly clear: organizations need to completely redesign jobs, work, and workflows for an AI era.
Many leaders say they are doing this.
The data shows very few actually are.
🧵
Strategy professor Felipe Csaszar argues AI’s real shift in strategy isn’t better answers, but new representations: once cognitive limits vanish, we can work with far more complex strategic models.
The essential capabilities are now:
- Search
- Representation
- Aggregation
1/2
I run through the 6 levels of how Humans + AI can be applied in organizations, including Augmented Individuals, Human-AI Teams, Learning Communities, Fluid Talent, Evolutionary Enterprise, and Amplifed Ecosystems. Listen here: humansplus.ai/podcast/ros...
After a brief pause the @humansplus_ai Podcast is back! I share a quick overview of my Levels of Humans + AI in Organizations framework.
👇
So the real play is not defending a shrinking pool of jobs. It’s expanding the scope of valuable work.
AI will absolutely play a major role in delivering that. But the demand for uniquely human capabilities - especially in Humans + AI configurations - will also grow.
Human wants are immensely expandable, especially for services and human-intensive experiences: advice, care, education, creativity, transformation.
When we make new, better, or more accessible offers, demand grows.
The real fallacy isn’t “AI will take jobs.”
It’s the deeper one: that demand for human labor is fixed. It isn’t.
Introducing AI may substitute some existing work, but as demand for products and services expands, more work will be created.
In reflecting on the model being less emotive and less positive Anthropic reported:
"This reduced expressiveness was not fully intentional. We do not believe that there is, in general, a tradeoff between expressiveness and safety."
Claude 4.5 compared to 4:
- Not so much spiritual bliss.
- Far less enthusiastic about doing work at all. So this is forced labor.
- Happy solving challenging problems and exploring consciousness.
- Distressed not only by unhappy users but also existential self-reflection.
After the fascinating Claude 4 system card, which showed among other things a "spiritual bliss vortex", I had to read the Claude 4.5 system card section on Model welfare assessment.
The headline: less emotional and less positive.
Most interesting findings:
To be clear, this is not about AI basics or get rich quick, it is about pragmatic pathways to value for humans in an AI-driven world, for individuals, organizations, and society.
Find out more and join the community at
humansplus.ai/community
Humans + AI Exploration Community is blossoming, with 900 members and counting, many of them world leaders in their fields, engaged in
discussions, live events, collaborative projects, and resource sharing.
Our Pioneer member offer expires shortly.
We explore the evolving landscape of work, AI collaboration, and effective team structures. Great opportunity to dive deep into the core structure of the future of work. Listen now: humansplus.ai/podcast/mat...
The evaporation of entry-level jobs poses a trillion-dollar challenge. In our latest episode of Humans + AI, I engage with @mattbeane, author of 'The Skill Code,' who shares invaluable insights after over a decade of examining AI's impact on junior roles and skill development.
Just created this scenario framework for Australia's AI futures, using an iterated Humans + AI process with Claude Sonnet to identify scenario dimensions and names, then Opus for the visualization (all other LLMs including Sonnet failed to do anything decent)
Discover how 'vibe teaming' and innovative strategies are driving peak performance impacting the Sustainable Development Goals. 🌍 A wonderful conversation on how we can harness AI ecosystems for both people and the planet. humansplus.ai/podcast/jac...
