Implications of AI amplification, job transitions, AI fluency, organizational collective intelligence, and more

📖In this issue

• Insights into job transition pathways

• AI to increase the collective intelligence of organizations

• Optimizing Humans + AI for knowledge graphs

• Kai Riemer on AI as non-judgmental coach, AI fluency, GenAI as style engines, and organizational redesign

• Ideas to execution faster, cheaper, better

🧠🤖Humans + AI

Insights into job transition pathways

New research on job transitions using real-time job posting data applies a "Skills Space" method that accurately predicts job transitions 76% of the time. It shows career paths are inherently asymmetric, meaning it's often easier to move in one direction between jobs than the other.

The study highlights how generalist skills act as bridges between specialized domains, and identifies clear pathways into specialized roles. For instance, a sheet metal worker could transition to becoming an industrial designer, moving from a high-automation-risk job to a more secure position. This kind of research is particularly important as work shifts, helping workers find new opportunities by leveraging their existing skills, targeting growing occupations and focusing on developing valuable skills.

Skill-driven Recommendations for Job Transition Pathways
Key insights from paper 

AI to increase the collective intelligence of organizations

The future of organizational performance hinges on amplifying collective intelligence through AI, as explored in Christoph Riedl's HBR analysis based on Anita Williams Woolley's research.

The framework identifies three pillars - collective memory, attention, and reasoning. While AI promises to optimize team dynamics and decision-making, success requires careful implementation that preserves cognitive diversity.

The key is treating AI not as automation, but as a catalyst for human creativity and experimentation, fundamentally reshaping how organizations create value through enhanced collective intelligence.

HBR: How to Use AI to Increase Your Organization’s Collective Intelligence
Anita Williams Woolley on Collective Intelligence
Summary of key insights

Better Knowledge Graphs using Humans + AI

Building useful Knowledge Graphs will long be a Humans + AI endeavor. A recent paper lays out how best to implement automation, the specific human roles, and how these are combined.

From human experts to machines: An LLM supported approach to ontology and knowledge graph construction
Key insights from paper

🎙️This week’s podcast episode

Why you should listen

Prof Kai Riemer has long focused on executive education on AI, and strong opinions and deep insights into how AI can best be used in a business context, including in strategic decision-making and augmenting cognitve workers.

💡Reflections

Ideas to execution faster, cheaper, better

Not just every year, but every day it gets easier to execute ideas.

One underlying factor is that the substrate of the economy and society is increasingly digital.

A large proportion of value is in fully or partial digital offerings. And even the most real-world experiences such as cuisine, haircuts, or community meetings are enabled or can be enhanced with technology.

The cost of developing and running software has fallen by orders of magnitude. In many cases a startup software development project that might have cost $1 million at the turn of the century would have cost $100,000 a decade later and perhaps $10,000 more recently.

Open source software - from operating systems to frameworks to modules - has created massive shortcuts.

Cloud platforms mean startups don’t need to buy, own, and maintain computers other than laptops.

No code and low code platforms and software libraries enable swift development.

Collaboration tools such as GitHub and Jira allow global teams to coordinate and allocate work to least-cost.

And now AI has dramatically accelerated the pace, with tools such as Cursor, Replit Agent, and Claude creating entire apps from text instructions.

This has long been an obvious trend, but the pace and extent will reshape the economy in coming years.

I am just beginning to think through the implications in more detail.

One of the most interesting is the new skills and capabilities to take advantage of this scaling.

You can’t use legacy methods and compete with those using the latest AI-enabled approaches.

Ultimately it lies in two domains:

Mindset. An attitude of always seeking maximum enablement, refining techniques, accelerating what can be done.

Technical competence. The level of technical capabilities required to build and scale software is rapidly reducing, to the point that coding ability is not required. But a certain level of competence and understand how tools work and integrate is essential.

I will be exploring this space in a lot more pragmatic detail, stay tuned!