Humans + AI Capability Spiral, AI for mental health, metacognitive agents, and more

June 30, 2025

“New jobs will appear — very valuable jobs … humans are infinitely adaptable.” — Demis Hassabis

Interacting with AI for capability development and growth

There is extensive discussion and debate on the potential for LLM use to negatively impact our cognition.

Our outcomes are significantly driven by our intent. In my Humans + AI Capability Spiral I suggest a range of ways we can apply and use AI in ways that amplify our capabilities and growth. We need to embed these both in the design of systems, and in how we use the tools.

I’ve been working hard on some exciting projects which will be announced soon, stay tuned!

Be well!

Ross

 

📖In this issue

  • Framework: Humans + AI Capability Spiral

  • 22 new jobs created by AI

  • The potential for Generative Collective Inteliigence

  • Using metacognitive agents to assist Humans + AI co-creation

  • Podcast: Matt Lewis on augmenting brain capital, AI for mental health, neurotechnology, and dealing in hope

💡Framework: Humans + AI Capability Spiral 

How we interact with AI determines the value, which can be negative as well as positive if it leads to dependence and cognitive atrophy.

I suggest 8 levels of types of interaction, focused on human capability development and growth as well as outcomes. This is not really linear, but each level is appropriate to a particular context.

This is intended to inform two things: how we design our AI systems, and how we develop our skills in interacting with AI.

🧠🤖Humans + AI

22 new jobs created by AI

A feature in New York Times Magazine on new jobs created by AI drew on interview with leading figures in the space. My post generated extensive discussion, including on how new these job ideas are, and whether these jobs would last.

The potential for Generative Collective Inteliigence

Some of the speakers at Human+Tech Week, including BCG Henderson chair Martin Reeves and MIT professor Sandy Pentland, wrote an excellent paper to preface their panel discussion. These provide an important framing for Humans + AI.

Using metacognitive agents to assist Humans + AI co-creation

Metacognition is central to our ability to use AI well. "Metacognitive agents" can help human mechancial designers, also surfacing valuable lessons on effective agent design.

🎙️This week’s podcast episode

Matt Lewis on augmenting brain capital, AI for mental health, neurotechnology, and dealing in hope  

Why you should listen

Matt Lewis has been deep in the trenches of applying AI to wellbeing and mental health for over two decades. He provides an inspiring vision of what is possible and what we need to be doing to get there.

Thanks for reading!

Ross Dawson and team