The Understanding Group

05/28/2026

Bob Royce's excellent presentation at the Allovance Summit about the use of strategic tools when planning AI implementations. Ethics! Centers of Excellence! Strategy! Check it out!

AI Strategy using Allovance

05/21/2026

More meditation on how LLMs aren't eliminating work, but fundamentally changing it. The smart companies will leverage this approach instead of a reductive replacement theory, because the outputs will just be better, not necessarily faster.

one of the ways to sense the shift in product right now — notice how much time you spend on IF you should build something. this one question actually has three components: #1 risk you accept, in relation to #2 how much effort is needed to build, and #3 will people want it in the age of AI-native w...

05/19/2026

Do we know how to formulate questions? Do we know what the scope of our need is and what kind of information we are trying to get?

Richard Saul Wurman and Dan Klyn's periodic table of questions gives you an amazing tool for figuring out the nature of what you want to ask, and why.

05/14/2026

New technologies fit into two boxes: do the thing you used to do more efficiently, or do what you used to do differently. stress test both those assumptions, and their impact on coding is having profound effects on the whole software development lifecycle. Or, as Damola says here,

"What’s left?
Context and observability. 👀
The new skill isn’t writing code.
It’s context engineering."

AI didn’t make developers 10x faster. It killed the SDLC. 💀 For 20+ years we optimized a linear pipeline: Requirements → Design → Code → Test → Review → Deploy → Monitor. Each stage had tools. Rituals. Owners. Entire industries. Now? Intent → Agent → Build + Test + Deploy → Ob...

05/12/2026

TUG's Emily Claflin reflecting on her experience as the chair after a successful, fascinating conference. Well done, Emily!

During a break at last year’s IAC - Information Architecture Conference, Jacob Geib-Rosch walked up to me and said, “I hear you might be one of the chairs for next year’s conference!” I hadn’t heard anything about that, and I didn’t know what a conference chair was, and I had never met J...

05/07/2026

This is asking more than "who reaps the productivity wins"; Alastair is asking "is the reward for being more productive an expectation that productivity will dramatically increase at the expense of workers' well-being?

Yesterday I built a dashboard for one of our datasets in less than 30 minutes using Claude Code. It was just one of three apps I created yesterday. Prior to AI coding that same work would have taken me days, or even weeks, to complete. It raises an interesting question: if I can do in an hour what w...

05/05/2026

Communication succeeds when it supports feasible action, not when it overwhelms decision-makers with competing options or unresolved uncertainty.
Seen this way, ineffective communication is not about weak storytelling. It is about misreading what policymakers need evidence for at different stages of the process.
Understanding this shifts the question from ‘How do we persuade?’ to ‘How do we enable defensible decisions?’

This is so true, and applies to any high-level decision maker.

To communicate effectively in policymaking systems, actors need to understand how policymakers process evidence and the environment in which they operate. Therefore, we combine psychology and policy studies to produce a three-step strategy. First, do not bombard people with evidence. Human bein...

05/01/2026

The thing that's most interesting about this Linkedin Essay is that the patterns it describes reflect a fundamental truth to the way that organize and share information -- aggregate, bayesian weighted neural nets.

But human creativity comes from similar systems, so another way to think about this is that there may be opportunities to discover or develop creativity in these neural nets similar to humans. I wonder if cognitive psychologists have any thoughts on this.

Is AI the "Mp3" of Human Thought? Here's a metaphor for semantic compression. The public has spent decades getting used to the "flatness" of an MP3. To save space, the algorithm performs perceptual coding—it literally throws away the frequencies it thinks our ears won't miss. It shaves off the son...

04/21/2026

Excellent insights from Christina Wodtke which are becoming part of a recurring theme, which is that AI shouldn't pressure you to speed up; instead it invites you to take the time to collaboratively figure out what is going on because the development process can be so efficient.

I had a ton of mixed feelings listening to this, and I recommend you listen to it yourself and decide for yourself. The story of designers being unable to keep up because engineering is now going fast was less interesting than her talking about how she spends a lot of her time jamming with engineers...

04/16/2026

A very good article about how to approach AI with Information Architecture concepts. Service Design is one of our favorite approaches to mapping systems and how they interact with users.

"AI cannot fix a broken process. It maps the chaos faster."

Last week I spoke at Rosenverse Live about something we are still in the middle of: a Human-Agent pilot running at university with industry partners from tech, government, and Mittelstand. Not a finished case study, but a live process with real friction and real learning. These four patterns show up...