Why AI Implementations Fail and How to Fix It | Unfinishe_ posted on the topic | LinkedIn Small Bites, Big Impact: Why AI Implementations Fail (And How to Fix It) We walked into an architecture firm with compelling numbers: significant annual administrative costs that AI could automate. They...

A really good high level description of how small AI projects can augment workflows.

People who use AI the most return, again and again, to the importance of deeply understanding what the work is in the first place.

Narrower Terms vs. Alternative Terms When creating or editing a taxonomy, it's important to know when to create narrower terms vs. synonym terms (alternative labels).

Great thoughts from Heather on #taxonomytuesday about the problem with using the word "term" when building taxonomies.

There is a key idea here there is the NAMING of a thing (labels) and then there is the STRUCTURE of the thing (concepts). They are distinct.

Great Stuff!

Narrating diagram on cognition and user research | Indi Young posted on the topic | LinkedIn Here's a narrated version of the diagram that shows where/when to make durable knowledge about people (cognition) versus the cyclical #UserResearch that comes in the team's build space. 📣 This is part...

A really great discussion of how to consider durable knowledge in technical projects by Indi Young.

We're glad designers are having this conversation more often, because it will make software projects better in general.

buff.ly/iSJZBUF

Archetypes of User Intention — The Understanding Group (TUG)

Archetype modeling ensures that the information structure at the core of a digital place is optimized to help users get things done. Learn to anticipate user needs in your digital places!

Starting November 13th.

More info: buff.ly/tsAjNar
#DesignNurture #TUGWorkshop

Vibe coding has turned senior devs into ‘AI babysitters,’ but they say it’s worth it | TechCrunch TechCrunch spoke to experienced coders about their time using AI-generated code about what they see as the future of vibe coding.

The key thing to remember about "vibe" anything is that LLMs are not procedural turing machines. They are naive, chipper magpies trying to find the things that will make you happy, for better or worse. Working with them in that mindset is critical.

Archetype modeling ensures that the information structure at the core of a digital place is optimized to help users get things done. Learn to anticipate user needs for your digital places. Starting November 13th. More info: buff.ly/tsAjNar
#DesignNurture #TUGWorkshop

Democracy in the Age of AI Democracy in the Age of AI

Taiwan has been at the forefront of using social media tools to strengthen their democracy, so much so that Audrey Tang was interviewed on the "Your Undivided Attention" podcast. This article is practical and hopeful in its ideas for using AI to improve democratic outcomes as well.

Why AI is conservative and how it affects tech choices | Jorge Arango posted on the topic | LinkedIn AI is inherently conservative. I don’t mean that in the political sense — this is lowercase-c “conservative.” Because of how they’re trained, LLMs favor and perpetuate long-established frameworks...

We're not sure that what Jorge is describing is "conservative" as much as "average to its very bones", but his key point stands: AI is not good for finding less well lit corners or surprising results. It literally can't, and we have to keep that in consideration as we use it.

buff.ly/V5p5FFv

Archetypes of User Intention - Remote Workshop — The Understanding Group (TUG) Learn TUG’s method for understanding the intentions of people as a way to make better digital places. Workshop Dates: Thursday, November 13, 2025, 12 – 3 p.m. ET Thursday, November 20, 2025, 12…

Archetype modeling ensures that the information structure at the core of a digital place is optimized to help users get things done. Learn to anticipate user needs in your digital places!

Starting November 13th.

More info: buff.ly/tsAjNar
#DesignNurture #TUGWorkshop

Design as governance: The next frontier for enterprises | Mark P. posted on the topic | LinkedIn The next frontier for enterprises is not digital innovation. It is governance of the digital, AI included. Design is the connective tissue here: mapping the terrain, orchestrating accountability, embedding...

Another post that reflects the current state of AI, which is basically "how do we MANAGE them". A reflection of the way that they generate outputs that require more work on the back end and in the original setting of constraints than creating the actual outputs themselves.

buff.ly/OYqoDqO

Enterprise Architecture of Agency: A New Unit of Design | Jesper Lowgren posted on the topic | LinkedIn Enterprise architects don’t live in diagrams; we live in the 𝐭𝐞𝐧𝐬𝐢𝐨𝐧𝐬 𝐛𝐞𝐭𝐰𝐞𝐞𝐧 𝐝𝐨𝐦𝐚𝐢𝐧𝐬. Business leaders want speed and new value. Data...

This short post may be a seminal work in the creation of agency-centered architectures. There are some assertions about the scope of agents here that may be challenged as they expand in capability and quality, but the overall vision of the model is deeply compelling.

buff.ly/6FgYUOw

Documents: The architect’s programming language - Stack Overflow Senior developers know how to deploy code to systems made of code. Architects know how to deploy ideas to systems made of people.

Systems documentation as an Information Architecture problem! It's really notable how many decisions map to IA issues, starting with simple things like the way that the author organizes his documentation (chronologically rather than by topic).

Mother of All Art and Art History Links Pages This page has been accessed times since August 13, 1996

This old site shows how many essential elements of architecture remain consistent, even in layouts approaching thirty years old. The only 90's constraint is that the page was designed for smaller screens.

It even looks good on a smartphone (if you hold it in landscape mode)!

LukeW | Common AI Product Issues At this point, almost every software domain has launched or explored AI features. Despite the wide range of use cases, most of these implementations have been t...

A great summary of the issues with the state of the art AI design pattern (and there is only one right now). Good stuff!

She started invoicing her company for data requests. | Justin Custer She started invoicing her company for data requests. $200 per PowerPoint. $500 per dashboard. What happened next: It began as a joke during her performance review. "You say I'm not strategic enough,"…

A remarkable meditation on the power of surfacing cost and value, regardless of your domain, job description, or role. A must-read.

buff.ly/9GkcA0y

I was thinking about this over the weekend - just how much extra effort it now takes to simply buy things. | Luke Reed I was thinking about this over the weekend - just how much extra effort it now takes to simply buy things. We’re told technology makes life easier. Faster. Smarter. More efficient. But has it really?…

The EXPERIENCE of using kiosks, online tools, etc. adds a level of cognitive friction that makes them feel burdensome, even though they provide tremendous value. So, is there a way to improve the experience so people don't feel so exploited?

buff.ly/oTgljkQ

One of the most powerful things TUG does is to create ways to generate "consent" in the understanding and apprehension of data.

buff.ly/dJOPsf5

Why we changed the format of the service blueprint Service blueprinting is a core method of service design. However, the traditional format of a service blueprint undermines the value of the…

Nonlinear service modeling problems have existed for quite a while, actually, but #AI makes the initial decision points even more stochastic and wooly. So how do we manage our planning and design? This article points to a potential path.

buff.ly/qGCOQkA

Using Claude Code for Information Architecture Repurposing a software development agent to help draft website taxonomies.

A lot of what #AI generates for #IA projects is so bone-dry that it's hardly worth using. Jorge's idea of using an AI add-on originally designed to architect unique code bases to generate new taxonomies was a brilliant approach to solving this problem.

LukeW | Agent Management Interface Patterns As an increasing number of AI applications evolve to agents doing work for people, agent management becomes a critical part of these product's design. How can p...

This article's patterns for running #AIagents are all project management tools (kanban, email, etc.)

By this reasoning, you could swap out "agent" throughout the article for "talented but inexperienced intern" and the "interfaces" would be the same.

Over our many years as design consultants, the limiting factor has (almost) never been the rate at which we can generate artifacts, but the pace at which clients can make decisions, which is a… | Erika Hall Over our many years as design consultants, the limiting factor has (almost) never been the rate at which we can generate artifacts, but the pace at which clients can make decisions, which is a…

A critical insight from Erika Hall.

We manage what we measure. "good" has become synonymous with "fast" because we measure fast now (we even call them "sprints"). But what if we want "good" to be synonymous with something else?

Here’s a wacky theory I can’t shake: I’m pretty sure many of the people steering tech are on the spectrum (and I’m pretty sure I am, too). | Christina Wodtke Here’s a wacky theory I can’t shake: I’m pretty sure many of the people steering tech are on the spectrum (and I’m pretty sure I am, too). Live human interaction drains our batteries, so we invent…

Has #HomoLogicus become Homo Introvertus? And, more specifically, is it really "us"? And even if it is, is it good for our digital systems to be designed as if we were all painful introverts?

Thoughtful ideas from Christina Wodtke

buff.ly/CjpkxW6

Beyond prompt engineering, AI is really about *context engineering.* Once you hear this, it may sound obvious. | Marshall Kirkpatrick Beyond prompt engineering, AI is really about *context engineering.* Once you hear this, it may sound obvious. But it's got very big implications. This idea rippled through top AI conversations on…

Every experienced UX professional already has the core cognitive skillset to do almost all of what the AI community is hailing as the cutting edge of practice. -- Jesse James Garrett

True? Untrue? Incomplete? How does the community feel about becoming an army of prompt engineers?

Some research practices age like wine. | Nikki Anderson Some research practices age like wine. Others age like milk left in a warm usability lab. Here are a few we should resurrect and a few we should let quietly rot. 𝗕𝗿𝗶𝗻𝗴 𝗶𝘁 𝗯𝗮𝗰𝗸: ↳ Paper…

We agree completely with list of "bring it back" from this article; in fact we're wondering why and if they ever left!

Everybody is building the same thing. | Luke Wroblewski Everybody is building the same thing. Meanwhile Bench Computing, Inc. • look at the 6 screenshots in this image • find the CEOs of each company • make an image of the Spider-Man pointing meme but…

A lot of this is because the underlying technology simply demands it. Current LLMs produce undifferentiated text blocks to answer prompts. Is there any other way that the interface should be expected to look?

buff.ly/OnkkLJN

You aren’t imagining it. | Jesse James Garrett You aren’t imagining it. Design leaders really have been losing ground in authority and influence in recent years, and industry analyst John Gleason has the data to prove it, as he shared with me and…

"Design leaders who do not define the value their teams create will continue to have that value defined for them by others."

"How do I help" is one of the most important questions any professional or organization can ask to be effective.

buff.ly/jp5mTBW

Websites are going to change completely in the next two years. | Drew Burdick Websites are going to change completely in the next two years. Not just a redesign. A full shift in how they work. I've had half a dozen conversations in the past few months with leaders planning to…

What are the design and IA implications of this prediction? What are the informational requirements needed to make it work?

buff.ly/2KKILPU

The Cost of "More with Less": A Reflection on Platform Scale and Hidden Complexity By: Jamil Jadallah Since I began working on platform teams, helping developers onboard, building internal tooling, and scaling infrastructure, one thing has become increasingly clear… The complexity…

The way TUG describes this is "make the complex" clear. The essence is the same: the world is complex, and that's okay. It's our job to work through that complexity in a way that works for the systems that the complexity contains.

“We tend to overestimate the effect of a technology in the short run and underestimate the effect in the long run.” -Amara's Law

h/t Scott Berkun, who also notes that long-term effects of technology are often not the anticipated ones when the tech is first developed.

A Vision For Product Teams - Silicon Valley Product Group Marty Cagan discusses the potential impact of generative AI on the roles on a product team and the team topologies of product organizations.

One of the better thought out explorations of the impact of #LLM and #AI on digital creation. Essentially, #AI will be largely responsible for the PRODUCTION of things, leaving the product team to focus on DISCOVERY of new ideas and concepts.