In this era of AI, we’ve been taught that LLMs are probabilistic, not deterministic, and that they will sometimes hallucinate. There’s a saying in AI circles that humans are right about 80% of the time, and so are AIs. Except when less than 100% accuracy is unacceptable. Accountants need to be 100% accurate, lest they lose track of money for their clients or businesses.

And that’s the problem Intuit had to solve to roll out their AI agent. Sean Michael Kerner, writing in VentureBeat:

Even when its accounting agent improved transaction categorization accuracy by 20 percentage points on average, they still received complaints about errors.

“The use cases that we’re trying to solve for customers include tax and finance; if you make a mistake in this world, you lose trust with customers in buckets and we only get it back in spoonfuls,” Joe Preston, Intuit’s VP of product and design, told VentureBeat.

So they built an agent that queries data from a multitude of sources and returns those exact results. But do users trust those results? It comes down to a design decision on being transparent:

Intuit has made explainability a core user experience across its AI agents. This goes beyond simply providing correct answers: It means showing users the reasoning behind automated decisions.

When Intuit’s accounting agent categorizes a transaction, it doesn’t just display the result; it shows the reasoning. This isn’t marketing copy about explainable AI, it’s actual UI displaying data points and logic.

“It’s about closing that trust loop and making sure customers understand the why,” Alastair Simpson, Intuit’s VP of design, told VentureBeat.

We’ve been hearing a lot about AI agents and now enough time has passed that we’re starting to see some learnings in industry. Writing in Harvard Business Review, Linda Mantia, Surojit Chatterjee and Vivian S. Lee showcase three case studies of enterprises that have deployed AI agents.

They write about Hitachi Digital and how they deployed an AI agent as the first responder to the 90,000 questions employees send to their HR team annually.

Every year, employees put over 90,000 questions about everything from travel policies and remote work to training and IT support to the company’s HR team of 120 human responders. Answering these queries can be difficult, in part because of Hitachi’s complex infrastructure of over 20 systems of record, including multiple disparate HR systems, various payroll providers, and different IT environments.

Their system, called “Skye,” is actually a system of agents, coordinating with one another and firing off queries depending on the intent and task.

For example, the intent classifier agent sends a simple policy question like “What are allowed expenses for traveling overseas?” or “Does this holiday count in paid time off?” to a file search and respond agent, which provides immediate answers by examining the right knowledge base given the employee’s position and organization. A document generation agent can create employee verification letters (which verify individuals’ employment status) in seconds, with an option for human approval. When an employee files a request for vacation, the leave management agent uses the appropriate HR management system based on its understanding of the user’s identity, completes the necessary forms, waits for the approval of the employee’s manager, and reports back to the employee.

The authors see three essential imperatives when designing and deploying AI agents into companies.

1. Design around outcomes and appoint accountable mission owners. Companies need to stop organizing around internal functions and start building teams around actual customer outcomes—which means putting someone in charge of the whole journey, not just pieces of it.
2. Unlock data silos and clarify the business logic. Your data doesn’t need to be perfect or centralized, but you do need to map out how work actually gets done so AI agents know where to find things and what decisions to make.
3. Develop the leaders and guardrails that intelligent systems require. You can’t just drop AI agents into your org and hope for the best—leaders need to understand how these systems work, build trust with their teams, and put real governance in place to keep things on track.

It’s interesting to me that Figma had to have a separate conference and set of announcement focused on design systems. In some sense it’s an indicator of how big and mature this part of design has become.

A few highlights from my point-of-view…

Slots seems to solve one of those small UX paper cuts—those niggly inconveniences that we just lived with. But this is a big deal. You’ll be able to add layers within component instances without breaking the connection to your design system. No more pre-building hidden list items or forcing designers to detach components. Pretty advanced stuff.

On the code front, they’re making Code Connect actually approachable with a new UI that connects directly to GitHub and uses AI to map components. The Figma MCP server is out of beta and now supports design system guidelines—meaning your agentic coding tools can actually respect your design standards. Can’t wait to try these.

For teams like mine that are using Make, you’ll be able to pull in design systems through two routes: Make kits (generate React and CSS from Figma libraries) or npm package imports (bring in your existing code components). This is the part where AI-assisted design doesn’t have to mean throwing pixelcraft out the window.

Design systems have always been about maintaining quality at scale. These updates are very welcomed.

As a follow-up to our previous item on Claude Code, here’s an article by Nick Babich who gives us three ways product designers can use Claude to code.

Remember that Anthropic’s Claude has been the leading LLM for coding for a while now.

With Cursor and Lovable as the darlings of AI coding tools, don’t sleep on Claude Code. Personally, I’ve been splitting my time between Claude Code and Cursor. While Claude Code’s primary persona is coders and tinkerers, it can be used for so much more.

Lenny Rachitsky calls it “the most underrated AI tool for non-technical people.”

The key is to forget that it’s called Claude Code and instead think of it as Claude Local or Claude Agent. It’s essentially a super-intelligent AI running locally, able to do stuff directly on your computer—from organizing your files and folders to enhancing image quality, brainstorming domain names, summarizing customer calls, creating Linear tickets, and, as you’ll see below, so much more.

Since it’s running locally, it can handle huge files, run much longer than the cloud-based Claude/ChatGPT/Gemini chatbots, and it’s fast and versatile. Claude Code is basically Claude with even more powers.

Rachitsky shares 50 of his “favorite and most creative ways non-technical people are using Claude Code in their work and life.”

Slow and steady wins the race, so they say. And in Waymo’s case, that’s true. Unlike the stereotypical Silicon Valley of “Move fast and break things,” Waymo has been very deliberate and intentional in developing its self-driving tech. In other words, they’re really trying to account for the unintended consequences.

Writing for The Atlantic, Saahil Desai:

Compared with its robotaxi competitors, “Waymo has moved the slowest and the most deliberately,” [Bryant Walker Smith] said—which may be a lesson for the world’s AI developers. The company was founded in 2009 as a secretive project inside of Google; a year later, it had logged 1,000 miles of autonomous rides in a tricked-out Prius. Close to a decade later, in 2018, Waymo officially launched its robotaxi service. Even now, when Waymos are inching their way into the mainstream, the company has been hypercautious. The company is limited to specific zones within the five cities it operates in (San Francisco, Phoenix, Los Angeles, Austin, and Atlanta). And only Waymo employees and “a growing number of guests” can ride them on the highway, Chris Bonelli, a Waymo spokesperson, told me. Although the company successfully completed rides on the highway years ago, higher speeds bring more risk for people and self-driving cars alike. What might look like a few grainy pixels to Waymo’s cameras one moment could be roadkill to swerve around the very next.

OK, so there’s workslop, but there’s also general AI slop. With OpenAI’s recent launch of the Sora app, there going to be more and more AI-generated image and video content making the rounds. I do believe that there’s a place for using AI to generate imagery. It can be done well (see Christian Haas’s “AI Jobs”). Or not.

Casey Newton, writing in his Platformer newsletter:

In Sora we find the entire debate over AI-generated media in miniature. On one hand, the content now widely derided as “slop” continually receives brickbats on social media, in blog posts and in YouTube comments. And on the other, some AI-generated material is generating millions of views — presumably not all from people who are hate-watching it.

As the content on the internet is increasingly AI-generated, platforms will need to balance how much of it they let in, lest the overall quality drops.

As Sarah Perez noted at TechCrunch, Pinterest has come under fire from its user base all year for a perceived decline in quality of the service as the percentage of slop there increases. Many people use the service to find real objects they can buy and use; the more that those objects are replaced with AI fantasies, the worse Pinterest becomes for them.

Like most platforms, Pinterest sees little value in banning slop altogether. After all, some people enjoy looking at fantastical AI creations. At the same time, its success depends in some part on creators believing that there is value in populating the site with authentic photos and videos. The more that Pinterest’s various surfaces are dominated by slop, the less motivated traditional creators may be to post there.

Sticking with the workslop or outsourcing our main work to AI, Douglas Rushkoff writes in Fast Company:

By using the AI to do the big stuff—by outsourcing our primary competencies to the machines instead of giving them the boring busywork—we deskill ourselves and deprive everyone of the opportunity for AI-enhanced outputs. Too many of us are using AI as the primary architect for a project, rather than the general contractor who supports the architect’s human vision.

People forget that it’s the process of doing something that helps us synthesize and form the connections necessary for innovation.

As the researcher behind MIT’s study “This is Your Brain on ChatGPT” explained at a recent ANDUS event, when people turn to an AI for a solution before working on a problem themselves, the number of connections formed in their brains decreases. But when they turn to the AI after working on the problem for a while, they end up with more neural connections than if they worked entirely alone.

That’s because the value of the AI is not its ability to create product for us, but to engage with us in our process. Working and iterating with an AI—doing what we could call generative thinking—is actually a break from Industrial Age thinking. We focus less on outputs than on cycles. Less on the volume of short-term results (assembly line), and more on the quality and complexity of thought and innovation.

Speaking of workslop, here’s an article from NN/g on how to avoid falling into over-reliance on AI in our design field. They call it the “7 Deadly AI Sins for UX Professionals.”

1. Outsourced Thinking
2. Wasted Time
3. Lost Details
4. Isolated Ideation
5. Naïve Trust
6. Bland Taste
7. Defensive Outlook

As Tanner Kohler writes:

It’s not about avoiding AI. It’s about maintaining your own growth and the quality of your work as you use AI. AI will constantly be changing. Never let yourself slip into repeatedly committing the sins that weaken you and your UX skills.

Definitely use AI at work if you can. You’d be guilty of professional negligence if you don’t. But, you must not blindly take output from ChatGPT, Claude, or Gemini and use it as-is. You have to check it, verify that it’s free from hallucinations, and applicable to the task at hand. Otherwise, you’ll generate “workslop.”

Kate Niederhoffer, Gabriella Rosen Kellerman, et. al., in Harvard Business Review, report on a study by Stanford Social Media Lab and BetterUp Labs. They write, “Employees are using AI tools to create low-effort, passable looking work that ends up creating more work for their coworkers.”

Here’s how this happens. As AI tools become more accessible, workers are increasingly able to quickly produce polished output: well-formatted slides, long, structured reports, seemingly articulate summaries of academic papers by non-experts, and usable code. But while some employees are using this ability to polish good work, others use it to create content that is actually unhelpful, incomplete, or missing crucial context about the project at hand. The insidious effect of workslop is that it shifts the burden of the work downstream, requiring the receiver to interpret, correct, or redo the work. In other words, it transfers the effort from creator to receiver.

Don’t be like this. Use it to do better work, not to turn in mediocre work.

Workslop may feel effortless to create but exacts a toll on the organization. What a sender perceives as a loophole becomes a hole the recipient needs to dig out of. Leaders will do best to model thoughtful AI use that has purpose and intention. Set clear guardrails for your teams around norms and acceptable use. Frame AI as a collaborative tool, not a shortcut. Embody a pilot mindset, with high agency and optimism, using AI to accelerate specific outcomes with specific usage. And uphold the same standards of excellence for work done by bionic human-AI duos as by humans alone.

The web is a magical place. It started out as a way to link documents like research papers across the internet, but has evolved into the representation of the internet and the place where we get information and get things done. Writer Will Leitch on Medium:

It is difficult to describe, to a younger person or, really, anyone who wasn’t there, what the emergence of the Internet — this thing that had not been there your entire life, that you had no idea existed, that was suddenly just everywhere — meant to someone who wanted to write. When I graduated college in 1997, the expectation for me, and most wanna-be writers, was that we had two options: Start on the bottom rung of a print publication and toil away for years, hoping that enough people with jobs above you would retire or die in time for you to get a real byline by the time you were 40, or write a brilliant novel or memoir that turned you into Dave Eggers or Elizabeth Wurtzel. That was pretty much it! Then, suddenly, from the sky, there was this place where you could:

• Write whatever you wanted.
• Write as long as you wanted.
• Have your work available to read by anyone, anywhere on the entire freaking planet.

This was — and still is — magical.

The core argument of what Leitch write is that while the business and traffic models that fueled web publishing are collapsing—due to changing priorities of platforms like Google and the dominance of video on social media (i.e., TikTok and Reels), the essential, original magic of publishing on the web isn’t dead.

But that does not mean that Web publishing — that writing on the Internet, the pure pleasure of putting something out in the world and having it be yours, of discovering other people who are doing the same thing — itself is somehow dead, or any less magical than it was in the first place. Because it is magical. It still is. It always was.

Noah Davis writing in Web Designer Depot, says aloud what I’d thought—but never wrote down—before AI, templates started to kill creativity in web design.

If you’re wondering why the web feels dead, lifeless, or like you’re stuck in a scrolling Groundhog Day of “hero image, tagline, three icons, CTA,” it’s not because AI hallucinated its way into the design department.

It’s because we templatified creativity into submission!

We used to design websites like we were crafting digital homes—custom woodwork, strange hallways, surprise color choices, even weird sound effects if you dared. Each one had quirks. A personality. A soul.

When I was coming up as a designer in the late 1990s and early 2000s, one of my favorite projects was designing Pixar.com. The animation studio’s soul—and by extension the soul I’d imbue into the website—was story. The way this manifest was a linear approach to the site, similar to a slideshow, to tell the story of each of their films.

And as the web design industry grew, and everyone needed and wanted a website, from Fortune 500s to the local barber shop, access to well-designed websites was made possible via templates.

Let’s be real: clients aren’t asking for design anymore. They’re asking for “a site like this.” You know the one. It looks clean. It has animations. It scrolls smoothly. It’s “modern.” Which, in 2025, is just a euphemism for “I want what everyone else has so I don’t have to think.”

Templates didn’t just streamline web development. They rewired what people expect a website to be.

Why hire a designer when you can drop your brand colors into a no-code template, plug in some Lottie files, and call it a day? The end result isn’t bad. It’s worse than bad. It’s forgettable.

Davis ends his rant with a call to action: “If you want design to live, stop feeding the template machine. Build weird stuff. Ugly stuff. Confusing stuff. Human stuff.”

Designer Ben Holliday writes a wonderful deep dive into how caring is good design. In it, he references the conversation that Jony Ive had with Patrick Collison a few months ago. (It’s worth watching in its entirety if you haven’t already.)

Watching the interview back, I was struck by how he spoke about applying care to design, describing how:

“…everyone has the ability to sense the care in designed things because we can all recognise carelessness.”

Talking about the history of industrial design at Apple, Ive speaks about the care that went into the design of every product. That included the care that went into packaging – specifically things that might seem as inconsequential as how a cable was wrapped and then unpackaged. In reality, the type of small interactions that millions of people experienced when unboxing the latest iPhone. These are details that people wouldn’t see as such, but Ive and team believed that they would sense care when they had been carefully considered and designed.

This approach has always been a part of Jony Ive’s design philosophy, or the principles applied by his creative teams at Apple. I looked back and found an earlier 2015 interview and notes I’d made where he says how he believes that the majority of our manufactured environment is characterised by carelessness. But then, how, at Apple, they wanted people to sense care in their products.

The attention to detail and the focus and attention we can all bring to design is care. It’s important.

Holliday’s career has been focused in government, public sector, and non-profit environments. In other words, he thinks a lot about how design can impact people’s lives at massive scale.

In the past few months, I’ve been drawn to the word ‘careless’ when thinking about the challenges faced by our public services and society. This is especially the case with the framing around the impact of technology in our lives, and increasingly the big bets being made around AI to drive efficiency and productivity.

The word careless can be defined as the failure to give sufficient attention to avoiding harm or errors. Put simply, carelessness can be described as ‘negligence’.

Later, he cites Facebook/Meta’s carelessness when they “used data to target young people when at their most vulnerable,” specifically, body confidence.

Ian Dean, writing for Creative Bloq, revisits the impact the original TRON movie had on visual effects and the design industry. The film was not nominated for an Oscar for visual effects as the Academy’s members claimed that “using computers was ‘cheating.’” Little did they know it was only the beginning of a revolution.

More than four decades later, TRON still feels like a moment the film industry stopped and changed direction, just as it had done years earlier when Oz was colourised and Mary Poppins danced with animated animals.

Dean asks, now what about AI-powered visual effects? Runway and Sora are only the beginning.

It’s always interesting to hear how others think about the design process from the outside. Eli Woolery and Aaron Walter interview creativity researcher and author Keith Sawyer to learn about what he’s found to be true after interviewing hundreds of art and design professors and students over a decade for his new book:

The creativity doesn’t come at the beginning. You don’t start by having a brilliant insight. You just dive into the process. And then as you’re engaging in the process, the ideas emerge.

Sawyer emphasizes that art and design schools are not just teaching students how to create, but how to “see.” He found that many professors believe students already possess creativity, but the role of art and design school is to help them realize and develop that potential by teaching them to observe, critique, and reflect more deeply on their own work.

When I interviewed these artists and designers, I would say, how are you teaching students how to create? And everyone was quite uncomfortable with that question. A lot of them would say, we’re not teaching students how to create. Or they’ll say something like, the students are already creative. We’re teaching them how to realize the potential they have as creatives.

Sawyer notes that the hardest thing for students to learn is how to see their own work—that is, to understand what they have actually made rather than sticking rigidly to their original idea.

When we talk about learning to see, you’re talking about learning to see yourself. The hardest thing to teach a student is how to see their own work, to see something that they’ve just generated. Because these studio classes, students have opportunities to share their work in interim stages along the way. You don’t go off and work for two weeks or four weeks and then bring back in the finished product. You bring in your interim and you get a lot of feedback and comments on it.

And what the professors tell me is these 18, 19, and 20-year-olds, they don’t realize what they put on the canvas. Or if they’re a graphic designer, they don’t realize what it is that they’ve generated. A lot of times, they’ll think they’ve done a certain thing. So they have this kind of linear approach—model of the creative process where I’m going to have an idea and I’m going to execute it so they’ll start with their idea and they’ll execute it. They’ll think that what they put on the canvas is their original idea, but in a lot of cases, it’s not. They can’t see what they’ve done themselves, so that’s kind of powerful how do you teach someone that what you put on the canvas isn’t what you say you’re doing.

You can’t just tell them, “Hey, you’re wrong. Let me tell you what you’ve done.” You have to lead someone through that. You have to walk them through it.

One way you do it is you put students in the classroom together and then have them comment on other students’ work so they will be on the other side. And they’ll see another student. talking about what they’ve done and not really describing what’s really on the canvas.

So I think that’s the hardest thing about learning to see is learning to see yourself, learning to see your own work.

I think that’s the power of art and design school, this studio learning environment. I’m biased, of course, because that’s how I learned. Those who are self-taught or have gone through bootcamps miss out on a lot of this experience. The other thing the design school environment teaches is how to give and take critiques. It’s about the work, not you.

In the scenario “AI 2027,” the authors argue that by October 2027—exactly two years from now—we will be at an inflection point. Race to build the superintelligence, or slow down the pace to fix misalignment issues first.

In a piece by Derek Thompson in The Argument, he takes a different predicted AI doomsday date—18 months—and argues:

The problem of the next 18 months isn’t AI disemploying all workers, or students losing competition after competition to nonhuman agents. The problem is whether we will degrade our own capabilities in the presence of new machines. We are so fixated on how technology will outskill us that we miss the many ways that we can deskill ourselves.

Degrading our own capabilities includes writing:

The demise of writing matters because writing is not a second thing that happens after thinking. The act of writing is an act of thinking. This is as true for professionals as it is for students. In “Writing is thinking,” an editorial in Nature, the authors argued that “outsourcing the entire writing process to LLMs” deprives scientists of the important work of understanding what they’ve discovered and why it matters.

…

The decline of writing and reading matters because writing and reading are the twin pillars of deep thinking, according to Cal Newport, a computer science professor and the author of several bestselling books, including Deep Work. The modern economy prizes the sort of symbolic logic and systems thinking for which deep reading and writing are the best practice.

More depressing trends to add to the list.

Our profession is changing rapidly. I’ve been covering that here for nearly a year now. Lots of posts come across my desk that say similar things. Tom Scott repeats a lot of what’s been said, but I’ll pull out a couple nuggets that caught my eye.

He declares that “Hands-on is the new default.” Quoting Vitor Amaral, a designer at Intercom:

Being craft-focused means staying hands-on, regardless of specialty or seniority. This won’t be a niche role, it will be an expectation for everyone, from individual contributors to VPs. The value lies in deeply understanding how things actually work, and that comes from direct involvement in the work.

As AI speeds up execution, the craft itself will become easier, but what will matter most is the critical judgment to craft the right thing, move fast, and push the boundaries of quality.

For those looking for work, Scott says, “You NEED to change how you find a job.” Quoting Felix Haas, investor and designer at Lovable:

Start building a real product and get a feeling for it what it means pushing something out in the market

Learn to use AI to prototype interactively → even at a basic level

Get comfortable with AI tools early → they’ll be your co-designer / sparring partner

Focus on solving real problems, not just making things look good (Which was a problem for very long in the design space)

Scott also says that “Design roles are merging,” and Ridd from Dive Club illustrates the point:

We are seeing a collapse of design’s monopoly on ideation where designers no longer “own” the early idea stage. PMs, engineers, and others are now prototyping directly with new tools.

If designers move too slow, others will fill the gap. The line between PM, engineer, and designer is thinner than ever. Anyone tool-savvy can spin up prototypes — which raises the bar for designers.

Impact comes from working prototypes, not just facilitation. Leading brainstorms or “owning process” isn’t enough. Real influence comes from putting tangible prototypes in front of the team and aligning everyone around them.

Design is still best positioned — but not guaranteed

Designers could lead this shift, but only if they step up. Ownership of ideation is earned, not assumed.

Is the AI bubble about to burst? Apparently, AI prompt-to-code tools like Lovable and v0 have peaked and are on their way down.

Alistair Barr writing for Business Insider:

The drop-off raises tough questions for startups that flaunted exponential annual recurring revenue growth just months ago. Analysts wrote that much of that revenue comes from month-to-month subscribers who may churn as quickly as they signed up, putting the durability of those flashy numbers in doubt.

Barr interviewed Eric Simons, CEO of Bolt who said:

“This is the problem across all these companies right now. The churn rate for everyone is really high,” Simons said. “You have to build a retentive business.”

I love this framing by Patrizia Bertini:

Let me offer a different provocation: AI is not coming for your job. It is coming for your tasks. And if you cannot distinguish between the two, then yes — you should be worried. Going further, she distinguishes between output and outcome: Output is what a process produces. Code. Copy. Designs. Legal briefs. Medical recommendations. Outputs are the tangible results of a system executing its programmed or prescribed function — the direct product of following steps, rules, or algorithms. The term emerged in the industrial era, literally describing the quantity of coal or iron a mine could extract in a given period. Output depends entirely on the efficiency and capability of the process that generates it.

Outcome is what happens when that output meets reality. An outcome requires context, interpretation, application, and crucially — intentionality. Outcomes demand understanding not just what was produced, but why it matters, who it affects, and what consequences ripple from it. Where outputs measure productivity, outcomes measure impact. They are the ultimate change or consequence that results from applying an output with purpose and judgment.

She argues that, “AI can generate outputs. It cannot, however, create outcomes.”

This reminds me of a recent thread by engineer Marc Love:

It’s insane just how much how I work has changed in the last 18 months.

I almost never hand write code anymore except when giving examples during planning conversations with LLMs.

I build multiple full features per day , each of which would’ve taken me a week or more to hand write. Building full drafts and discarding them is basically free.

Well over half of my day is spent ideating, doing systems design, and deciding what and what not to build.

It’s still conceptually the same job, but if i list out the specific things i do in a day versus 18 months ago, it’s almost completely different.

Care about the outcome, not the output.

In an announcement to users this morning, Visual Electric said they were being acquired by Perplexity—or more accurately, the team that makes Visual Electric will be hired by Perplexity. The service will shut down in the next 90 days.

Today we’re sharing the next step in Visual Electric’s journey: we’ve been acquired by Perplexity. This is a milestone that marks both an exciting opportunity for our team and some big changes for our product.

Over the next 90 days we’ll be sunsetting Visual Electric, and our team will be forming a new Agent Experiences group at Perplexity.

While we’ve seen acquihires and shutdowns in either the AI infrastructure space (e.g., Scale AI) or coding space (e.g., Windsurf), I don’t believe we’ve seen one in the image or video gen AI space have an exit event like this yet. Obviously, The Browser Company announced their acquisition by Atlassian last month.

I believe building gen AI tools at this moment is incredibly competitive. I think it takes an even stronger stomached entrepreneur than in the pre-ChatGPT moment. So kudos for the folks at Visual Electric for having a good outcome and getting to continue to do their work at Perplexity. But I do think this is not the last that we’ll see consolidation in this space.

Tim Berners-Lee, the father of the web who gave away the technology for free, says that we are at an inflection point with data privacy and AI. But before he makes that point, he reminds us that we are the product:

Today, I look at my invention and I am forced to ask: is the web still free today? No, not all of it. We see a handful of large platforms harvesting users’ private data to share with commercial brokers or even repressive governments. We see ubiquitous algorithms that are addictive by design and damaging to our teenagers’ mental health. Trading personal data for use certainly does not fit with my vision for a free web.

On many platforms, we are no longer the customers, but instead have become the product. Our data, even if anonymised, is sold on to actors we never intended it to reach, who can then target us with content and advertising. This includes deliberately harmful content that leads to real-world violence, spreads misinformation, wreaks havoc on our psychological wellbeing and seeks to undermine social cohesion.

And about that fork in the road with AI:

In 2017, I wrote a thought experiment about an AI that works for you. I called it Charlie. Charlie works for you like your doctor or your lawyer, bound by law, regulation and codes of conduct. Why can’t the same frameworks be adopted for AI? We have learned from social media that power rests with the monopolies who control and harvest personal data. We can’t let the same thing happen with AI.

I’m happy that the conversation around the design talent crisis continues. Carly Ayres, writing for It’s Nice That picks up the torch and speaks to designers and educators about this topic. What struck me—and I think what adds to the dialogue—is the notion of the belief gap. Ayres spoke with Naheel Jawaid, founder of Silicon Valley School of Design, about it:

“A big part of what I do is just being a coach, helping someone see their potential when they don’t see it yet,” Naheel says. “I’ve had people tell me later that a single conversation changed how they saw themselves.”

In the past, belief capital came from senior designers taking juniors under their wing. Today, those same seniors are managing instability of their own. “It’s a bit of a ‘dog eat dog world’-type vibe,” Naheel says. “It’s really hard to get mentorship right now.”

The whole piece is great. Tighter than my sprawling three-parter. I do think there’s a piece missing though. While Ayres highlights the issue and offers suggestions from designer leaders, businesses need to step up and do something about the issue—i.e., hire more juniors. Us recognizing it is the first step.

Jason Spielman put up a case study on his site for his work on Google’s NotebookLM:

The mental model of NotebookLM was built around the creation journey: starting with inputs, moving through conversation, and ending with outputs. Users bring in their sources (documents, notes, references), then interact with them through chat by asking questions, clarifying, and synthesizing before transforming those insights into structured outputs like notes, study guides, and Audio Overviews.

And yes, he includes a sketch he did on the back of a napkin.

I’ve always wondered about the UX of NotebookLM. It’s not typical and, if I’m being honest, not exactly super intuitive. But after a while, it does make sense. Maybe I’m the outlier though, because Spielman’s grandmother found it easy. In an interview last year on Sequoia Capital’s Training Data, he recalls:

I actually do think part of the explosion of audio overviews was the fact it was a simple one click experience. I was on the phone with my grandma trying to explain her how to use it and it actually didn’t take any explanation. I’m like, “Drop in a source.” And she’s like, “Oh! I see. I click this button to generate it.” And I think that the ease of creation is really actually what catalyzed so much explosion. So I think when we think about adding these knobs [for customization] I think we want to do it in a way that’s very intentional.