Human-Centered, Robot-Driven: Ethical Considerations for ML in Design

Cyborg face, generated by Midjourney

UPDATE 2.6.23—I’ve decided to correctly refer to these systems using their actual technologies (i.e. ML—Machine Learning) rather than the market-speak and false narrative of AI (Artificial Intelligence). The article has been updated accordingly.

Modern ML systems like ChatGPT and Midjourney changing how we design, and how we think about design today and in the future. This comes at a cost. If we’re not willing to play this game right we have no business playing it at all. Below I discuss some of the tools used today, some of the issues we’ve seen with these systems, and how we can work together to have our ML-generated cake and consume it too.

At the end of the article, I’ve listed just a few of these systems already in play today.

What are these ML thingamabobs anyway?

Thingamabob, generated by Midjourney

In a nutshell, systems like ChatGPT and Midjourney can generate human-like output (e.g. text and images respectively). Their models are trained on large amounts of existing data and can generate new, derivative content based on said data.

While these models have the potential to enhance the design process, they also raise several ethical, moral, and practical issues.

Bad actors

A shady character, generated by Midjourney

One of the main concerns with ML-generated content is the potential for the proliferation of propaganda and misleading information. These systems can also be used to impersonate real people and organizations, not only spreading falsehoods but also violating privacy with the potential to cause real harm. Additionally, it raises questions about the authenticity of any and all content, and the ability to trace it back to its original author.

Bias

Unbalanced, generated by Midjourney

Additionally, ML-generated content can perpetuate bias and discrimination. ML models are only as unbiased as the data they are trained on, and if the data is biased, the generated designs will also be biased. This could lead to designs that exclude or marginalize certain groups of people, which is a major ethical concern. A good example of this is using ML to recommend people for a promotion to management. Sounds brilliant right? Remove any notion of gender, race, physical attributes, or ability…perfectly objective yeah? Amazon discovered this isn’t so. Their system was trained on résumés of the people who’d historically done well in those positions in the large tech company. Care to guess which gender dominated at the tech giant? So even without the system being aware of gender, the bias was built into the training models simply because of how things worked in the past. Bias has found its way into our judicial system, mortgage lending, resident screening, and more affecting real people’s lives—and generally not for the better.

Theft

Man in black stealing art, generated by Midjourney

Some artists have gone so far as to launch a lawsuit against some of the ML generators like Stability AI and Midjourney. They claim their rights were infringed upon by having their work (and likely millions of others) scraped illegally by these companies to train their models. 

Without the billions of pieces of work generated by these artists, the models would not be able to function and there would be no tool, but not a single artist has been attributed let alone compensated for the billions of hours taken to produce the original work.

Recognizing the problems

Weapons of Math Destruction, by Cathy O’Neil

In "Weapons of Math Destruction", Cathy O'Neil describes three checks to identify "Weapons of Math Destruction" (WMDs) which are any systems, technologies, or models used in harmful ways, often to marginalized groups. The three checks are:

1. Opacity: A WMD is opaque if it is hard to understand how it works, and if its creators are unwilling or unable to explain it. This makes it difficult for people to question, investigate, or validate the model's assumptions, biases, or decisions.

2. Scale: The more people the tool has access to, and the more it's used, (the bigger the scale) the bigger the risk of the tool harming large groups of people.

3. Damage: A WMD causes damage if it’s used to make decisions with a negative impact on people's lives and if the people who are most affected by the model are not the ones who are best positioned to understand or challenge it. This makes it difficult for people to fight back against the model's decisions or to change the model itself.

Technically Wrong, by Sara Wachter-Boettcher

Sara Wachter-Boettcher's "Technically Wrong" also highlights the importance of considering the ethical, moral, and societal implications of technology.

One of the key points they both make is the importance of using diverse and unbiased data sets when training ML models to reduce the potential for bias and discrimination in ML-generated designs. They both emphasize the importance of transparency and accountability when using ML, by being transparent about the technologies being used, and by monitoring and evaluating the ML-generated content to ensure they’re inclusive, unbiased, and not harmful.

O'Neil's WMD model highlights the importance of addressing the root causes of bias, such as the use of biased data sets, and the lack of diversity in tech companies. Wachter-Boettcher emphasizes the importance of designing technology with a human-centered approach, considering the potential consequences and impact on society when using technology, and the need to be transparent and accountable for the products they create.

What can we do?

Robot and Human Handshake”, generated by Midjourney

Much like the monstrosity that is the image above, humans and machines alike have a lot of work to do to ensure we don’t make a giant mess of everything. Several actions can be taken today when leveraging ML in our work to avoid ethical, moral, and practical problems.

If you’re a data scientist or leader:

• Use diverse and unbiased data sets: When training ML models, use diverse and unbiased data sets. This will help to reduce the potential for bias. This. Must. Be. Done.

• Consider the social and ethical implications: Consider any potential consequences and impact when building generative systems. Ask yourself how a bad actor might use this for nefarious purposes.

• Encourage ethical guidelines for ML usage: Work with industry groups and organizations to establish ethical guidelines for the use of ML.

If you’re a designer:

• Be transparent about the use of ML: Clearly disclose when ML is being used in your process and the specific ML technologies being used. This will help to promote transparency and accountability.

• Continuously monitor and evaluate generated content: Regularly monitor and evaluate the generated content to ensure they are inclusive, unbiased, and not harmful.

• Consult with experts: Seek advice from experts in ML ethics, privacy, and legal issues when implementing ML in design work.

• Invest in your education and professional development: Stay current on the latest developments and best practices in ML-based design to stay informed about the ethical and practical issues surrounding ML-based design.

Elevate artists and designers, don’t exploit them

Bots lift us up where we belong, generated by Midjourney

• Implement clear attribution and copyright policies: Clearly state how ML-generated content will be attributed and ensure that the original creator is credited for their work.

• Use ML to augment, not replace, human creativity: ML should be used to assist designers in the creative process, not replace them. This will ensure human creativity and artistic expression are still valued, and keep humans centric in the process.

• Educate artists and creators about ML: Educate artists and creators about the capabilities and limitations of ML so they can make informed decisions about how they want to use it in their work.

• Encourage collaboration between artists and ML experts: Encourage collaboration between artists and ML experts to ensure that ML is used in a way that supports and enhances the artist's vision.

• Encourage Fair Use and Open-source policies: Encourage the usage of open-source ML technologies to ensure accessibility and fairness. Transparency into the algorithms will help prevent them from being maliciously used.

• Protect intellectual property and provide compensation: Provide artists and creators with attribution and compensation for the use of their work in training models.

ML has the potential to enhance the design process, but it raises several ethical, moral, and practical issues. It’s paramount that everyone, designers, developers, leaders, and end-users alike, is aware of these issues and actively takes steps to mitigate them. This includes being transparent about how the ML models work, being accountable for the generated content, and being aware of and addressing bias in the data and generated content. Additionally, it's important to consider how ML-generated content may impact artists and creators and to work towards fair compensation and attribution for their work. By taking these steps, we can ensure ML is used responsibly and ethically, while still reaping the benefits of this powerful technology.

Sidenote: Not for nothing, the courts are literally still out on who exactly owns the output from generative systems. While OpenAI’s terms seem to indicate users own their output, the law is a lot more divided at the moment in terms of actual copyright.

And finally, it’s up to the technologists to take a bigger role in policing ourselves, and asking whether something should be done as often as we ask if it can be done, as well as how we do it.

Author’s note: This article was written with the assistance of ChatGPT and has a GPTZero score of 268.4: “text is likely human-generated”.

---

Some existing ML design and content systems:

GPT-3 (Generative Pre-trained Transformer 3): A language model developed by OpenAI, it can be used for NLP (natural language processing) tasks such as text generation, language translation, and language understanding.

Autodesk Dreamcatcher: A generative design tool that uses algorithms to generate design solutions based on design constraints and goals. It allows designers to explore a wide range of design possibilities, leading to more innovative and unique solutions.

Microsoft Sketch2Code: An ML-powered design tool that can turn a hand-drawn wireframe into a functional website. It uses ML (machine learning) to understand the design and automatically generate the corresponding code.

Midjourney: An ML-based generative design tool that can generate images and videos based on input like text prompts, or other images or videos. It's used to generate new and unique designs and art.

The Designer's Secret Weapon: How ML is Revolutionizing Web Design

Robot brains connected to screens - Generated with Midjourney

UPDATE 2.6.23—I’ve decided to correctly refer to these systems using their actual technologies (i.e. ML—Machine Learning) rather than the market-speak and false narrative of AI (Artificial Intelligence). The article has been updated accordingly.

As I write this ML (machine learning) is revolutionizing the way we approach UX (user experience) research and design. One of my biggest hopes for the technology is for it to help designers overcome the dearth of creativity in the modern web, creating new experiences that are unique, usable, and accessible.

One of the ways ML is doing this is through services like ChatGPT. ChatGPT is a large language model that can generate human-like text, which can be used to create more natural and engaging interactions between users and digital products. Imagine integrating it into existing sites to replace some intake forms. This can help humanize the web by making interactions with digital products feel more natural and personal, even conversational. Or instead of working to tweak your Google query to find the right website, just ask chat your question in plain language and get the data you're looking for directly.

Screenshot of a ChatGPT prompt to summarize this article

Another great way to utilize ML like ChatGPT is by integrating it into existing sites in the form of support, chat, and even forms. This can help humanize the web by making interactions with digital products feel more natural and personal. Imagine filling out a form that can understand and respond to your input in a conversational manner, making the process less tedious and more enjoyable. By using ML in this way, we can create a more seamless and enjoyable experience for users on the web, making them feel like they're interacting with a real person, rather than just a machine.

Here are just a few ways it's already happening today.

ML can assist designers in creating accessible experiences. ML-powered tools and plug-ins can analyze designs and identify accessibility issues, providing designers with instant, actionable feedback. This can help designers ensure their designs are inclusive and can be used by a diverse range of users.

Blind woman reading with ML-enhanced glasses - Generated with Midjourney

ML can nudge designers to expand their creativity by equipping them with tools and insights to inspire new ideas and approaches to design. ML could analyze user data and behavior, providing designers with valuable insights into how people interact with digital products more quickly.

ML-powered tools like Ando can inspire designers in generating new design solutions through the use of generative design techniques. These techniques use algorithms to generate multiple design options based on a set of design constraints and goals. This can help designers explore a wider range of design possibilities quickly leading to more innovative and unique solutions faster.

ML can work in the implementation of designs by automating repetitive tasks like alternate version generation, layout, and optimization. This can free up designers to focus on the more creative aspects of their work, leading to more efficient and effective design processes where designers spend more time on the things they love, and less on the things they don't.

Mobile app to purchase stock - Generated by Midjourney

ML can be a valuable tool for designers, providing them with new insights, inspiration, and automation capabilities that can help them create more unique, usable, and accessible experiences on the web. As designers, we should rejoice in the future as ML tools become more available (and usable) to create truly unique and usable experiences accessible to everyone.

If you have other suggestions of how ML can positively enhance modern design work drop me a line! As to the negative aspects, I'll need to address that in a follow-up post.

Author’s note: This article was written with the assistance of ChatGPT, with a GPTZero score of “81.4: text is likely human-generated”

Landing page for a window company - Generated on Midjourney by @ToggleWords

Can Great UX Be Novel and Risky Too?

In my previous post Delivering Quality Experiences, there’s an overarching sentiment that introducing novelty and unique behaviors in your UX should generally be frowned upon. I still believe that (within the context of that article, largely focused on enterprise, software, web apps, and the like).

I’ve come to feel that we, the UX community, have been sacrificing experiential pleasure and meaning at the altar of usability and convenience.

But when it comes to the web, and digital media at large, I’ve come to feel that we, the UX community, have been sacrificing experiential pleasure and meaning at the altar of usability and convenience. At least, more than necessary. We have done, and continue to do so, to the detriment of our craft and the continued evolution of our collective digital experiences.

This is not to say that pleasure and meaning can only be derived from a novel, bold, or risky design or experience. But, take a look at the web around you and tell me when last you landed on a page or opened an app and felt enthralled. Giddy? Surprised? Excited? I cannot.

In writing this, I’m not saying anything that hasn’t already been said before. But this is the first time I’m saying it publicly, and wanting to do something about it.

What’s to blame?

Templates
Design templates and libraries are easy, obvious targets. These alone literally make websites look and feel similar. Every time we reuse, repurpose, and revert to existing and common, we perpetuate the mundanity.

Photo: OozleMedia.com

Grids
Grids are another culprit, but we’ve been designing cool stuff for thousands of years using grids (tile, mosaics, mandalas, buildings, etc.), so I say that’s a cop-out and not worth mentioning further.

Screens
Mobile and responsive breakpoints are yet another suspect. To me though, these should be thought of as opportunities for surprise and elegant delight. There’s no reason my mobile experience should have to function similarly or the same as the desktop experience, in fact, they really shouldn’t. Time and money are at play here though, and no amount of argument will change that. So suffice it to say that properly leveraged, device accommodation should be a motivator for exceptional and amazing design, not an excuse to get lazy.

Us
Mostly dear reader, the real answer is us. The UX community.

UX Design—the cause of, and solution to, all of the web’s problems

Specifically, ignorant UX is the cause of all the problems. I don’t mean ignorant UX designers or that UX is inherently ignorant. I mean UX design (and research, and testing et al) that ignores the top of the UX Hierarchy of Needs: Pleasure and Meaning. This ignorance relegates the majority of our output to what Stephen Andersen referred to as the Zone of Mediocrity. Is that where we want our designs and experiences to live? Or, would you rather take a little risk, actually design something new, even challenging for a change, and propel your product into a category all its own?

To those who would argue accessibility prevents them from taking risks, I call shenanigans. Making something work for people with different abilities doesn’t force you into a box. In fact, understanding those with various challenges to traditional interfaces may be exactly the boost your creative mind needs to try something daring and new! We know that mobile-first design can help solve a number of UX/UI issues by helping us make hard choices about how we architect information and lay out interfaces. Accessible-first design is the logical next and better step. When something works for everyone, it will work for anyone! Try it and watch your usability success rates skyrocket.

Yes. Great UX can also be novel and risky.

So next time we run through our processes and procedures, our research and readouts, our methods and our madness, let’s inject something new, something risky, something bold and outlandish. It might not work, and that’s okay. Fail early, fail often. But dare to fail. Could there be danger in the water where we can’t see the bottom? Sure, but it’s also where treasure sits awaiting discovery! I think the greater danger lies in the static banality of a ubiquitous internet where everything looks and feels like a slice of the same bland quiche.

Here’s to the designers who go beyond what and why, and start asking what if and why not!

Delivering Quality Experiences

Photo by Michaela Baum on Unsplash

1. The Basic Design Process

The Brief

We've all seen the cliché memes even if we haven't experienced them ourselves:

Make the logo bigger! Make it pop! Add more sizzle! Make it fresh! Wow me!

And, honestly, we've also seen some sites that do just that, some even do it well. Everything from auto-playing music and sound effects, to parallax, animated gradients, interactive video, even websites that are almost full-blown video games in their own right.

It can feel as if every call with the client begins in a similar vein.

"We're building <APP> for <VERTICAL>. It's like <OTHER APP>, but ours will be different because <POP/SIZZLE/GAMIFICATION>."

We nod our heads. "It'll have dynamic, user-content-driven dashboards!" Yep. "And we want it to work on mobile and tablets." Absolutely. "It'd be cool if we could have some ML and AR involved there too, but we're not sure how yet. We're hoping you can help us figure that part out." Nothing new here. You want it to be accessible too right? “Oh yeah, accessibility! If we have time left, do that!” <sigh>

Research

You do your research. You talk to the stakeholders. You ask hundreds of questions of your client's customers. You unravel the knot of unknowns, and patterns of user behavior emerge. You begin to understand what the business actually needs, and what the users actually need. You rough out an IA and test it. You see the pitfalls in the current journey. You identify inefficiencies and problems, er, opportunities across the service layers. You present your findings. "Yes! You get it!" the client says. We're all on the same page. You understand the problem, and you begin solving it.

Photo by Leon on Unsplash

UX Design

You wireframe. You prototype. You assure the client that, no, these aren't the actual colors. No, that's not English it's called Greeked text. “Why does it look like Latin if it's called Greeked text? Why don't the numbers in those columns add up?” So you replace all the lorem ipsum with real-ish copy and adjust the numbers to make a little sense. Okay, now they get it. No, that's likely not the font you'll use, we're just trying to get a sense of how things work first. Visual design will come later. Okay, they get it. Sort of. So, you begin testing.

Photo by Amélie Mourichon on Unsplash

Testing

Testing goes great. You uncover some problems with the IA. No biggie, super-easy change. You see some small issues with some labels, so you wordsmith those. There are a couple of show-stoppers that make you feel stupid. How did you not see that before!? But you see it now, and you've already got 2 or 3 ideas for how to solve that. Things are smokin! You report your results. It tested through the roof, and you're very confident users will experience very little friction. They'll be able to easily grasp the tasks with no supervision. You've organized things with clear paths. Sequential work is laid out so it's obvious what came before, what they're doing now, and what comes next. The client begins to see how amazing this new vision is. So you move on to visuals.

Photo by UX Indonesia on Unsplash

Visual UI Design

The visual designers kill it. Maybe it's another team. Maybe it's you. Maybe it's Maybelline. They've transformed your usable, utilitarian, efficiency-driven low-fi wires into gorgeous, pixel-perfect renderings of the final vision.

And here's where things can get dicey.

Photo by Daniel Korpai on Unsplash

2. The Road To Destruction

Make it Pop!

"I like it..." says the HiPPO. "...but I don't love it." This is nothing new. We're designers, and this is the path we've chosen. Okay, that's fair. What don't you love about it?

"Oh, I don't know. It just doesn't pop/wow/sizzle/whatever. Have you been to <THAT HOT VIDEO GAMESITE>? They've got lasers, and real-time AR video that places you right in the middle of the screen and when you move your mouse around your little guy runs after it! And when you scroll, it scrolls horizontally, not vertically, which gives it a panoramic feeling! It's really cool!"

Everyone is forcing smiles and nodding...

"I know our accounting platform isn't a video game, but why can't we do something cool like make the avatars move around, or use their webcams to show a real-time view of their face so they don't have to upload a photo? What if when they scrolled, the data grid used that cool parallax motion so the columns moved underneath one another like those cool Miyazaki films? Anyway, all our competitors scroll vertically...what if we laid all our stuff out horizontally? It'd really set us apart!"

Okay - so I may be exaggerating some client requests and expectations, but honestly, not by very much.

The infamous LingsCars.com

Note: I don’t hate Lingscars.com for a number of reasons. Perhaps that’s another article.

Also, I’m not talking about getting the right feedback at the right time. That’s a different issue, nicely addressed in this article by James Cook.

3. The Path To Quality

The Truth

It’s times like these my boss Tim is wont to say

“Effective UX has to walk the line between fresh and novel, and usable. Novel is interesting initially, but usable has to be useful when the shimmer has worn off. We strive to achieve both of those valuable experiences.”

— Tim Doll

The first thing about introducing anything "fresh and novel" is that you've created your own usability debt. By its very nature, the user is being introduced to something they may not be familiar with and will have to discover and learn. If this is a game, or a marketing site, maybe that's a great thing. If it's accounting software, it's not. In fact, in most (if not all) cases I'd argue introducing your own obstacles to clarity and efficiency, not managing to your user base's existing mental models is bad. It’s like tying your own shoelaces together then trying to sprint.

Remember Stephen Anderson's hierarchy of UX?

Concept and design by Stephen P. Anderson

We assume the app will be built to function reliably. Our job is to make it first usable and convenient. Only when we've established that base platform can we even begin to explore pleasure and meaning. If you try to design for uniqueness and stand-out visuals prematurely, you'll compromise your own foundation that your research and design teams spent so much time and effort establishing.

Does this mean we don't try to cross the chasm of convenience and push our apps into the pleasure zone? No. In fact, apps like Word and Excel could really use a healthy dose of pleasure and meaning, and dare I say convenience as well. But managing design and experience at this level gets exponentially harder. Your baseline reliable functionality that's relatively usable is table-steaks. It absolutely has to do that or nobody will use it at all. But if you don't even try to shoot for some novelty, some fresh expression, they may use it a bit, but have no desire to come back. This is one of the biggest problems with MVPs (minimum VIABLE products). When was the last time you really enjoyed an experience or app or site and said to yourself, "Wow, that was a really viable experience!"?

The Product Roadmap

This is why we choose to design and build Minimum VALUABLE Products (I'd also have accepted Minimum Lovable Products). Because MVP is so common though, we don't even use that acronym. We use Cupcake, Birthday Cake, and Wedding Cake.

Cupcake (your minimum valuable product) is what absolutely must ship, otherwise, there's no point. It's important to note here that Cupcake isn't a horizontal cut of the hierarchy though, sacrificing convenience, pleasure, and meaning for a baseline product that merely satisfies at an intellectual level. No, Cupcake products cut vertically up the pyramid, capturing a bit from every layer.

The cake analogy is so powerful, because at their core, all three types of cake are the same. You've got flour, eggs, sugar, cocoa, icing, maybe even a creamy filling and a topper. A bite of a cupcake and a wedding cake are essentially the same experience, just in a less substantial form.

• Cupcake offers a compelling experience

• Birthday Cake enhances that experience

• Wedding Cake is the full realization of the product vision

Cupcake encompasses the things you KNOW you must deliver to provide real value.

Once that core value is delivered, you can begin on some Birthday Cake revisions, adding additional features, functions, dare I say, even sizzle, insomuch as they enhance the core experience of your app. You scale. You are able to handle more clients. But when you're building Cupcake, as much as you want that slick new feature or novel AR avatar experience, you have to justify that it's part of the core foundational experience and nobody would bother with your app without it, or, if it's an enhancement that can build onto the base in a fast-follow release (Birthday Cake). Wedding cakes are the "big show" of cakes.

Wedding Cake is what you envision your product to be two to five years from now, with all the bells and whistles that make Wedding Cakes so much more substantial than your humble cupcake. But all the while, you're really providing the core value in your initial cupcake offering, and getting hung up on fresh trends and unique, fun, and sizzling design can detract, and more often, even degrade, even break your core experience.

Have Your Cake & Eat It Too

To wit:

1. The Basic Design Process

   • The Brief—Effective planning and setting expectations

   • Research—Discover insights into the problems you’re solving

   • UX Design—Design like you know you’re right

   • Testing—Test like you know you’re wrong

   • Visual UI Design—Enhance the experience with brand guides and style

2. The Road To Destruction

   • Make it Pop!—Here be monsters

3. The Path To Quality

   • The Truth—Keep it real, make sure you stay focused on what matters

   • The Product Roadmap—Cupcakes! Delicious, moist cupcakes!

Let's seek to build Cupcakes that span the layer-cake hierarchy from functional, to reliable, to usable, past the chasm of convenience, even to pleasure and meaning.

Once we've designed the thing the users actually need that supports the business' goals, don't let sparkle get in the way of delivering a great, valuable product.