Digital Accessibility - we have to talk about Jakob Nielsens Ideas

Jakob Nielsen - the guru of usability research - has written a provocative article. Even if I don't agree with all of his statements, I think that the majority of his critics are wrong. Nielsen's core thesis is that digital accessibility has failed and that one should try AI-generated, automatically adapted user interfaces - so-called generative UIs.

Nielsen is not an accessibility expert

First of all, Nielsen hasn't stood out for great commitment to accessibility. I haven't heard anything from Nielsen since my last intensive involvement with usability almost ten years ago. His statements do not lead to the conclusion that he is really deep into the topic:

Before turning to my recommendation for helping disabled users in general, let me mention that two huge groups of users can indeed be helped with current approaches: old users and low-literacy users.

By current approaches he probably means WCAG. However, there are few starting points for older people and in the WCAG AA none for illiterate persons. What follows is usability and not accessibility, he seems to have mixed something up.

As I wrote elsewhere, I am also sure that the high degree of customizability of User interfaces will become increasingly important in the future. This applies not only to disabled users, but to all users. My guess is that users will create profiles at the operating system or browser level, which will then be automatically applied to interfaces. This is not a thing of the future; it is already happening to a certain extent today; the algorithms are getting better and better. I also suspect that the trend in design is moving more and more towards adaptability rather than pixel accuracy, just as we talk about responsive design today. This is particularly important for the three groups that have been essentially ignored by our accessibility experts: the visually impaired, the cognitively disabled and the neuro-diverse. They do not take place in WCAG 2.x. If this happens on the client side, I don't see any data protection issues. It is true, however, that such adjustments must be as error-free as possible. For example, no relevant content may be removed. This is particularly relevant when it comes to linguistic comprehensibility: content must always be correct, even if it is automatically translated into understandable language. Although we have made enormous progress here, it is still far from the case that we now have an acceptable error rate.

Nielsen is wrong, however, when he suggests that this would solve all accessibility problems. Automatic adjustments are also helpful for blind people, but more in the sense that they remove redundant content. For example, when I buy a ticket or do online banking, the content from the navigation and the footer bothers me, I only need the form. It would already work today for websites with semantically correct labels, but I don't know of any corresponding functions for hiding them.

Nielsen writes:

Accessibility is doomed to create a substandard user experience, no matter how much a company invests, particularly for blind users who are given a linear (one-dimensional) auditory user interface to represent the two-dimensional graphical user interface (GUI) designed for most users.

It is correct that blind people only perceive information linearly, but this is due to the technology and does not apply to touchscreens. Although two-dimensional Braille displays are now readily available, they have not yet become widespread. I don't currently see how generative UIs are supposed to change that. I don't understand what he means by auditory interfaces, perhaps a kind of voice assistant like Alexa. But there's a reason why Alexa is rarely used for more complex tasks.

In conclusion, Nielsen does not make a clear distinction between UX and accessibility, but rather mixes them up in his article:

Moving to second-generation generative UI will revolutionize the work of UX professionals. We will no longer be designing the exact user interface that our users will see, since the UI will be different for each user and generated at runtime. Instead, UX designers will specify the rules and heuristics the AI uses to generate the UI.

As I said above, I would agree with Nielsen when it comes to the design of graphical user interfaces. Where, in my opinion, he is completely wrong is the topic of screen readers and voice control: In other words, where it is currently the code level and not the GUI that is important. It may be that other AI technologies can help here, but Nielsen is not talking about that.

Nielsen also speaks of a not too distant future. So there is no reason to abandon current approaches to accessibility unless we have something better.

Nielsen's critics

I fear that much of the accessibility scene is resistant to criticism. It is often observed that things go down to the personal level and people then simply forego arguments altogether. An example is the compilation by Adrian Roseli, which I Would recommend to visit a yoga class. Well, that stays in the A11Y bubble anyway and will probably neither cause any particular harm nor interest to Nielsen. Leonie Watson's contribution was more amusing, saying that Nielsen would be wrong because she can order her lunch online. As long as Leonie can do this, everything is okay with accessibility.

I found one of the few objective reviews written by Brian DeConinck. He writes:

In any case, WCAG certainly does have deficiencies. It is not sufficient to ensure an accessible experience for all users. But it’s a well-trodden path of patterns to which users are accustomed that ensures some baseline consistency in how websites behave and how assistive technologies are supported.

And further:

Beyond that, when used by an experienced practitioner, WCAG is a tool for identifying things beyond just “letter of the law” conformance. WCAG provides a series of pass/fail tests, but as a sum of its parts it also describes a philosophical approach for ensuring accessible outcomes.

I would contradict both statements: The WCAG primarily helps users who would probably be able to cope with a moderately barrier-free application. It does not help the majority who are not technically experienced or poorly equipped. The second statement is also wrong: I almost only know experts who interpret the WCAG literally like the Bible. There is no philosophical or holistic approach. Today's testing procedures break down the surface into 60 testing steps instead of taking a comprehensive approach.

Conclusion

I'm admittedly not a fan of Nielsen, but I'm also relatively disillusioned with the accessibility community. The aggressiveness and personal attacks on him do not seem justified to me. The reason is probably not the thoughts in the article, but the heading, nobody likes it when somebody said that he has failed. Instead, as I did, one should see it for what it is: a contribution to a long-overdue discussion. As I have shown, you can analyze and critically evaluate Nielsen's theses even without foaming at the mouth. It offers enough attack points.

I predict that the scene will be increasingly challenged in the future, probably by new technologies and newbies who no longer want to accept the way things have been. And then even fewer people will be interested in what say have to say.

Over the last few months, I've had the impression that the scene can't get over the fact that it no longer has control over the discourse. We have been in our ivory tower for so long that both the overlay providers and many other players have pushed us out. In the end, it doesn't matter: the technologies that Nielsen describes are not specifically aimed at disabled people, but at all users. And they will use it as naturally as they us ChatGPT does today, even if we say it doesn't work.

It's been a long time since we've heard anything constructive from the scene: The experts just tell us that this doesn't work and that doesn't work, sometimes they're even right, but sometimes they're trapped in their dogmatism.

Read more