Don Norman is a technological optimist. The author of The Design of Everyday Things and head of UC San Diego’s Design Lab believes that artificial intelligence might only take the worst parts of our jobs, and when it gets smart enough, it will pity us rather than destroy humanity. On these points, the scientist in him admits that he might be wrong, but Norman would prefer to live his life hoping for the best. Because nobody wants to go to sleep at night expecting a Terminator in his bed in the morning.
But that doesn’t mean Norman isn’t critical of the direction of Silicon Valley’s powerhouse companies often celebrated for merging technology and design. In a casual, stream-of-consciousness conversation with Co.Design, Norman dissected what’s wrong with Apple, Google, Microsoft, and, as a little bonus we pushed for, Tesla.
In 2015, Norman penned a scathing review of Apple’s design. We asked if he’d seen the company improve since then.
“I think Apple’s products have gotten worse, not better.
“I think in a number of areas, the quality of design has deteriorated [worldwide], and I think it’s because traditional designers have come back. Design is a complex field, and design itself sort of emerged accidentally as a craft. Only recently has it become a real field with real principles. There’s a real tension between the people I represent, who care a lot about whether or not people can understand and use a product, and the more traditional view, and the look and feel–the aesthetic.
“There’s nothing the matter with [aesthetic design]. Doing that doesn’t mean it has to be unintelligible, or making something intelligible and easy to use makes you lose aesthetic beauty and nice feel, but the people who understand how to make something understandable are a different type of people. They come from psychology, computer science, science. They understand the need of testing and data analysis.
“More traditional designers are more applied artists. They claim they take a human-centered approach, but they do it by thinking about the people, which is not the same things as using or testing with people.
“As design has suddenly become popular, the wrong designs have suddenly become popular [as they have at Apple].
“I’d argue that those of us trained in the science side are not really good at making wonderful, delightful, emotionally pleasing objects, because we lack those design skills, but the designers with those skills lack the understanding of making those things usable and understandable.
“If only we could bring those two groups together! And that’s happening more and more. But Apple is driven by someone [Jony Ive] with a very traditional design background.”
We asked Norman: Hadn’t Google’s design gotten better since it merged its data-driven design with decent, consistent UI and aesthetics?
“I agree with you. I think Google in many ways is better than before. I’m talking to you on a Google Pixel mobile phone!
“I have a different problem with Google, which is that they never finish the job. This is especially true in their software. Google Docs is a good example. Google Docs is a really brilliant concept; it’s really well done. For the first time, multiple people can work on the same doc at the same time. But I can never find a Google Doc. They have no organizational structure. A lot of us hate to use it because it’s 90% wonderful. If you could only finish the last 10%.”
“I think that today, the [design] leader is, surprise, Microsoft! I think they’ve always had very excellent people in their research groups. But the researchers barely talk to people doing the products. Microsoft has done a good job connecting the two. And Microsoft has had a good group of people doing human-computer interaction. They’re among the leaders in this field. And Microsoft is finally listening to them.
“Microsoft in the past, which maybe because of Bill Gates, was very argumentative. They divided themselves into divisions, they fought with each other. I heard lots of stories of where there were great products, but the Office division might say, ‘No, you can’t do that, because that might decrease the sales of our product.’
“The people I’ve talked to believe that Microsoft is under a completely different management structure that’s a lot more collaborative.
“It also helps to have a bunch of bad years. I’m really serious. Nothing is worse for a company than a couple years of fantastic success–you get complacent and think you can do no wrong. Historically, companies go up and down and up and down. It’s really good to have hit the bottom, to have had trouble–that makes you more willing to experiment.”
Given that Norman was on a roll, and he does a lot of research and consulting on self-driving cars, we asked what he thought of Tesla.
“Tesla bragged about releasing autonomous cars, but they weren’t fully ‘autopilot.’
“I think Tesla had too many engineers. And they were too logical. And they really didn’t recognize the way that people behave. Those of us who’ve been–a lot of my friends and I have worked long in aviation safety–we know the issues, we saw Tesla releasing stuff and said, ‘This isn’t going to work.’
“I wrote a paper 20 years ago that said automation gets more dangerous the better it gets. Because you rely upon it more and more. And I wasn’t the first person to say that!
“Tesla, though, has a tremendous virtue: They’re really a fast-learning company. Last night, at the dinner I was at, one of the people complained, in a good-humored way, ‘My Tesla changes almost every week when there’s a new software release. They just released a new Autopilot, and it’s much much better than before. It really requires me to pay attention, but it won’t break the law. I’m driving down the highway at 45, which is the speed limit, and I’m used to going 50. I’m the slowest person on the road.’
“That is a problem; if you look at Google’s car, a tremendous amount of the accidents it’s been in is that it’s obeying the law, and is very cautious as well. And many of us have experienced being behind an extra cautious driver in an intersection. It’s timid and so it waits, but behind that is another car and another car.
“I think Tesla is starting to learn its lesson. It’s unfortunate it had to learn it the hard way.
“All the other car manufacturers are cautious, though there is a pressure, and I see it because I work with car companies. The researchers are clearly aware of the dangers and unknowns, and they’re trying to exercise caution, but the car companies are in a race now to bring out the fully autonomous cars, so [they’re prioritizing speed over safety.]”
Notably, Norman believes that everything from the design of the autonomous driving experience to the algorithms powering it are a long way off from the level of perfection we need to take over driving across all roads.
“It was interesting to see Toyota’s new research group, founded on AI . . . made the pronouncement that we’re not going to see a fully self-driving car for 20 years. And that’s what I think. I think the first 90%, 95%, or 98%, that will be solved quickly. In some cases, we’re already there. But the last couple percent is really really hard. That’s what people don’t understand.”