Notes on “The Invisible Computer”

“The Invisible Computer” is a book by Donald A. Norman published in 1998. It’s basically a critique of personal computers at that time and the vision of future computing by Norman, just like what the subtitle says: “Why Good Products Can Fail, the Personal Computer Is So Complex, and Information Appliances Are the Solution.”

I have to admit, this is not the most influential work by Don Norman. But I’m really surprised that so many reviews on Amazon.com says this book is not worth reading. According to my understanding, many ideas discussed in this book were so important that we’re still facing them again and again today. And it’s encouraging to find that Norman was insightful on most of the issues and provided his solutions for us. We’re living in a world that things are changing to the direction very much as Norman predicted.

Here’re what I’ve learned from this book:

We must change as technologies matures

When technologies first came out, it’s amazing and wonderful. Technology enthusiasts dominated this field. But the majority of customers are not geeks. They are not going to pay for the devices that are fancy but difficult to use. However, the early adopters ignored that. Tech companies, media and fans were crazy about specs and features, not usability, not design.

There is a transition point (or the “chasm”) when technology could cover most customs’ basic needs. At the point, technology advancement alone is no long so attractive as it was. Other factors start to take effect, such as user experience and marketing.

Therefore we must change, from a tech-centered development model to a user-centered one. And we must change now, because this is what we see today: Apple is winning by delivering products that are not only technological advanced, but also much easier to use. You can’t imagine a smartphone without essential features like cut-and-paste, multitasking and MMS could be called a “smart” phone in 2006. But iPhone changed the entire game before the rest of the world realized what had just happened: an epic fail for technocentrism.

Creeping Featurism is highly addictive but unhealthy

Everybody loves features, that is what most tech companies thought. A feature comparison table was always needed for promoting products. Products keep being updated not mainly because there are bugs — it is because new features will be introduced. This logic worked, but not anymore.

The new features are added merely because of user’s demands, not careful consideration. The more features we have, the more complex the product is, and users will be confused. When it inevitably becomes huge, slow and hard to use, users will leave and find better solutions.

That’s part of the answer to a recently heated discussion on Quora: Why is Dropbox more popular than other programs with similar functionality?

Here’s what Isaac Hall, co-founder of Syncplicity, said:

In the end, it really came down to one incredibly genius idea: Dropbox limited its feature set on purpose… Our company had too many features and this created confusion amongst our customer base…

If you’re starting a new company, the best thing you can do is keep your feature set small and focused…

I think that explains the idea.

Update: Hall removed his answer on Quora later, but it’s already everywhere on the Web.

AI (or Cloud, 3D, smartphones) wouldn’t save the world

They will make the world better, but they will not save us from the complexity of computers.

AI (artificial intelligence) is great for simplify our tasks, help people achieve something. But it’s very unlikely that computers would really understand this world in the near future.

Watson, the IBM computer designed for “Jeopardy!” game show beat its human competitors on Feb 14–16. What I’ve concluded from the event is that, even if Watson could understood all the questions and answers in some sense, it did not understand the world as we do. Human beings and computers are completely different types of creation. Computers are good at things we aren’t but we are doing better in things that they can’t. In Norman’s words, “together, we are a more powerful team than either of us alone.”

AI makes computers clever, but “more intelligence” does not lead to “more user-friendliness”. I don’t see the point for computers to simulate human kind for better usability. We’d better take advantage of computers instead of introducing the sophistication of humanity to computers.

The drawbacks are similar with other proposed solution. In general, the problem caused by technocentrism could not be solved by technology inside out. We need to consider this from outside in.

Cloud is the infrastructure of pervasive computing

Though Norman didn’t mention “Cloud” directly, lessons of infrastructure is discussed in the book and I think it’s closely related to what we call “Cloud” today.

Cloud computing is an overly-used term that already starts to cause all kinds of misunderstanding among customers. But I don’t think Cloud is overestimated. It’s not just another concept invented for selling software (like some people believe), but it could be the foundation of every future major development in this field.

The vision of pervasive (or ubiquitous) computing is likely to be:

Integrate computation into everything.

Literally, everything. Machines. Walls. Skins. Or even dust, (though there’re some issues. See my explanation below.) In the meantime, my interpretation of Cloud computing is also very simple:

Supply generalized computation power via the Internet.

It basically means that once you’ve got an Internet connection, you could get all the power of the Cloud, just like the supply of electricity.

So it becomes obvious that if we want pervasive computing, we’ll need computation power provided by the Cloud. When Cloud becomes infrastructure, pervasive computing is just two steps away: an Internet connection and the integration. The latter one is not easy, but become more straightforward with the support from infrastructure.

The invisible computers do not hide from view

Many of us are trying to make smaller embedded systems that could be easily hide from the users view. Or some others are trying to put computers into walls, buildings, roads. These are all very meaningful work, but does hiding computers make them invisible?

No, it’s not the implication of “invisible”. The invisible computers are the ones that nobody notices but are still working for us. They don’t hide from view, they hide from our minds. We should make them as intuitive to use as possible, not necessarily as small as possible.

Watch your customers and talk to your non-customers

This topic is discussed from time to time because the term “user-centered design” doesn’t directly imply the role of users in the design process. But this is pretty clear if you read Norman’s books.

Fans (aka. the most satisfied users) don’t always tell truth because they are easily misguided. They will tell you what they want, but what they want is not necessarily what people really need. We should listen to what they say, but we shouldn’t always do like that. This is similar to the relationship between patients and physicians. Patients will tell the doctor about his symptoms. But patients are not experts, doctors are. The diagnosis should base on facts, not what they say.

So why even bother talking to your non-customers? Because they don’t actively give feedbacks like fans do. We should understand why they don’t use the product, which is often ignored.

Apps are sweet

At the end of his book, Norman gave a few examples of “Information Appliances”, which depicts his vision of the future computers. Computers won’t be one single device that are used for all sorts of tasks. They’ll be part of every information appliances which will be designed and made specifically for one particular task. He believes that this is the only way that we could solve the complexity of computers.

I agree with him on the main idea of information appliances. Actually, the apps we have now on multiple platforms are kinds of early forms of informations appliances. With all kinds of apps, iPhone could become all kinds of information appliances which is designed specifically for some tasks. In the appendix of the book, Norman talked about appliances of photography, medical advisor, weather/traffic, shopping, gardening, reference and financial… Do they look like names of Apple iOS App Store categories?

Unfortunately, none of those happened before most readers read this book. When this book was published in 1998, how could people understand what Norman was talking about? I just can’t imagine.