Unusable at Any Speed
The author provides a bid of detail about activists (Ralph Nader, Rachel Carson) who wrote exposes on widespread problems that the industry was either blind to or refused to discuss. He asserts that software can also have "deadly effects," which seems melodramatic, except for the case of the Therac-25 (a treatment device that, because of bad UI, killed a number of patients by radiation poisoning).
Unusable Interfaces
As the user base of the Internet increases, the individual user is, on average, less sophisticated and cannot be counted upon to interpret the UI correctly (or even to read instructions).
For applications in which physical harm could result from misuse (military, medical, transportation, power systems, etc.), there is a clear need for careful design, testing, and usage monitoring. (EN: the author does not go on, but there are other types of "harm" that can be considered, such as financial harm, compromise of private or sensitive information, etc.)
Computer software has traditionally been difficult to figure out: cluttered interfaces, confusing terminology, and absent or incomprehensible instructions. Web pages have tended to follow suit, and are further complicated by the commercial factor (attempts to draw the user's attention away from the task at hand to something the site operator wants them to see or do). The author provides a couple examples of error messages that tell the user nothing at all (usually indicating what went wrong with the program, not how the user can act to overcome the problem.)
One study suggested that 10% to 30% of working hours are lost to users who are either attempting to figure out how to do something or attempting to recover from a problem they were unable to avoid. Another study concluded that "even the easiest-to-use applications pose significant barriers to the user of online services." Other studies cite productivity increases resulting from the redesign of business systems with an eye toward usability.
Getting to the New Computing
Convincing those who control (by financial means) the development of technology that usability is a worthwhile investment is difficult. Too often, it depends on customer pressure: it is only when customers turn to a competitor, or return products, that it becomes clear that usability is a problem (EN: interestingly, the software industry's widespread policy of "no returns of opened software" was not motivated to protect against piracy, but to compel users to struggle with the software that they found to be unusable).
The author returns to Da Vinci as an example of the painstaking attention to detail: he didn't paint rough shapes, but paid close attention to the expressions on the faces of his subjects, the way that garments hung naturally on the body, the effect of light sources, and all the other details that seem inconsequential to conveying the general sense of a scene. Quality in software products is of a similar nature: it is not sufficient to deliver the bare minimum functionality to accomplish a task, and rely upon the user to figure it out. Quality is the result of attention to detail.
Historically, software has been inherently buggy and difficult, and developers have relied on this expectation and tolerance to continue to produce badly designed applications. The author suggests that bad design is the reason for such lengthy and detailed user agreements: the manufacturer knows their product to be bad, and wants to escape liability.
The author suggests putting a bounty on usability problems, in the same way some companies offer a bounty for finding software bugs. (EN: This is a thoroughly idiotic notion that would encourage developers to make errors in order to later get a bounty for pointing them out. It also assumes that the programmers have the knowledge and skills to make products usable). The author advocates activism: consumers should complain, lobby for consumer protection, form consumer groups to pressure the industry into doing its job properly. Alternately, he recommends recognition for those who have done well, such as the "Webby" award for well designed Web sites.
EN: Each of these suggestions are thoroughly idiotic. When legislation is imposed, industry undertakes the least possible effort to satisfy legislation, and does not address the problem (if required to hire a usability expert, they would ... and then ignore that person's advice and use them as a scapegoat). The consequences of bad software design are not sufficient to whip consumers out of their apathy, and "awards" tend to be faddish. The solution is simply for some companies to do it right and steal the business from those that do not, until such time as it becomes a natural imperative.
The Skeptic's Corner
Skeptics insist that you cannot change the way industry leaders operate, and that they will always seek to provide the least possible value for the highest possible price. Design is of little value, and usability a mere afterthought, and both of them are seen as waste of resources.
The author returns to Nader and Carson, an instance in which a single activist called attention to problems that others preferred to ignore, and had a significant and widespread impact.
Another common escape is the argument that perfecting the underlying technology will yield a more user-friendly experience. This, too, has not borne out in reality. Technology focuses on efficiency - doing things faster, using fewer resources, etc. - but this does not, and will not, make software more usable.