Metaphorically Speaking

What's the use of all that electronic information if you can't get at it?
Economist StaffNovember 9, 2004

The two biggest consumer-technology successes of recent times are a
white page and a wheel. The white page belongs to Google, the world’s
most popular search engine; the wheel to Apple’s iPod, the world’s most
popular portable music player with a hard disk. Both form part of
so-called “interfaces” — metaphorical gateways through which humans
enter and navigate around a technology. Both are also picture-book
examples of simplicity concealing complexity underneath.

The white page is said to have come about as follows. In its early
days, Google kept receiving strange anonymous e-mails containing only
the number 53. Sometimes they stopped coming, then they started again.
Eventually, one of Google’s geniuses figured out that the e-mails
arrived whenever Google had made changes to its web home page that
expanded its word count beyond 53. The anonymous adviser was telling
Google to keep down the clutter (although why he picked 53 as the
cut-off point remains a mystery). In August this year, Google made the
biggest stockmarket debut of any technology firm in history. The
current word count on is 27.

As for the iPod, “It is successful because it’s simple,” says Paul
Mercer, the brainfather of its interface and the founder of Iventor, a
technology-design firm. “It does few things, but some subtle things,
and it is fluid.” The simplicity comes from the wheel itself; the
subtlety comes from features such as the acceleration built into the
wheel, so that it seems to sense whether the user wants to scroll
through songs slowly or fast. The genius lies in what is absent — there
is no “fast-scroll” button. Instead, says Mr. Mercer, the “technology
materializes only when needed”, and thus “seems to intuit” the user’s

Drive Business Strategy and Growth

Drive Business Strategy and Growth

Learn how NetSuite Financial Management allows you to quickly and easily model what-if scenarios and generate reports.

Google and the iPod are successful because each rescues consumers from
a particular black hole of complexity. Google does it by putting a
white page on top of the googol (the number 1, followed by 100 zeros)
of potential web pages. The iPod does it by letting music lovers, in
effect, carry all of their CDs with them in their pocket. Both
solutions require an enormous technological apparatus behind the
scenes. Google is said to operate some 100,000 servers. And Apple had
to configure the iPod so that it automatically and fluently talks to
iTunes, the music application that runs on users’ PCs. Transferring
songs from the PC to the iPod now requires nothing more than plugging
in a single cable. (Both companies, incidentally, are notoriously
secretive and refused to be interviewed for this survey.)

More Flops Than Hits

Perhaps the most startling thing about Google and the iPod, however, is
the fact that they stand out so much. There are very few other recent
examples of interfaces that have opened up entirely new avenues for
technology to change human behavior. Yet breakthroughs on this scale
are needed if technology vendors are to see their visions come true.
Those visions, remember, assume that people will increasingly connect
to the internet through devices other than the PC. These gadgets will
either have smaller screens, as with iPods, mobile phones or watches,
or larger and more remote ones, as with TV sets or even, perish the
thought, car windscreens.

Small screens require simplicity for two reasons, says Mr. Mercer. One
is the “lack of real estate”, i.e., very restricted space, meaning that
not much fits on to the screen at one time. The other is that the
method of input is different, because there is either only a tiny
keyboard or none at all. Mary Czerwinski, a cognitive psychologist at
Microsoft who calls herself the “visualization and interaction boss”,
has also found big gender differences. For whatever reason, women
struggle with small screens, whereas men do almost as well on them as
on PC monitors.

Large screens, for their part, require simplicity because they tend to
be further away than a PC monitor and operated by a remote control, or
because of the context in which they might be used. “Simplicity is a
must-have when you’re driving,” says Jack Breese, Microsoft’s research

Even for the traditional PC, however, a new interface is needed. The
present “metaphor”, in designer-speak, of a desktop surface was Apple’s
key commercial breakthrough that launched the PC era in 1984. This
broad metaphor also lent itself to sub-metaphors, including
object-icons such as a rubbish bin (also the work of Mr. Mercer when he
worked at Apple in the 1980s), folders and files. Microsoft eventually
copied these metaphors and brought them to the mass market, thus
helping to make millions of computer users more productive.

But now that the internet era, in which everything is connected, is
taking over from the PC era, in which computers were mostly isolated,
these old metaphors are becoming increasingly redundant. PCs are
turning into crowded repositories of family photographs, songs and
e-mails alongside word documents and spreadsheets, and point to
locations on their own hard disks as well as to computers far away.
This is too much to keep track of on one desktop. “Making everything
visible is great when you have only 20 things,” writes Mr. Norman in
“The Invisible Computer”. “When you have 20,000, it only adds to the
confusion. Show everything at once, and the result is chaos. Don’t show
everything, and stuff gets lost.”

The desktop metaphor is collapsing under the weight of data overload,
says Tim Brown, the boss of IDEO, a design firm in Silicon Valley.
“Browsing in the old sense of the word becomes pointless,” he says, and
“filtering becomes crucial.” This applies both to items that are stored
on the user’s PC and to those on the internet because, in an always-on
world, the distinction becomes irrelevant.

Hence the excitement about Google. Its algorithms have so far been
directed only at websites, but it plans to deploy its search technology
to help people find their own documents as well. Google is currently
soft-launching Gmail, a free e-mail service that offers one gigabyte of
free storage. This could be a first step towards letting customers
store all their data on Google’s servers, where they will be easily
searchable, instead of on their own PCs. In a parallel move, earlier
this month Google offered free software that searches the local hard
disks of PC users and displays the results much like those of a web

Naturally, this has struck fear into Microsoft, whose Windows system
runs 94% of the world’s PCs and which sees itself as the ruler of the
desktop. Yet Microsoft understands the threat that data overload poses
to Windows’ current metaphors. Bill Gates, Microsoft’s chairman and
software boss, regards this interface crisis as one of the biggest
challenges for his firm, alongside the security holes in Windows and,
perhaps, the threat from Linux, an open-source operating system.

His plan was therefore to introduce new metaphors in the next version
of Windows, code-named Longhorn. Instead of files and folders, it would
use fancy new search algorithms to guide users through their PC. This
technology, called WinFS (which stands either for “file system” or
“future storage”), was to turn Longhorn into relational databases so
that users would no longer need to remember where they put things,
because the interface would automatically retrieve data for them as
needed. Alas, in August Microsoft announced that Longhorn would be
delayed until 2006 and that its gem, WinFS, had been dropped from it
altogether. Gleefully, rivals now refer to Longhorn as either Longwait
or Shorthorn.

Honey, We Need to Talk

Even the mockingbirds, however, cannot agree on what metaphor should
replace the desktop. One favorite seems to be some kind of “personal
assistant”. But that may be promising too much, because what makes
real-life assistants helpful is that they are able to make sense of
their bosses’ inchoate ramblings. In computing, says Microsoft’s Mr
Breese, “the holy grail of simplicity is
I-just-wanna-talk-to-my-computer”, so that the computer can “anticipate
my needs”. The technical term for this is speech recognition. “Speech
makes the screen deeper,” says X.D. Huang, Microsoft’s expert on the
subject. “Instead of a limited drop-down menu, thousands of functions
can be brought to the foreground.”

The only problem is that the idea is almost certainly unworkable.
People confuse speech recognition with language understanding, argues
Mr. Norman. But to achieve language understanding, you first have to
crack the problem of artificial intelligence (AI), which has eluded
scientists for half a century. In fact, the challenge goes beyond AI,
according to Mr. Norman, and to the heart of semantics. Just think how
difficult it would be to teach somebody to tie a shoelace or to fold an
origami object by using words alone, without a diagram or a
demonstration. “What we imagine systems of speech-understanding to be
is really mind-reading,” says Mr. Norman. “And not just mind-reading of
thoughts, but of perfect thoughts, of solutions to problems that don’t
yet exist.” The idea that speech recognition is the key to simplicity,
Mr. Norman says, is therefore “just plain silly”.

He concludes that the only way to achieve simplicity is to have gadgets
that explicitly and proudly do less (he calls these “information
appliances”). Arguably, the iPod proves him right. Its success so far
stems from its relative modesty of ambition: it plays songs but does
little else. In the same vein, other vendors, such as Sun Microsystems,
have for years been promoting radically stripped-down devices called
“network computers” or “thin clients” that do nothing but access the
internet, where the real action is. Such talk horrifies firms such as
Microsoft, whose financial fortunes rely on clients getting thicker so
that they can sell software upgrades. But in the end the minimalists
may be proved right.