Friday, November 09, 2012

The Myth of Multitasking

As an engineering summer student, I had some fantastic summer jobs at Glenayre electronics in Vancouver, where 8-bit microprocessors were already changing how telecommunications systems worked. Part of my job was about learning to use, and debug, what were called multitasking operating systems. They had names like MPM and Kadak.  And there I found a secret.

There ain't no such thing as multitasking. All these system do is quickly jump from doing one small part of a task to to another small task, and because these things are so fast compared to we carbon-based life forms, if you squint your eyes a bit you can imagine they are doing those thing at the same time.


Why did developers of so-called multitasking operating systems do it this way? Why didn't they let the computer finish one thing before it started another, like the annoying construction foreman in Tonka Construction? I think it is because since people are so slow compared to computers, that the CPU might as well try to do something useful rather than wait for billions of nanoseconds while a human's fingers hover over a keyboard awaiting inspiration.

Making a computer change from one task to another is surprisingly easy. It's called a "context switch", and for a computer it is a simple, quick task. Most everything the actual CPU does is stored around it, in memory (RAM), or on the hard disk, or flash memory, things that don't change if you leave them alone, like the way you've spread papers around on your desk. The actual information a CPU has to "swap out" to change context is a few hundred bytes. That's because computers are not deep thinkers. They don't pore over ideas, and hold several in their little imaginations, and consider the technical or artistic merit of one over the other. They just move memory around, and compare, add and subtract numbers.

The illusion of multitasking only works if the amount of time you can spend on a task is far greater than the time it takes you to switch from one task to another. If you give a computer so many tasks that it spends all its time switching from one task to another instead of accomplishing anything, it might seem very busy, but it is accomplishing nothing.

Which leads to this obvious but important statement. You are not a computer. You are not as fast as a computer, and for you, a context switch takes anywhere from several seconds to several minutes, and requires millions of neurons.

If you think you are a multitasking machine, guess again. All you can do is switch your attention from one thing to another. Whichever thing you are paying attention to is swapped in, and nothing else is going to get that kind of attention.

Be human. Focus. 

Monday, January 23, 2012

Kodak's Self-Destruction, an Inside Story

I worked for Kodak for four years, but it wasn't by choice.

Kodak purchased a Vancouver company named Creo, as part of their vision to convert Kodak to a digital company before the roof caved in from falling film revenue. The strategy was sound. The execution - a disaster. There are three kinds of mistakes we saw as Creo-ites, mistakes that were an anthema to the values that the founders of Creo sought to avoid, and did so successfully, for the years that led to Creo's phenomenal growth and eventual move to a public company. While speculation runs amok in the popular press about Kodak’s demise, I can tell you what it looked like from the inside.

Kodak executives were not naive about the future of digital cameras, and indeed Kodak was in the best market position to bring out the “brownie” of digital cameras - the one that would have made Kodak the household name in point-and-shoot digital. What was lacking was not vision, but courage. Kodak management just could not bring themselves to take a terrific new camera to market, even though Kodak had the Brownie, Instamatic, and Starmite lines in their DNA. In the end they just could not muster the courage to actively end the reign of film themselves. The camera is dead - long live the camera! should have been the battle cry, but no executive vice-president was prepared to be the person perceived to be the reason for the demise of the film cash-cow. Instead, they handed it to Fuji to decimate the Kodak film business, instead of doing it themselves, and losing the market share of film and future digital cameras to their major competitor.

The second thing we learned about how badly things were in Kodak was the exposure to the incredibly large, self-perpetuating bureaucracy that took us from 4 layers between the CEO and staff to about 12. Executive VPs, VPs, executive assistants reflected the 1950’s model of a happy Kodak family living off 90% gross-margin film and loved by America. We watched competent professionals, who made rational arguments and economic decisions, replaced by Kodak company-bots who made a career out of squeezing the innovation out of one division, showing an increased profitability, then moving onto another group before the other collapsed. The kind of pyramid promotion scheme we had only seen in movies was being played out every day, and the shell-game players rewarded each other with promotions and bonuses. A very abrupt insight into American corporate politics, and a foreshadow of the inevitable time when the music stops and the carefully-planned moves of the survivors ensure their bonuses were safe.

Finally, the last element that destroyed Kodak is that they lost sight of who they are. Antonio tried to remake the company into a digital imaging force in the market, to prove that his ouster from HP was a mistake, but HP executed on the digital imaging strategy far more easily, with their purchase of Indigo, and their active self-cannibalization of the LaserJet by transferring energy to the DeskJet line, in a move that preserved their overall consumables strategy. Kodak was not a digital imaging company any more than it was a film imaging company. By losing sight of who they were, Kodak let the 1984 Olympics torch pass to Fuji, starting Fuji's American ascent at Kodak's expense. Kodak’s retired CMO Carl Gustin I think understood this: Kodak was the memories company. Kodak moments were the true American spirit behind the yellow boxes, not the chemically-treated plastic inside them. Kodak glimpsed that vision with the Gallery - at least the Kodak Gallery that could have been. Once again there was no lack of vision for Gallery, from the digital generation of marketing folk. One could imagine the thousands of images uploaded every minute worldwide, correlated by time and location for shared current events, like a tsunami or a rally; by location over time to build a retrospective of a famous landmark through the years. Instead, while YouTube recklessly let people upload anything without limit, the Gallery threatened users that their precious images would be deleted (deleted from their shoebox!) if they didn't buy stuff. The Kodak Gallery could have been the Story of America, shared, over time, instead of lost in shoe-boxes in the attics of the previous generation.

In the end, being stuck in the previous generation is what did Kodak in. Kodak dies an old man, friendless, having squandered his children’s inheritance on dreams of past glory, instead of investing in that same next generation, and treating digital as a tremendous new opportunity instead of an old threat to be managed.

RIP Kodak, and thanks for the company that invented continuous film, the Brownie camera, the Instamatic, and motion picture film that captured nuance and shadow to tell great stories.