Digital Assistance

Posted by: by John Parsons on February 21st, 2012

Do you remember the Knowledge Navigator video? In 1987, shortly after Jobs’ first departure from Apple, the Scully regime produced this futuristic look at collaborative technology.* In hindsight, it seemed to anticipate Mosaic, WiFi, Skype/FaceTime, Siri, and many other technologies. For geeks of the Mac persuasion, it was a positive, HAL-like experience, without all the scary Stanley Kubric bits.

An appealing aspect of the video was the deferential “digital assistant,” a bow-tied avatar that “understood” your schedule, your current projects, your trusted colleagues and their work, plus a myriad of information sources that might be relevant to your needs. (Yes, it also looked like a younger version of Bill Nye, but without the sense of humor.) It added a human touch, of sorts, creating an interface that simulated human interaction—freeing the actual human to ponder, speculate, and apply himself to actual problems.**

Let’s leave out the cosmetic issues for the time being. Voice recognition, voice and visual synthesis are not small problems, but they are gradually becoming mainstream. What intrigues me is the filtering or “tuning” of data that such a digital assistant would need, in order to carry out the tasks in the video.

First of all, the data—and its related metadata—would have to be standardized to a high degree. Statistics on rainfall, land use, and the like would have to be not only accessible but recognizable at a high level of abstraction—in this case for a world map simulation that allowed for various “what if” scenarios. This kind of visual rendering can be done today, but at the cost of many, many programming hours.

Second, there needs to be a more flexible, human-like approach to varying conditions or assumptions in the data. Computing today is by nature linear and logical, while human thought—at least the kind we value most—is intuitive, spontaneous, and able to reach conclusions with incomplete data. While it may often be flawed, the human process is decidedly non-digital.

Technophobes fear the competition of a computing process that competes with our own intuitive process. After all, back in 2001, we all shivered when HAL decided we were expendable, right? (Come to think of if, how do you know this blog wasn’t written by a cyber agent, softening up humanity for the takeover?)

Just kidding. Computing that transcends today’s linear model will eliminate more mundane jobs, to be sure, but it will also open up new ones. If digital assistance is truly that—giving us time to do the important stuff—then we’ll be better off.

– John Parsons

* Go to to see the 5+ minute video. Spoiler alert: To those accustomed to Apple’s way-cooler videos under Jobs 2.0, it is corny and clichéd, but still worth a look.

** The video is not without irony today. The lead human character is a white and evidently very affluent professor living in an expensive, high-carbon-footprint home—discussing climate, deforestation, and policy changes that other countries should adopt.

Leave a Reply

Copyright ©2019 Nimbleware Consulting