Are You Machine Wise?

From Harvard Business Review, September/October 1997

“Are you machine wise?” The question stares out at me from an ad pitching mimeograph machines in the issue of the Literary Digest dated June 11, 1927. Its unknown author answers the question with a comforting certainty that the “machine wise” will buy the mimeograph, trusting as a matter of faith that it will perform important new work in the service of lower operating costs and increased efficiency.

Naturally, computer and software companies are making the same pitch today: Buy our machines to improve your business. But today’s and tomorrow’s machine-wise managers will be wary of that claim, because the nature of the machine and the challenge has changed.

In 1927, the challenge was to reduce the labor cost of collecting and disseminating information. Freed from the drudgery of manual reproduction, human clerks could be assigned to higher-value tasks. Tools were simpler then, and their benefits were more obvious.

Seventy years later, the challenge is very different. Machine use is essential but not sufficient. As our tools become ever more complex and interconnected and more central to the conduct of business, their benefits also become harder to recognize. Furthermore, executives need to know and understand the logic of the work done by machines–and, above all else, the limits beyond which those tools cannot be pushed.

Meanwhile, the volume of information continues to expand exponentially, generated by machines conversing with other machines on our behalf. Every business activity leaves behind a wake of information, from data spinning off production-line process controllers to transaction records generated over retail-credit-card networks. And the growing centrality of the Internet for business purposes will only add to the flood.

All our innovations have left us afloat in a growing sea of information, which we must navigate with tools that are far from being up to the task. We don’t even fully appreciate our predicament, wrongly labeling it “information overload” when it is not a consequence of the amount of information confronting us but rather of the gap between the volume of information and effectiveness of the sense-making tools that technology has built for us.

Better tools can narrow the gap. In the next decades, the most important new sense-making tools will be those that help people visualize and simulate. Visualization techniques reduce vast and obscure pools of data into easily comprehended images. And simulations systems will become intellectual training wheels for executives, allowing them to experiment with strategies in the forgiving world of cyberspace, in much the same way that pilots in the Gulf war ran practice missions before flying the real thing.

The gap can be narrowed, but the machine wise also know that it can never be closed, for the very tools we use for sense making generate even more information of their own. Information breeds yet more information, and information tools are formidable breeders.

If we are not careful, we will chase our new tools, Alice-like, down a digital rabbit hole of infinite information regress, and here is how it may happen. The temptation, as simulation and other sense-making tools become more sophisticated, will be to substitute them for human judgment. That would be a mistake.

Seven decades ago, machine-wise managers could embrace nearly any information technology, such as the mimeograph, with the faith that it would help them collect and disseminate information–the more information the better–and thus improve their businesses. Today machine-wise executives will not only know when and how to use the new tools technology brings them but also when to switch off their computers and take their own counsel.