Google: Trapped in Its Own Walled Garden with Lots of Science Club Alums

August 30, 2023

Vea4_thumb_thumb_thumb_thumb_thumb_tNote: This essay is the work of a real and still-alive dinobaby. No smart software involved, just a dumb humanoid.

I read “MapReduce, TensorFlow, Vertex: Google’s Bet to Avoid Repeating History in AI.” I found the idea that Google gets in its own way a retelling of how high school science club management produces interesting consequences.

image

A young technology wizard finds himself in a Hall of Mirrors at the carnival. He is not sure what is real or in which direction to go. The world of the House of Mirrors is disorienting. The young luminary wants to return to the walled garden where life is more comfortable. Thanks, MidJourney. Four tries and I get this tired illustration. Gradient descent time?

The write up asserts:

Google is in the middle of trying to avoid repeating history when releasing its industry-altering technology.

I disagree. The methods defining Google produce with remarkable consistency a lack of informed control. The idea is that organizations have a culture. That cultural evolves over time, but it remains anchored in its past. Thus, as the organization appears to move forward in time, that organization behaves in a predictable way; for example, Google has an approach to management which guarantees friction. Examples range from the staff protests to the lateral arabesque used to move Dr. Jeff Dean out of the way of the DeepMind contingent.

The write up takes a different view; for example:

Run by engineers, the [Google MapReduce] team essentially did not foresee the coming wave of open-source technology to power the modern Web and the companies that would come to commercialize it.

Google lacks the ability to perceive its opportunities. The company is fenced by its dependence on online advertising. Thus, innovations are tough for the Googlers to put into perspective. One reason is the high school science club ethos of the outfit; the other is that the outside world is as foreign to many Googlers as the world beyond the goldfish’s bowl filled with water. The view is distorted, surreal, and unfamiliar.

How can a company innovate and make a commercially viable product with this in its walled garden? It cannot. Advertising at Google is a me-too product for which Google prior to its IPO settled a dispute with Yahoo over the “inspiration” for pay-to-play search. The cost of this “inspiration” was about $1 billion.

In a quarter century, Google remains what one Microsoftie called “a one-trick pony.” Will the Google Cloud emerge as a true innovation? Nope. There are lots of clouds. Google is the Enterprise Rent-a-Car to the Hertz and Avis cloud rental firms. Google’s innovation track record is closer to a high school science club which has been able to win the state science club content year after year. Other innovators win the National Science Club Award (once called the Westinghouse Award). The context-free innovations are useful to others who have more agility and market instinct.

My view is that Google has become predictable, lurching from one technical paper to legal battle like a sine wave in a Physics 101 class; that is, a continuous wave with a smooth periodic function.

Don’t get me wrong. Google is an important company. What is often overlooked is the cultural wall that keeps the 100,000 smartest people in the world locked down in the garden. Innovation is constrained, and the excitement exists off the virtual campus. Why do so many Xooglers innovate and create interesting things once freed from the walled garden? Culture has strengths and weaknesses. Google’s muffing the bunny, as the article points out, is one defining characteristic of a company which longs for high school science club meetings and competitions with those like themselves.

Tony Bennett won’t be singing in the main cafeteria any longer, but the Googlers don’t care. He was an outsider, interesting but not in the science club. If the thought process doesn’t fit, you must quit.

Stephen E Arnold, August 30. 2023

Comments

Comments are closed.

  • Archives

  • Recent Posts

  • Meta