César Hidalgo on personbytes and knowledge networks

It’s sponsor week at the Media Lab, the semi-annual “open house” where Media Lab students and researchers share their work with the foundation and corporate folks who pay for it. It’s Joi Ito’s first sponsor week as director of the lab, and there’s an emphasis on making this week more open and visible to the outside world. Our sponsors are now “members”, and our hope is to be sharing the new research happening here at the Lab with the members and with the wider world. Most of the meeting is being streamed online, if you want to follow along. There are journalists in the crowd, for one of the first times. And my blogging is my modest attempt to help on this front.

César Hidalgo, a year into his teaching career at the Lab, is the program chair for the conference, and has brought us together around the idea of knowledge networks. He invites us to think about what “media” means. He shows us a Van Gogh painting and points out that the gallery label associated with the painting. It lists the artist, the work’s name, the date painted and the media, oil on canvas. “What is left if you remove the oil and the canvas?” Hidalgo wonders. Nothing physical is left – what remains is the media as a vehicle for knowledge, the unique perspective Van Gogh had.

We can think of physical objects as containers for insights. Michael Faraday’s work on uniting magnetism, electricity and light is embodied in the electric motor. As a result, we can see a vacuum cleaner as a vehicle for Faraday’s laws, or a harp as a vehicle for Pythagorean thought about geometry and harmonics.

How do you get knowledge into a knowledge vehicle? “If you have a bad dental infection, would you rather have a good article on root canal surgery or a dentist?” You want knowledge embodied in people. And you want those people to have knowledge embodied in their equipment – the metalurgy to build dental tools, the skill to build an x-ray system. Dentists don’t usually build their own tools – we each hold only a little knowledge personally. If knowledge means understanding something well enough to build it, most of us don’t know enough individually to do everything we need to do.

To function in the modern world, we need an enormous amount of knowledge. Hidalgo suggests we consider the “personbyte”, the amount of knowledge a person can know. Generations ago, it might be possible for a person to know most of what was known by people. Now there’s no possible way one person to know all of human knowledge. A project like a root canal requires may peoplesbytes of knowledge, embedded in tools and systems. Rather than knowing everything, as we did in prehistoric days, we distribute knowledge through networks.

If we understand that knowledge lives in networks, we discover that markets make us wiser and organizations make us smarter. Knowledge began to accumulate as people got together in towns and cities, but now we organize within organizations. But there’s a limit to the ability of that model to scale. Add 50,000 people to a 50,000 person organizationm, and you are unlikely to double the amount of knowledge you can hold. At very high levels of knowledge, people need to share knowledge between firms, to learn through networks of organizations.

This is an unfamiliar situation for humans. We’re used to trying to horde our knowledge. At low levels of knowledge, this works. If you know how to make a really good spear point, protecting that knowledge gives you a huge advantage over your comeptitors. But once knowledge gets bigger, you need to share knowledge within your firm, but you’re unlikely to share it more broadly. But we’re now reaching a world of knowledge where we can only understand what we need to know by building networks of networks and networks of firms. He leaves us with the provocation, “Anything that is worth doing can not be done alone.”

This entry was posted in Africa. Bookmark the permalink.

One Response to César Hidalgo on personbytes and knowledge networks

  1. Paul May says:

    Very interesting; there are just so many fascinating challenges to be thought about and solved – from the obvious friction of gathering, sharing and representing huge volumes of data – to the less obvious (and far more complex) challenges sharing and representing understanding. We can put people in contact with understanding, but can we store it or represent it? I certainly don’t know!

Comments are closed.