Joi Ito's conversation with the living web.

Here are some thoughts about leadership as I prepare to participate in IBM's THINK Forum.

The Internet has enabled the cost of the production and distribution of ideas and information to plummet nearly to zero-resulting in an explosion of ideas and a low cost of collaboration. This has prompted a great deal of innovation, but also a complexity, speed and capacity for amplification that makes the world a difficult and dangerous place for many organizations and human-made systems designed for a slower and simpler era.

The cost of planning, predicting and managing rapidly changing, complex systems often exceeds the cost of actually doing whatever is being planned and managed. In fact, it can be often easier to try something and iterate than to try to predict the outcome and manage the risks. Most great ideas as well as dramatic failures have been unpredictable and are only obvious in hindsight. (Don't get me wrong: foreknowledge and planning are useful and, often, necessary; they're just not sufficient.)

In such a world, leadership hinges on the ability to master a broad set of skills and character traits necessary for fostering a robust system, including courage, flexibility, speed, values and a strong vision and trajectory. It's more important to have a strong compass than a detailed street map since the map is probably outdated and wrong.

These kinds of decentralized models of leadership have been evolving and emerging in a variety of situations ranging from battle (virtual and real) to religions. The Internet has just super-charged the importance of this type of leadership in almost every organization.

Managers in large corporations no longer have the promise of promotions and long-term employment to keep employees obedient and hard working. Central corporate R&D and planning organizations can no longer provide detailed maps of the world to their staff and partners. Innovation is happening in the most unlikely parts of the organization-often outside of the organization.

Leadership today is about empowering those around you share your vision, embrace serendipity, have the courage to take risks and learn from failure rather than be crushed by it. Diversity must be embraced and organizational borders made porous. Assets such as intellectual property and lines of software code must not prevent aggressive agility. Organizations must be willing and able to pivot away from attachment to such assets lest these assets become liabilities holding back innovation and progress.

In this new world, leaders must be courageous, visionary and comfortable in an environment where control and complete knowledge are impossible and their pursuit futile and counterproductive.


Once the network infrastructure and dynamics have been mostly understood by people, I have often seen them trying to modify the network to meet their own power structure. A good example of that is marketing professionals applying the property of the network to achieve the same crap, instead of grasping the nature of the network itself. Another example would be putting an encryption system on digital content instead of maximizing the property of the network to circulate content. I have seen managers using the opportunity of an always connected employee to better monitor his/her time at work instead of using the positive property of the network allowing the employee to be in contact with more people.

Maybe there is a tendency to reuse the constraining systems they are used to instead of really adjusting to the distributed nature of the network. It is also deeply recorded in our mental structures. We classified information for centuries in physical boxes (because we had to keep it in order). The content could go in one box only and not another one. It's either here and there. A first process of categorization is to put all similar boxes in the same room, and then all similar rooms in the same building. A taxonomy, a categorization tree, an organization chart, etc. all of these are classification *tree* (hierarchical).

Once you digitize the interaction, once you make the possibility to connect structures and information with a graph (point to point), the tree disappears it is not becoming necessary at all. You may have an important role as a hub, someone helping the information to go in the right directions and connecting people, but not anymore as being the only node where the information flows. The information in a graph will follow the easiest immediate path. Authority being one source of slowing things, people will choose another node which ease their interaction.

"Leadership today is about empowering those around you share your vision, embrace serendipity, have the courage to take risks and learn from failure rather than be crushed by it."

It is so refreshing to see this perspective permeating our business culture today. People are finally realizing that people are involved with leadership! Who would have thought, right! haha.

Hi Joichi Ito,
I am writing from the Economist Group based in Hong Kong. We have an invite that we would like to invite you to join. Would you please email me your details on or call me on (852) 2585 3893 and I will share more details.
Kind regards

The iterative approach cannot be used in all situations. In many cases, you have only one chance and if you fail, you can't backtrack and iterate.

Another issue with the iterative approach is that, it can be hard on employees. Edison failed 1000 times until he finally invented the light bulb. What percentage of people can tolerate 10 failures and continue working on the problem to get it right at the 11th iteration? Add to that the lack of the "promise of promotions and long-term employment". Then, the employees will have the option of believing the manager again, that they will get it right next time, or simply moving to a new company and starting from scratch. Most employees will pick the latter, specially because the grass is always greener on the other side of the fence.

Leave a comment