Abstract Knowledge networks have two contradicting demands, increasing flow and decreasing cost. Maximizing the sum of these demands will result in an optimized network. On the basis of the evolutionary optimization algorithm, we suggest a quantitative way to produce an optimized technological knowledge network where optimality is defined as the maximization of overall benefit. Applied to Korea’s ICT network, such a network results in a total knowledge flow increase of up to 75% in absolute measures and 150% with due regard to cost. Our result supports the notion that self-organizing networks can be far from optimal, and thus active intervention by policymakers or managers is necessary to take full advantage of the potential benefits. Finally, with clear-cut suggestions about where to add new links and remove old ones according to potential benefits, this study sheds light on future policies with regard to evolutionary technology networks.