From a simple request, a wonderful thread has ensued on technology, policy and the market inbetween. I was on vacation while it grew, so let me capture the thread before making some points... Cory emphasizes policy Kevin intertwingles technology Jason drives by technology danah drives by demand Wendy is concerned with the effects of technology But, Kevin did also say   Joi drives by supply Policymaking Happens Policymaking happens, sometimes enshrined in code, by public, private and non-profit institutions. All three sectors compete in providing policy and code. Code is global, policy increasingly local. Social norms pull, like my aggregator, not push decisions in policy nor code. The market, with enough choice, has always decided policy. What's new is users as developers are increasingly forming policy. What's new is falling search costs and switching costs as low as click allow rapid abandonment of bad policy. What's new is the cost of feedback, expressing demand, group forming and networking is dramatically falling to drive powerful emergent patterns. Through localized decisions of feedback we can put issues through our social filter, deliberate upon policy, allow social norms to emerge and take collective action at a larger scale. In my previous post I asserted that if a tool weakens social capital more than it strengthens it, its doomed from the start. Will Davies correctly pointed out the opposite is usually the case: Doesn’t recent (and maybe less recent) technological history suggest that tools are successful precisely when they relieve us of our reliance on one another, and foster independence?. This is true for such things as telecom and transport, and a good warning, but doesn't apply for most forms of social software, where group forming and user controlled networks enhance social capital. The best a toolmaker can do is be part of the feedback cycle, design for adaptability, and where possible, provide the act of making to users themselves. Making doesn't have to be traditional code it can be user control in the case of social networking services or information architecture in the case of wikis and other social software. The greatest cost of any business is change. Think through the policy implications you can of the code you write, but specifically give users control where you can and don't hard code policy too early. You can't predict all uses, but you can try to find a balance between openness/control, identity/anonmity, centralization/decentralization, explicit/implicit and other key issues. The most challenging axis of design that effects the social compact between toolmaker and tooluser is between openness and control. Its a complex issue, so here's some food for thought... Traditionally, tools have primarily given only control to toolmakers and have been open for use. Tools that are restrictive for use tend to come at a cost of growth although they realize efficiencies in providing them. Openness for users allows growth and access, but with negative externalities such as spam and allowing third parties to make private details transparent. An almost inevitable reaction is a reassertion of greater control for makers allows them to curb abuses. This change in policy is initially driven by users and if the vendor doesn't implement it other sectors will impose it (read: regulation). Another approach would be to turn this model on its head: put control in users hands and make openness part of being a maker, but that's a subject of a much longer post. For example, a social networking service that gives users control to form their own networks provides value in the form of social capital to organizeers. By giving organizers the openness to promote them has the risk of the negative externalities of publicity. By not holding on to control as a service provider it avoids privacy concerns. This model is good for the growth of the service, but it does come at a cost of market education. Right now, with many social networking services, you don't get spam from the service but from its users (albeit in exceedingly convenient ways). We are still getting used to these new tools and setting the norms for their abuse. My guess is social norms are arising to punish what we collectively believe to be outside convention, service providers will provide some constraints in policy and the model that does more good than harm will prevail. When the network is the market it can sense when someone is selling something with negative policy. It can take coordinated action, be it by exit, adapting social norms or code itself (building or supporting alternatives). The market isn't just about demanding consumers, its about empowered users. Technology, code and policy are outcomes from user networks and how they are served.