Answers to those questions will define the CIO's supply chain design and implementation agenda for the next decade. The "cost-effective" supply chain is a seductive trap. CIOs who invest as if the goal of their supply chain infrastructures is to ensure least-cost design and delivery of just-in-time ingredients are either short-sighted fools or long-term incompetents. China's rise as a global sourcing superpower requires CIOs to fundamentally rethink what kind of transparency, visibility and accountability they want to build into their supply chain.

Why? Because it's far too easy to design supply chains that get the immediate job done at the price of irreparably altering the balance of power between the multinational customer and Chinese vendor. George Bernard Shaw once defined success as the ability to take the path of maximum advantage rather than the path of least resistance. By that measure, too many global supply chains are miserable failures.

How so? Listen and learn from the Chinese. Companies will increasingly have to decide what information they wish to conceal or reveal to their links on their supply chains.

Determining the appropriate conceal/reveal ratio is a calculation that, frankly, is more difficult than most organizations are willing to admit. Legal, manufacturing, sales and finance may each have different notions of what data is appropriate - or necessary - to share. If your supplier asks for financial information, sales projections or manufacturing batch schedules as part of its efforts to be a faster, cheaper source for you, just how accommodating are you prepared to be? Does IT provide the necessary data in an integrated ERP fashion? On an as-needed data retrieval basis via the Net? Or through souped-up XML process drivers?

What complicates these questions even further is the inarguable reality that there are "grey markets" in supply chains. The Chinese understand a lot about grey markets. One Chinese manufacturing head who works closely with his IT chief tells me that his company builds a dynamic database for each of its major accounts and uses it to make a business case for tighter links to its customers' networks. The executive's goal is to become a design partner - not just a supplier - to his key customers. He believes the best way to do that is to go deeper into the client company. He's not alone.

If you're the CIO of the client company, how does this info-intimacy play? Do you want to encourage informal interactions? Or does the prospect of having your supply chain management support half a dozen data-sharing "grey markets" increase your level of risk? Do you just say: "That's not my responsibility. I just make sure the infrastructure works?"

Your answers around the conceal/reveal ratio would no doubt be influenced by the question of whether your company wants to collaborate with your customer - or supplier. Or whether your customer wants to commoditize your company's product. Collaboration likely means more revealing; commoditization means more concealing. After all, you don't want to give your customer the ingredients that turn you into just another supplier. But if you're trying to commoditize a key supplier's offering, then you're looking for more information that would make it easier for your company to switch suppliers.

In other words, what role should IT play in helping reduce dependence on your suppliers while increasing the dependence of your customers? I don't know the answer to that question. But I do know that every single manufacturer I've met in Beijing thinks along those dimensions. RFID tags are a technical means to a business end. The conceal/reveal ratio and the collaborate/commoditize axis, on the other hand, are key business criteria to determine technical investments.

Observing the evolution of this debate in China - and learning how this supply chain superpower implements answers to those questions - represents the best way for you to discuss the future of the supply chain in your own company.

Michael Schrage is co-director of the MIT Media Lab's eMarkets Initiative