The Internet makes exploiting labour something we can all do. In the old days, we used to use people as our access control mechanisms. We had butlers, servants, and cleaners to deal with the dirty work of answering the door to strangers and cleaning up from the random detritus that is life. Now, we are trying to write automated responses to automated problems like spam. Suppose instead we simply hired someone cheap from the Third World to do our dirty work for us, but over the Internet? What is actually cheaper?
What if we exploited labour on an industrial scale. We already have OffShore call centres and data entry centres strewn around the world. How many digital jobs are done cheapest in countries we never see but routinely recolonize?
But could we use those some exploitive practices in the West? All we would need is a contingent of cheap but stupid labour that is easily diverted to do our bidding. While minimum wage in most of our countries is too high, really the goal is to exploit other people's labour for less than its worth. The only way to do this is to download the capital expenditure on to the labourers themselves. Essentially, we want to make the labourers pay for themselves.
The easiest way to do this is through a competitive tendering process. Bidders compete with each other for a chance at a larger prize. Government frequently does this, forcing bidders to spend large sums of money with the hopes of landing a contract. In practice, government makes this easy by rotating contracts through each of the major bidders in order to ensure there is a semblance of democracy and a maximum amount of employment. The more capitalist automotive industry also forces bidders to spend money wooing a contract that may never come, but there the bidders may never see any money at all ever.
Of course, this process only gets you as much work as it will take to land a contract, not the actual desired end result. Also, companies cannot afford to blow the bank on a chance they will recouperate a bigger amount. As a result, most bidders only spend a small fraction of the potential winnings. What we need is a process that will result in a finished product.
Only people with nothing to lose and everything to gain would do something like this, and the easiest pool to tap are junior employees like students who have lots of time, no skills, and a desperate need for money. Rather than hire a junior employee and go through the expense of hiring, training, enculturing, and providing them with benefits, if you just need something simple done you can hold a contest. For a free laptop, redesign my website. While you're at it, you can have them usability test it for you as part of the contest rules. This way, rather than go through the hassle of trying several design approaches to find the most usable, you can just have a competition amongst several concurrently built designs and choose the best one. The cost to you is $3k rather than $25k. Many students will jump at the chance at a free laptop and some cred. What's more, they will actually feel happy and excited about doing it.
[ed: My roommate redesigned his call centre's website for $500 CAD, despite costing him about 80 hours to do it on top of 60 hour work weeks. He was quite excited to win the money. -- SunirShah]
Now, the problem with students is that they produce dubious quality. Another untapped labour market is the unemployed who have copious spare time and are desperate for an in at a company. But in a tight capital market, one wants to be extra careful not to hire a bad apple. It's risky to interview and hire a candidate you barely know. It's much better to hire someone whose results you can visibly see. The OpenSource developer community is rife with un(der)employed coders vying for a better résumé. If you offer a decent paying job for the best developers of the best projects, this creates a contest just like the free laptop above. You just need to sit back and wait to see what emerges, and then skim the cream off. Since the underemployed aren't negotiating from a point of strength, you can probably offer them a lower salary than the best employed developer as well.
[ed: What I was at Wizards of OS 3, the CTO of RedHat was showing over their recent fiscal results, which were growing impressively. He bragged to me about the value of their (well-known) strategy of skipping the pain of the hiring process by only hiring the leaders of the important OpenSource projects. -- SunirShah]
Further, the developers who lose the competition gain nothing. Despite their labour, they hold no real equity in the project. Only the social leader has any equity, and the sum total value of that equity was $0 + promise of exchange of more work for some money. There is no rationalized (and therefore accountable) way to distribute money back to contributors, say like is done with traditional corporations, particularly with equity sharing programs like stock options.
Thus, a TragedyOfTheCommons occurs. Competition between developers creates a race to the bottom, where developers increasingly trade large amounts of labour for free. Unlike normal capital practices where this labour is converted into assets, thereby generating wealth that frees labour to go onto other things, the lack of revenue as well as the lack of strong IntellectualProperty means this labour has been converted into nothing of economic value for the developer. All he or she is trading for is the potential of exchange of labour for revenue in the future.
The irony is that Marxist FreeSoftware rhetoric is very anti-capitalist, but that 'give it all away' rhetoric encourages the lack of accumulation of assets, making the labourers only more vulnerable to capitalism. Money will always find a way, since it is in fact necessary.
This is not to say that OpenSource is intrinsically vulnerable to this problem, but it is to say that the FreeSoftware rhetoric is unnecessarily harmful because it is dishonest. It's better to look at how to ExchangeValue by looking at OpenSourceAsPowerSharing, because that will make it clearer how to respond as a group against exploitive capitalists (i.e. unionize--with corporations and IntellectualProperty, FairSoftware, CommunitySource run by a NonProfit co-operative, or any other scheme that unites TheCollective).
But there's not just one company trying to optimize the ratio of work to cost; they're all doing it. So if this scenario holds, there is a competition on both sides: for jobs and for worthwhile labor. So there will be tensions that will guard against the unfettered downward spiral that you paint. Admittedly, that depends on how exhaustible the labor pool is.