The Wall St. Journal has an interesting op-ed by Nansen Saleri disputing the "Peak Oil" myth. Fundamentally, he points out that the theory is seriously flawed due to the inherent difficulties in forecasting the four questionable assumptions for which the theory has to define answers.
"The trivial becomes far more complex because the four factors -- resources in place (how many barrels initially underground), recovery efficiency (what percentage is ultimately recoverable), rate of consumption, and state of depletion at peak (how empty is the global tank when decline kicks in) -- are inherently uncertain."
Despite short-term price fluctuations brought on by speculative market forces, (primarily due to political factors rather than supply issues) there are not only huge reserves available, we are drastically improving our ability to recover them, as well as develop new unconventional sources. In fact, the price fluctuations are greatly assisting our ability to do so, giving national oil firms in particular both the incentive and the ability to make technological improvements to improve recovery methods.
Saleri correctly states that we've used around 1 trillion barrels of oil - out of around 12 to 16 trillion barrels of total supply (about half conventional, half not). In addition, the traditional figure of recoverable petroleum from a given well of 33 percent is steadily improving as new technological methods are being brought to bear, and may eventually double. Increasing the amount of recoverable oil even 10% would result in an additional 50 years of supply at current consumption levels. Doing the math, even without a increase in recoverable conventional supplies, we're used only about a quarter to a sixth of recoverable conventional supplies.
This brings the question of how much oil will we need over the coming years. Current consumption is 86 million barrels per day, with consumption forcasts increasing this figure up to 2% per year, leading to a figure of up to 117 million barrels in the next decade. In the event of any economic slowdown, this figure would likely be much less. "Peak Oil" theory holds that the tipping point will occur when 50% of the recoverable oil has been consumed, but with modern extraction methods, this figure is likely to reach a higher figure - and even a 5% increase in total recovery would delay the onset of the "panic" by 20-25 years.
All of this is not to say that research work on alternative sources of energy is misguided or ill founded, simply that the pessimism and alarmism found in some quarters is certainly not justified - at least not at present. Much like the almost crazed religious obesssion in the global warming debate (and I'm fairly certain there is a large degree of overlap between the two schools) the "Peak Oilers" are simply convinced that the end of civilization is coming soon. In my view, a gradual transistion from petroleum will almost certainly take place over the coming century for a variety of technical and other factors, and I'm quite optimistic the forces of civilization and innovation will almost certainly save us from the "impending catastrophe".