Phil asks; "Robert ...add up the arable land, give us your best efficiency of conversion from your best net primary productivity - does it add up?" I'm not sure why arable land is necessary to generate electricity from solar. Indeed, I'm more inclined to use wasted space. My rambling 1960s sunny Southern California home could go "off the grid," generate more than it consumes with about 50 M^2 of cells unfortunately that would also cost about 400 times what the typical monthly enegy bill runs so it doesn't make sense. I won't go into the egregious public policies that forbid one from going off the grid as well. At the current rate I expect the costs to halve in about a decade so I expect the crossover point for western residential energy installations to occur 6-8 years from now. Even sooner if source point electricity utilities get smart and agree to subsidies in exchange for peak demand capacity sharing. Then there's the future of GM crops and animals. Drought resistant corn that can be used to produce some plastics comes to mind as a way to both reduce oil dependency and increase the amount of arable land by putting marginal areas to better use. Fuel cells do indeed currently need special metals but I thought that complaint was dealt with decades ago with the famous bet. In 1980, economist Julian Simon and biologist Paul Ehrlich decided to put their money where their predictions were. Ehrlich had been predicting massive shortages in various natural resources for decades, while Simon claimed natural resources were infinite. Simon offered Ehrlich a bet centered on the market price of metals. Ehrlich would pick a quantity of any five metals he liked worth $1,000 in 1980. If the 1990 value of the metals, after adjusting for inflation, was more than $1,000 (i.e. the metals became more scarce), Ehrlich would win. If, however, the value of the metals after inflation was less than $1,000 (i.e. the metals became less scare), Simon would win. The loser would mail the winner a check for the change in price. Ehrlich agreed to the bet and chose copper, chrome, nickel, tin and tungsten. By 1990, all five metal were below their real price level in 1970. Ehrlich lost the bet and sent Simon a check for $576.07. Prices of the metals chosen fell so much that Simon would have won the bet even if the prices hadn't been adjusted for inflation. Remember when we were going to run out of platinum because of catylitic converters for automobiles? Does anyone remember that the Club of Rome in 1975 said in no uncertain terms that we would be absolutely out of gold by 1997? That was before they even knew about the explosion in commercial consumption for electronics. Phil, you are right that we don't have enough land to say grow material for bio-diesel to replace what we now extract from the ground but we don't need to, we just need to price alternatives at the margin. At current prices we have some 37 years of proven extractable worldwide reserves. At the low end of the 35-45 year range that we've maintained since 1920. Interesting eh? In 1920 we had exactly the same world oil supply balance we have today. We've consumed more oil since 1970 than was proven to exist at that time. Someday the peak oil theory might be right about finiteness of the resource but even then the theory will be completely wrong as to the consequences.