You need something like half of the Sahara to produce the global primary energy consumption from solar panels alone. [1] ( Or the next few largest deserts combined.) So this is certainly not easy, but it is on the scale of doable. On the other hand, there are reasons why no one has ever tried to finance a nuclear power plant out of his own pockets, it is basically a bet on the interest rate of the next 30 years. ( And besides, we see in Germany that nuclear power gets slaughtered in the market when there is too much renewable energy. [2])
[1] 100 TW primary energy, 100 W/m^2 for solar irradiation on earth and 20% efficiency of the solar cells. Then you need 5e6 m^2.
An article by MIT claims 14TW - http://web.mit.edu/newsoffice/2011/energy-scale-part1-1024.h... - and I've seen figures from 12TW to 20TW bandied around and I tend to use 18TW when running the numbers, but I have never seen an estimate as high as 100TW. Where's that figure from?
Ups, good catch. It is from a old blog post from myself [1]. The reasoning was 10 billion people, with first world energy needs ( 10 kW). The alternative to use this number is, that you get a safety factor of 6 to account for using solar as baseload, weather and everything else that dissipates energy.
Edit: Additionally, the power of solar panels used is for England, not the Sahara. The purpose of the posting was essentially a back of the envelope calculation with pessimistic assumptions.
You don't need to get everyone in the world to be consuming more energy than the average European to have converted most energy to solar though. I am all for using fudge-factors to allow for rough data, but I don't see the point in starting a calculation with completely fictional data when there is data available.
[1] 100 TW primary energy, 100 W/m^2 for solar irradiation on earth and 20% efficiency of the solar cells. Then you need 5e6 m^2.
[2] http://www.bloomberg.com/news/2010-04-22/windmill-boom-curbs...