And that's not necessarily true. It assumes that everyone mines bitcoin for the purpose of obtaining bitcoin, not to support the network.
Like someone is intentionally donating money in hardware and electricity in order to support the project? How many dollars a month would you personally spend to do this? Maybe there's some rich people involved who have the money to spare? At the current cost of mining, how many dollars a month are are required to keep the economy going (I have no idea), and would it be possible to get that much as "charity" donations to support the bitcoin project (I have no idea either)?
Speculation has driven the cost of mining way up, but the speculation isn't inherent to mining. If the network was run entirely on a "charity" basis, the costs to mine would be much, much lower.
Mining involves some speculation about future difficulty and future exchange rates (since you have to pay for electricity in real money), but at least there are some fundamentals that you can build a model around. With simple calculations (that ignored transaction fees BTW) I was able to correctly predict when mining would and would not be profitable and mine accordingly.
I think the interesting thing the article points out is that bitcoin mining itself is inherently a form of speculation though -- you are counting on the value of the bitcoins you mine being at least as high as the money you had to expend to mine the bitcoins (hardwar, power, network access), or mining is a losing proposition.