Reasonable people say they'd like to tax or regulate carbon, but alas it's politically impossible. They invoke President Bill Clinton's humiliating failure to secure an energy tax in 1993. They declare that carbon taxes or regulations would cripple the economy.

These reasonable defeatists should meet Vinod Khosla, the Silicon Valley venture capitalist who bet big and early on Google and Amazon.

Khosla's current bet is on next-generation ethanol. He believes, with all the passion of a techno-evangelist, that we can get most of our vehicle fuel from the Midwest rather than the Middle East, and we can do so simply by growing it. He shows doubters a photograph of a bamboo-like crop that sprouts 11 feet in just one year. If South Dakota were planted with this stuff, our dependence on Saudi fundamentalists would fall -- and so would our output of climate-warming carbon.

Khosla has made a fortune betting on technology, and he's invested tens of millions of his own dollars in the futuristic ethanol that will replace today's corn-based version. Yet he's the first to say that we can't address climate change with technology alone. Government must deliver the right tax and regulatory fixes to persuade people to adopt ethanol.

So the first lesson that Khosla teaches is that politicians have to act. But the second lesson is more pleasant: The government fixes that Khosla seeks are not actually so burdensome.

Khosla wants government to require auto companies to make more flex-fuel cars that run on gasoline or ethanol. Well, thousands of flex-fuel cars are sold in Brazil, and they're barely more expensive than ordinary ones. Khosla wants government to require big gasoline distributors to install ethanol pumps at a tenth of their gas stations. Well, it costs less than $50,000 to convert a pump. Assuming 20,000 conversions, this is a $1 billion problem -- a flyspeck on a $12 trillion economy.

The other government policy that Khosla wants is a reform of the ethanol subsidy. At the moment, the subsidy is worth 51 cents per gallon, but Khosla proposes that it be lower when oil prices are high and vice versa. By switching to a variable subsidy, the government would insure ethanol investors against the danger that OPEC might cut prices to drive them out of business. Subsidies would rise and fall, but the long-run cost of this reform would be about zero.

Khosla's enthusiasm for next-generation ethanol may turn out to be overblown, though his track record would suggest otherwise. His variable ethanol subsidy is less sensible than a variable carbon tax, which would spur hybrid cars, hydrogen cars and other carbon-cutting progress, rather than just ethanol. But Khosla's pitch illuminates a little-known fact. Relatively cheap tax and regulatory fixes can trigger substantial cuts in carbon emissions.

When scientists and economists estimate the cost of stabilizing carbon emissions, their conclusions tend to support Khosla's claim that it's not actually expensive. In 2004, for example, the U.S. government's Energy Information Administration analyzed a carbon-cutting plan advanced by Sens. John McCain and Joe Lieberman, which aimed to stabilize greenhouse emissions. The energy administration estimated that reaching this target would cause U.S. GDP to be 0.4 percent less than it would otherwise have been in 2028. Since GDP was projected to grow by 90 percent between the time of the study and that year, this meant that the nation could address climate change and still experience growth of 89.6 percent over the period.

In 2001 the Intergovernmental Panel on Climate Change, the most prestigious authority in the field, carried out a similar exercise . It calculated that stabilizing carbon emissions at an acceptable level -- defined as slightly higher than today's -- would cause world GDP to be 4 percent lower than it would otherwise have been in 2050. Again, that is a modest cost -- roughly one year of decent growth for the world economy.

So the reasonable defeatists on climate change turn out to be unreasonably pessimistic. The cost of cutting carbon emissions is not actually crippling. And if that's not enough to convince doubters that a responsible policy is politically viable, consider the fact of Khosla's trip last week to Washington.

Like most Silicon Valley digerati, Khosla spent years ignoring politics. But last week he was bouncing between Senate offices; he's pitched the White House on his ideas; he's backing a California ballot initiative that would raise royalties for oil companies. In short, Khosla is part of an emerging lobby for emerging fuels -- a lobby that includes companies such as General Electric, which has green technologies to sell, as well as Archer Daniels Midland, an agribusiness giant that knows how to work Washington.

In the past month, I've spoken with congressional staffers, White House veterans and the boss of General Motors, and all of them agree that serious carbon policy is not on the horizon. But sooner or later logic will prevail. The folks who said it was impossible are in danger of losing face -- that, and maybe money.