Everyone agrees that we are running out of oil, what no one knows is how quickly. Pessimists say it could only be a few years until we are plunged into bitter darkness, causing a typhoon of war and famine to sweep the Earth.
Optimists think it will be many decades before this happens, so take your pick. When I say "no one knows" how long we've got, I really mean no one has got a clue, no matter how confident they might sound.
As Shell recently, uncomfortably, made plain, we don't even know if the reserves claimed by oil companies are real or fiction.
The latest stage of this corporate calamity saw Shell wipe 4.5 billion barrels from its "proven" reserves, meaning it's got a quarter less oil on hand than previously claimed.
I asked a few Wall Street types this week if it is really only Shell of which we can no longer be sure. To a man, they said that no other big oil companies will have to confess to overstating reserves, though this seemed to be a matter of faith rather than proof.
Faith is a big factor in the oil market. Saudia Arabia, for instance, claims to have 290 billion barrels in stock, though if anyone asks to actually inspect these supplies, the answer is: sure, as soon as green flying pigs jump over the moon.
Colin Campbell, a geologist who has worked for Texaco and BP, puts the Saudi's true reserves at less than 100 billion barrels, and predicts a "chronic long-term shortage" beginning in 2010.
At the moment, the big issue is not reserves but production. Demand is soaring, not least from China, and whether it can be met is far from clear.
In 1956, a Shell geologist called M King Hubbert predicted that oil production in America would peak in the early 1970s.
No one wanted to hear this, so the industry took the standard way out when faced with such a predicament. It laughed in derision.
Irritatingly for the oil men, Hubbert was right. US oil production peaked in 1970 and has been in steady decline ever since, which is why America has to import 60pc of its oil. That world oil production will also hit Hubbert's peak is inevitable - the big question is when.
Ken Deffeyes, whose book Hubbert's Peak: The Impending World Oil Shortage came out in 2001, has his bet on Thanksgiving Day (that's late November) 2005, though we won't know if he's right until well after then.
Deffeyes thought Shell was the least likely of the main players to have inflated its reserves, but says the bigger problem is countries, not companies. "Iran, Iraq, Saudi Arabia, they are all exaggerating," he says, both about what they've got in reserve and how much they could produce in a pinch.
Another who shares this view is Matt Simmons, who heads Simmons & Company, a Houston based energy-investment bank. Simmons is working on a book that is based on 40 years of industry documents and claims his findings are "truly scary".
Be clear that Matt Simmons is not one of those characters who actively wants us to run out of petrol because he gets such a kick from the idea of a car run on prune juice.
I don't know if he was wearing a cowboy hat and sucking on a cigar when I spoke to him, but it felt like it.
He points out that no big Saudi oil fields have been found in 30 years, and the biggest, Ghawar, is responsible for half of all Saudi oil production since 1948. "It is clear to me that we should be preparing for the day that Ghawar goes into a tailspin. We are not at all prepared for that," he said.
The CIA has been troubled about Saudi reserves for decades, which may be why Simmons was one of the contributors to Dick Cheney's secret energy task force meetings in 2001.
Simmons thinks that Shell "did the world an enormous favour by showing how fuzzy the idea of proven reserves is."
I asked the American Petroleum Institute if Shell could really be an isolated case. A spokesman said it "doesn't have sufficient information to form a view" (why the hell not, you might be asking) but referred me to the Petroleum Industry Research Foundation.
The PIRF doesn't have a press office and no one was very keen to give their name. It also declined to comment.
Although there will still be plenty of oil to go around when Hubbert does peak globally, that's the point at which production can no longer meet demand and oil prices leap. If you think it's expensive to fill up the motor now - just wait.
Oil fell well below $40 a barrel this week, but don't bet on it staying there. Simmons reckons that the correct price for oil so that demand is controlled while humankind comes up with another plan is $182 a barrel.
Simmons, Campbell and Deffeyes are not loved by the oil business, though they are being taken more seriously than ever before.
This week there was a Peak Oil conference in Berlin, a gathering of worrywarts wearing metaphorical "we are doomed" placards.
One unusual attendee was Fatih Birol, the chief economist of the International Energy Agency, who gave a speech saying that everything is fine, before admitting afterwards to a BBC reporter that everything is not. "This is not for the press," he said, after blurting out that the Saudis need to increase supply by 3m barrels a day to avert an oil crisis by the end of the year.
If they don't or they can't, there will be more at stake than the cost of the school run.