While everyone is gripped by fear over whether they still have a job, the mining industry is quietly wondering if this time is different.
Having been spared every other technological advance, mining still operates much the same as it always has. You get the rights to a parcel of land and dig around until you find something. If you find something then you keep digging. When you're done you fill in the hole and move onto the next one.
Can artificial intelligence really upend such a simple model?
Bill Gates, Jeff Bezos, and some other luminaries of Silicon Valley believe it can. The plan is to not only map out the surface of the earth, but to chart out the concentrations of minerals at depth using predictive models and vast amounts of data.
This plan would practically eliminate the need for exploration, itself a $10 billion industry. Which seems dramatic until you consider global revenues from mining clocked in at $2.5 trillion last year. Exposing the entirety of earth's subterranean layer is in the realm of the discovery of fire. Yet somehow it sidesteps the greatest challenge of mining: getting permission.
In mining we call this a license to operate, as if it were a single piece of paper you received in the mail, like a driver's license. Wouldn't that be simple. You do, in fact, receive your permits in the mail. The only difference is there are hundreds of them and collecting them can take up to twenty years. Or never. When I first started in the industry I assumed this was a gross exaggeration. Maybe it was nostalgia from old timers. But as of this writing, I have yet to see any mine break ground on an accelerated time table.
Mining is a business where you quite literally throw money in the ground. And we accept that. The reality today however is that a deposit can be proven in mere months, while the following years are spent dancing through a series of obstacles set by government officials who wouldn't know a mine if they fell into one. All the data in the world analyzed by the most powerful predictive models can't hope to outsmart the caprices of modern day bureaucrats.
These problems aren't going away either. Most of us are painfully aware of this when we consider the housing sector, transportation networks, and any large project essential to our way of life. I grew up in a mining district, and it was not unheard of for an entire town to be moved for a mine site. A shaft was directly opposite my school, and another downtown. This was considered normal. Today, the mere concept is inconceivable.
Of course, no one wants a mine in their backyard. And over time that preference has strengthened from higher populations, organized citizen groups, and an internet which fosters activism over longer distances. Increasingly, no one can have a mine anywhere even if they wanted it.
Still, the problem has never been finding minerals. Every year thousands of new discoveries are published. But less than 1 in 10,000 of these discoveries ever become mines. Which is why minerals in the ground are next to worthless. They only have value if you can get them out.
With such a miserable rate of attrition, it would be a mistake to conclude the mining industry is closed to innovation. Not long ago, my company was approached by a firm promising to unlock the value of our deposit. If only we would lend them our data they would do the rest. And our expectations weren't high. Imagine our disappointment when we were told, "try twinning the holes." After a year the best they could come up with was try drilling all the same holes but a few inches to the left. It's even more disappointing when you consider they had access not just to our data, but all the mining data published over the last century.
While the triumph of big data and its analysis may appear inevitable, there is more than one proposal on the table. Deep sea and asteroid mining have recently attracted sizeable investments. And there are recycling initiatives which promise to eliminate mining altogether. All of these proposals have successful precedents in their favour: we know how to drill oil in the ocean, we can land on extraterrestrial bodies, and we already recycle all the lead used in car batteries. Yet whatever merits they possess, none hold the allure of artificial intelligence.
But let's say we do chart the surface and underground of the entire planet. What then?
Today, mineral supply and demand is predictable. And so are the prices we pay for the products that use them. This was not always the case. The gold rushes in Peru, and later Australia caused wild fluctuations in currency values and rippled through financial markets for decades. Less known is the story of aluminium which went from an adornment at the table of Napoleon to a kitchen material valued at twelve cents a pound.
What this new solution proposes would be even more dramatic. You could wake up one morning and find out the ground beneath you increased in value several thousand percent. And before you have a chance to celebrate you might learn that you don't own the rights. Someone else does. Since many countries draw a distinction between surface and underground land rights this could quickly turn ugly. Everyone acts civilized until they discover their neighbour lives on top of several billion dollars. On this, historical precedent leaves little to the imagination.
That probably won't happen, though.
As with most things, it is often the most likely explanation. That same outfit which proposed charting the entire earth and received Silicon Valley backing is now partnered with a mining company, exploring Greenland.
Perhaps they realized mining is best left in the hands of the miners. Much like we have learned to leave taxes to the accountants, and non-disclosure agreements to lawyers. You pay the accountant, the lawyer, and the geologist not just to do the job but to take full responsibility for the consequences that follow. It's true, we can automate anything. But it doesn't mean we should.
I'm not sure who is the best at finding deposits, but I know who has the best track record of getting them out of the ground.