As long as I’ve been alive, I’ve always believed that intelligent life exists in more than one place in the universe. We are of course not alone. Whether that was a child’s fervent wish or the reasoned response of an adult faced with a near-infinite universe, the idea of intelligence as a natural consequence of life has always made sense to me.

Indeed, because we know life is essentially a higher expression of chemistry – that, given the right circumstances, the right set of chemicals will create DNA-based, self-replicating structures – it is almost insane to think life can’t exist elsewhere. Our galaxy alone contains an estimated 100 billion stars and is one of 100 billion estimated galaxies in the universe. The math almost guarantees that life exists elsewhere.

Scientifically, there are some logical problems with these beliefs. For a start, even if we earnestly believe that life must exist elsewhere, we see no evidence of it. Surely if life is a consequence of chemistry, and intelligence is a consequence of life, there must be some space-faring species that could have made contact by now? This is what is sometimes referred to as the Fermi Paradox: the apparent discrepancy between the high probability of life and the utter lack of its evidence.

Still, life very clearly exists on Earth and does so in abundance and variety. Once life gets a foothold, it seems clear, it can adapt and thrive. But is intelligence a natural consequence of life?

Here again, despite the multiplicity of life on Earth, we see evidence of only one form of higher intelligence. Why do we not see multiple, coexisting forms of intelligent life? Or else evidence that our one form of intelligent life out-competed another? Perhaps then, intelligence just one possible outcome of the evolutionary process. Like the swollen abdomens of honey pot ants, human intelligence may merely be a unique adaptation, instead of an inevitable next step.

Even if we can say that intelligent life is demonstrably possible, it’s also possible that the inefficiencies of the human brain are prohibitive to reproduce. As adaptations go, intelligence is a resource hog. Despite comprising only 2 percent of the body’s total weight, the brain demands about 20 percent of the body’s resting metabolic function. That means that if you burn 1,300 calories on a lazy Sunday, your brain sucked up 260 of those calories. (math helpfully provided by Scientific American)

So, brain power requires tremendous resources to maintain. Perhaps too much for intelligence to be common in the universe and even here on Earth, enough that training young brains cannot be separated from feeding young brains. Now that we find that 51% of American school children live in poverty, it should not surprise any of us that school performance in low-income neighborhoods is declining. Children who are either not eating or eating junk food with inadequate nutrition are being deprived of the precious resources that keep the gas-guzzling engine of learning running.

To the extent that our nation is interested in improving education, it’s worth keeping in mind that intelligence exists on a knife’s edge of impossibility because we have the wealth to feed it. If kids are going hungry, they’re going to fall behind. Simply raising testing standards only compounds the problem for students who cannot bring the brain power to bear that their richer neighbors can. Beating up on teachers may have some electoral appeal in certain quarters, but it won’t change the statistic staring us in the face: our kids’ brains are getting starved.

Watson, the IBM artificial intelligence supercomputer that defeated Jeopardy’s smartest contestants, ran on 90 IBM servers each requiring a megawatt of power. One Stanford scientist predicted that replicating a fully-functioning brain would take all the energy produced by a small hydroelectric plant. Intelligence is not efficient. Are we prepared to provide the fuel we need to actually improve on our educational system?