In 2016, I became the lucky parent of a newborn who slept horribly. Of course, this meant that my wife and I slept horribly, too. We rested in small snatches and were constantly irritable. We were a mess.

As a result, I became consumed with the idea of minimizing my need for sleep as much as possible. I had always required less sleep than my wife, but I thought that if I could just find some clever solution, some trick or tool, I might be cured of this time suck forever. I wanted to hack my need for sleep.

Eventually, I researched polyphasic sleep, a trend among the kind of people who quantify every aspect of their nutrient intake. Taken to its extreme, the practice promised the magic I was looking for: Simply sleep for 20 minutes or so every few hours, and eventually you’ll only need two or three hours of sleep a day. Sleep would be conquered! In its place, productive bliss.

I never ended up attempting polyphasic sleep. Its daunting requirements seemed destined to interfere with any semblance of normal family life. But the siren song of the quick, easy fix through a simple behavioral change or chemical consumed continues to appeal to me, as it does to many others: In Silicon Valley, this subset of biohacking is as strong as ever. Often coupled with its pharmacological sibling of nootropics (chemicals for cognitive enhancement), this trend of attempting to reengineer and overclock one’s physiology promises to make your body faster, stronger, and better in nearly magical ways.

Often going hand in hand with efforts to quantify the body and behaviors through all manners of technology, this hacking ethos relies on the idea that if people can just collect more data to better understand themselves, perhaps they can engineer themselves to perfection. We can hack our technologies, and even our societies, so why not ourselves?

Alas, things are not so straightforward. While there are many similarities between the biological and the computational, biological systems are complex on an entirely different level. They have evolved over millions of years, with a great deal of feedback and a mind-bogglingly large number of interacting components. And, in many ways, they’re already well-tuned and optimized.

Biology is a game of trade-offs. Any change in an interconnected system can yield both positives and negatives. Ways around sleep, for instance, come with costs. Taken to their extreme, practices like polyphasic sleep can take a month of adapting before you stop feeling like a zombie, and can severely mess with your social life. Caffeine, which more than 50 percent of Americans consume daily in coffee, is no panacea. Coffee makes you feel awake, but it can make you jittery, with the occasional sensation of your eyes bugging out from your skull. It makes you poop. It can even cause pretty strong symptoms during caffeine withdrawal, like headaches or irritability.

When it comes to hacker types parachuting into biology, especially for the purpose of improving the human body, failing to account for the inherent complexity of biology can mean failing to recognize that a messy system might not be easily modified in the way they want or expect. Fasting—which is often associated with increasing longevity—or some chemical of the week, whether hyped for improving concentration or weight loss, might help, but approaches like these are often feeble attempts to fit a highly nonlinear system with a single line.

Moreover, it’s hard to truly know what the many apparently redundant systems within our bodies have been optimized for. One alternative approach to handling all of this biological complexity is to use general rules of thumb. Comparatively simple guidelines have sometimes been passed down from generation to generation, such as adhering to traditional diets like the Mediterranean or Japanese. In one study, four basic lifestyle factors—smoking status, BMI, diet, and physical activity—were found to reduce the onset of major chronic diseases, including cancer and diabetes, by a massive amount. This simple, moderate kind of approach is also seen in the elegant dictum of the food journalist Michael Pollan, who advises, “Eat food. Not too much. Mostly plants,” or the haiku for how to run a fast marathon, from the Mayo Clinic anesthesiologist Michael Joyner: “Run a lot of miles / Some faster than your race pace / Rest once in a while.”

When it comes to biology and nutrition, while there is no doubt so much more to learn (and not everything that stands the test of time is necessarily correct), an incremental tinkering approach, tempered by a lot of humility, might be much more effective in the long term than the quick fixes of biohacking. Still, while these rules are straightforward, they’re not always the easiest to maintain. Perhaps that is why I was tempted—as are many others—by the search for ways of overcoming the trade-offs that are our evolutionary heritage.

Even in the face of the folly of biohacking, it needn’t be Luddism all the way down. There is still a place for massive data sets and complex models and machine learning and tech start-ups. But they’re only going to be effective in the long term when the massive complexity of biology is taken into account. For example, the chef Adam Melonas runs Chew, an R&D lab that is at the cutting edge of food science. The way he frames the company’s approach to food seems to strike the right balance between technological ambition and humility. Rather than claiming it’s going to reinvent food by turning eating into a problem of macro- and micronutrient intake, Chew aims to combine everyday ingredients with advanced technology and tries to come up with healthier and more sustainable products.

Or take Zymergen, one of several companies using artificial intelligence to speed up the scientific process. Its machines search for microbes that produce specific chemicals for such things as drugs or biofuels. This approach to science can be controversial: Sometimes, it’s not possible to immediately, fully understand how a microbe produces a specific chemical. But such complex systems demand an iterative understanding, as opposed to instantly and completely figuring out the biochemistry. They don’t operate under a misguided assumption that a system can always be understood and engineered into submission.

My son now sleeps better, so I’m no longer teetering on the edge of insanity. I’m still not the biggest fan of slumber—think of all the books or television you could catch up on while everyone else is asleep! But I’ve come to recognize that my search for a magical cure is not going to result in some simple fix. Hubris versus humility is an ancient story. As long as biohacking does not give in to its overconfident tendencies, it will be better able to internalize the best of the advances made about the human body and the subtleties of its operation.