The nightmare sickens me. A small child trusts a man to protect her, take care of her, and shield her from harm. The man, for incomprehensible and useless reasons, neglects her unto death. For me, when I think of Germanwings flight 9525, I am haunted by the photo of a single tiny girl, taken in the last days of life; the obliteration of that perfect life’s potential. Waste, tragedy, evil.

In the days that followed the crash, the blogosphere, media talking heads and airline safety experts, obsessed about preventing another murder by pilot. Always put two people in the cockpit. Three people, one with a gun. Give them all have guns. Remote technology on the cockpit door. Psychiatric testing, yearly. Monthly? Weekly? Clever, expensive workarounds to “fix” the capacity of the human mind to stray from rational. However, it occurs to me that the answer is more obvious, if more threatening. An empty cockpit. No pilots at all.

A computer would not get depressed. A computer would not get angry. A computer would not get tired. A computer would not forget or get confused. A computer would not develop the aberrant logic for which killing is an expression. A computer would not fly a plane into the side of the French Alps.

A computer would also not have flown a plane into the World Trade Center North Tower on September 11, 2001. Or the South Tower. Or the Pentagon. Or have to be forced into a catastrophic crash in a Pennsylvania field. Every plane would take off, fly and land perfectly.   It would not matter of someone’s wife leaving him, failing health, toxic medication or some fanatical calling requiring the blood sacrifice of innocents.

It may be that the only rational conclusion is that men cannot be trusted with complex technology on which depend the lives of other men. And women. Or children. Or a tiny girl for whom I weep.

Put in the context of the Institute of Medicine (IOM) 1999 report, To Err is Human, the 150 people killed in 9525 are only symbols, footnotes. When it comes to avoidable death by human failure, the American health system overshadows all the airplane deaths in history, including in war. As many as 98,001 preventable deaths per year, by medical error. That is 269 per day. 11.19 every hour. One every 5 minute, 22 seconds. That is an A320 Airbus, twice, every day, for always.

While the IOM study has been extensively debated, the concept of such needless human waste terrifies patients, doctors, and health care planners. It is the focus of innumerable projects, Lean and Six Sigma analyses, and an expanding culture of “quality.” The entire medical establishment, and even those not of “the establishment,” is struggling to reduce this ongoing incomprehensible tragedy to “acceptable loss.”

Nonetheless, we must not forget: The airline industry nearly perfected Six Sigma. They already get quality right. The lesson may be that that there is a limit to making man better. Perhaps medical men cannot be trusted with complex technology on which depend the lives of other men. Women. Children. Or, perfect little girls.

Of course, health care is nowhere as “simple” as flying a machine through a series of predictable and repetitive maneuvers. There is much more than terrain and weather to consider. Having practiced medicine for 35 years, I know each day and each patient are different.

However, I also understand that medical technology is incredibly complex and is so rapidly changing that no human can possibly keep up. I know doctors get tired, physicians are flawed and, while they rarely, if ever, are motivated by the twisted logic by which a pilot kills his passengers, they nonetheless make mistakes by which their patients suffer and occasionally die.

What does this mean for the future of medical care? What is the role of the doctor? The computer? The empowered patient? Do we rely on software, perhaps apps in smartphones, to order tests, make diagnoses, prescribe care? Will robots perform intimate exams? Do machines in the back of pharmacies do surgery? Can artificial intelligence lead research, produce innovation, inspire leadership? If so, who or what do they lead? Who educates families in the middle of dysfunction, crisis, and emotional distress? Who sits at cancer’s bedside, hand on shoulder, listening, providing courage, relieving fear, giving hope?

We have a conundrum. While we may be repulsed by the idea that doctors might go the way of travel agents, transcriptionists or bookstores, 100,000 deaths each year in a nation with arguably the world’s most advanced health care system is unacceptable. Around the world, over time, millions die from preventable error. Hardware, software, and big data, could be the solution. Myriad lives might be saved. Nonetheless, as the doctor exits, replaced by the wall-mounted camera, sensors on the exam table and a monitor on the desk, what does it mean for the humanity left behind?

P.S. As I complete this piece, I come upon the story of Mr. Iftikhar Hussain, 64 years old, from Chicago, Illinois. It seems he went for a drive on Sunday, to visit the family in Indiana. Carefully following GPS instructions given by the computer in his 2014 Nissan Sentra, he turned onto a ramp to a bridge over the Indiana and Ship Canal. Iftikhar deftly swerved around several bright orange barrels, which almost, but not completely, blocked the road ahead. The car burst into flames, and his wife Zohra died as their car plummeted off the bridge, which had been demolished several months before. There are some fatal errors during which man and machine collude.

James C. Salwitz is an oncologist who blogs at Sunrise Rounds.