A human marvel that's hard to replicate

I’m a tool guy. I love tools, not for their own sake but for the way they make jobs easier. The right tool for the right job, as many a grandfather has counseled.

In my garage, I have tools I inherited from my father-in-law, some from my father (which likely means from my grandfather), a mini-pitchfork from my aunt that I use in gardening, and some power tools that were gifts from a daughter who understood well the right-tool adage. All have helped hugely over the years.

But through all the jobs I’ve completed with all the tools I’ve used, I’ve often remarked to myself that there is no tool quite like the human hand. It’s an absolute marvel.

For versatility and effectiveness, it has no equal. It is capable of an astonishing variety of tasks — lifting heavy objects, applying torque, threading filament through the eye of a needle, writing, playing piano, opening a jar, pushing a broom, painting, folding clothes, flipping a pancake, scratching an itch. The list is endless.

Aristotle called the hand the instrument of instruments. The great Greek physician Galen was reported to have found the complex system of muscles, ligaments and bones in the hand “amazing and indescribable.”

So I will admit to no small amount of pleasure that the makers of humanoid robots — the machines meant to replace us, or at least take over a lot of things we do at home and in the workplace — are finding it hard to make a truly helpful humanoid robot because they cannot replicate the human hand.

Elon Musk, whose Tesla company is developing the Optimus humanoid robot, seemed to put aside his customary bravado in an investor call in October, saying that “the more you study the human hand, the more incredible you realize the human hand is.” Designing a hand and forearm for Optimus, he admitted, “is an incredibly difficult engineering challenge.”

There’s a lot at stake. Musk said in November that Optimus might turn out to be “the biggest product of all time by far, bigger than cellphones, bigger than anything.” Jensen Huang, chief executive of chipmaker Nvidia and a man less given to hyperbole than Musk, said in June that “humanoid robotics is going to potentially be one of the largest industries ever.” After reports that Apple was looking into robots, a Morgan Stanley analysis last year found that the tech giant could earn as much as $133 billion a year from them by 2040.

With tantalizing advances in artificial intelligence, the gold rush is on. In 2024 alone, robotics startups attracted more than $7 billion in investments. But what good will a thinking robot be if it cannot lift the kitchen garbage out of its receptacle, tie the bag, unlock the back door, bring the bag out to the trash can, return, unfurl a new bag, shake it open, and put it into the receptacle? Or if it does not understand the different grips needed to pick up an egg without crushing it or hold onto a balloon being buffeted by Long Island’s winds?

Musk’s challenge lies in the numbers. The human hand has 27 bones, 27 joints, 34 muscles, more than 100 ligaments and tendons, and lots of blood vessels and nerves. The palm alone contains 17,000 touch receptors and free nerve endings which sense pressure, vibration and movement. The National Institutes of Health says that our fingers are bent and stretched about 25 million times in a lifetime.

All that intricate dexterity took eons to evolve. Modern robotics scientists are working against a much shorter time frame. They’re making progress. But it’s not like they’re trying to reinvent the wheel.

The hand isn’t just another tool. It’s the tool of tools. And the ones I have are good enough for me.

Columnist Michael Dobie is a retired member of the editorial board.

By Michael Dobie

michael.dobie1@newsday.commwdobie

Michael Dobie is a retired member of the Newsday editorial board.