Humans are astonishingly efficient machines. The human brain runs on an average metabolic power of just 12 watts ... there are flashlights that use more power than that. (Meanwhile, a high-end PC will suck down hundreds of watts, while being way dumber than I am.)
(And chemical energy is way more energy-dense than even the best batteries ... an iPhone's battery can store just two and a half Skittles of energy - or a quarter-teaspoon of butter - and it takes way longer to charge than it takes me to eat two and a half Skittles)
Biological muscle is also an astonishingly good actuator; the main reason we've had trouble developing effective prosthetics is that no technology currently in existence is capable of matching the strength, dexterity, compactness, lightness, and speed of human muscle. (Certainly there are ways of being better at a few of those things - but not all of them at once.)
We are astonishingly efficient, flexible, and adaptable devices made of intricately complex self-repairing nanotechnology. As eager as I am to abandon this meat prison and enjoy the Glorious Transhumanist Future:tm:, it's going to be a long time before we can do better than meat.
@SenorOblong Phages, retroviruses, and wriggly little stem cell critters with an 'Eat and Replace' method of operation, and Plants that manufacture all the chemicals and pharma precursors we need.
The future is green and wiggly.
@Motodrachen Honestly, I think it's more likely that metal-air batteries of some variety or other will wind up being more efficient (when combined with the fact that electrical motors are like 90% efficient vs. 20-50% efficient for combustion engines and that electrical transmissions are way lighter than mechanical transmissions) and/or that electromagnetic wireless power transmission will become good enough to allow us to largely replace combustion engines.
@Motodrachen My personal sci-fi hobbyhorse is combining electrosynthetic bacteria (and the fact that cellular metabolism is more-or-less driven by voltage gradients and electron transport) to directly drive pseudocells off of electrical current.
Photosynthesis is only about 3-6% efficient; if we could drive biological processes directly off of solar panels, we'd already be way ahead.