Nanosystems, or Nanobots?
Goatchurch on the MundaneSF blog posted about his skepticism of molecular nanotechnology. He asked why we think a self-propelled miniature machine can exist, when we still don't have a robot that can vacuum the living room. (Given the existence of Roomba, that's another claim which, in itself, I would highly debate.) Here is the response I put there.
The failure of modern robotics is a software problem. But whether or not that is solvable is not relevant to nanotechnology, which rarely considers molecular robots. There is vast potential in nanotech products and materials that are not only dumb, but downright inert and permanently motionless. The potential of nanosystems which, while not motionless, are nevertheless dumb and sessile, is even more vast.
Look at this computer animation concept. Note that nothing in the assembly line depicted here involves individual nanorobots with independent self-propulsion, self-guidance or decision-making. Independence and intelligence are not required of molecular machinery. Don't confuse the "Universal Assembler/Disassembler" myth with the nanosystems that are actually being proposed.
One reason these systems will work so much better than existing technology is the simplicity of an atom in a vaccuum chamber in its interaction as a component. (One of the challenges of building nanosystems and nanofactories will be establishing a molecular vaccuum chamber to prevent interference from the churning of liquid and gaseous collisions.) If we have small enough fingers (which have only recently been developed) atoms act like snapping together Lego bricks, and are very reliable when placed with precision. The large-scale robots you refer to are made of unruly mountains of atoms, shaped with backhoes which, from the molecular level, are the size of skyscrapers.
A cloud of dumb molecular machines that would eradicate a tumor is likely to resemble an advanced form of drug to those who don't know what it is. It would likely be controlled and powered from outside the body with ultrasound or another directionally-focused emitter.
The failure of modern robotics is a software problem. But whether or not that is solvable is not relevant to nanotechnology, which rarely considers molecular robots. There is vast potential in nanotech products and materials that are not only dumb, but downright inert and permanently motionless. The potential of nanosystems which, while not motionless, are nevertheless dumb and sessile, is even more vast.
Look at this computer animation concept. Note that nothing in the assembly line depicted here involves individual nanorobots with independent self-propulsion, self-guidance or decision-making. Independence and intelligence are not required of molecular machinery. Don't confuse the "Universal Assembler/Disassembler" myth with the nanosystems that are actually being proposed.
One reason these systems will work so much better than existing technology is the simplicity of an atom in a vaccuum chamber in its interaction as a component. (One of the challenges of building nanosystems and nanofactories will be establishing a molecular vaccuum chamber to prevent interference from the churning of liquid and gaseous collisions.) If we have small enough fingers (which have only recently been developed) atoms act like snapping together Lego bricks, and are very reliable when placed with precision. The large-scale robots you refer to are made of unruly mountains of atoms, shaped with backhoes which, from the molecular level, are the size of skyscrapers.
A cloud of dumb molecular machines that would eradicate a tumor is likely to resemble an advanced form of drug to those who don't know what it is. It would likely be controlled and powered from outside the body with ultrasound or another directionally-focused emitter.