Why I'm Skeptical about Nanotechnology
I believe that nanotechnology is possible, but not feasible. (That's why my blog is called infeasible.) The difference between possibility and feasibility is money. For example...
A typical CPU is about an inch square and a few millimeters thick. However, the vast majority of the cost of making the CPU goes into just a few very thin layers of silicon, silicon dioxide (that make up the transistors) and copper interconnects which is maybe 10 microns thick total. The rest of the packaging, substrate and wires cost pennies.
A typical CPU costs about $100.
This means, by my calculations, that 1 gallon of transistors would cost a whopping 58.6 million dollars! A swimming pool would run about half a trillion dollars.
Using a density of 5000 kg/m^3 for copper averaged with silicon I get almost $100,000 per ounce! Gold is about $650/ounce.
Now this isn't fair to the semiconductor industry since transistors aren't purchased by the gallon or pound. However, it nicely debunks the myth at the heart of Drexler's vision of nanotechnology: that small is cheap.
The universe doesn't want these tiny structures to exist and it will extract an enormous price during their construction. The universe (through the 2nd Law of Thermodynamics) wants energy to be spread out, not packed in small tight spaces, and all those interfaces have energies associated with them. This is why it is highly doubtful that nanostructured materials will ever be cheap compared to more garden-variety materials.
Most technologies flop because they just aren't worth the increased expense not because they've run into some physical limit. One can theoretically show enhanced material properties by designing the correct nanostructure, but will it be worth the added expense?
A typical CPU is about an inch square and a few millimeters thick. However, the vast majority of the cost of making the CPU goes into just a few very thin layers of silicon, silicon dioxide (that make up the transistors) and copper interconnects which is maybe 10 microns thick total. The rest of the packaging, substrate and wires cost pennies.
A typical CPU costs about $100.
This means, by my calculations, that 1 gallon of transistors would cost a whopping 58.6 million dollars! A swimming pool would run about half a trillion dollars.
Using a density of 5000 kg/m^3 for copper averaged with silicon I get almost $100,000 per ounce! Gold is about $650/ounce.
Now this isn't fair to the semiconductor industry since transistors aren't purchased by the gallon or pound. However, it nicely debunks the myth at the heart of Drexler's vision of nanotechnology: that small is cheap.
The universe doesn't want these tiny structures to exist and it will extract an enormous price during their construction. The universe (through the 2nd Law of Thermodynamics) wants energy to be spread out, not packed in small tight spaces, and all those interfaces have energies associated with them. This is why it is highly doubtful that nanostructured materials will ever be cheap compared to more garden-variety materials.
Most technologies flop because they just aren't worth the increased expense not because they've run into some physical limit. One can theoretically show enhanced material properties by designing the correct nanostructure, but will it be worth the added expense?
I'm not a Drexler/Kurzweil apologize, but this article doesn't make sense to me. Sure, I get the concept that chips are expensive, but living cells are just as small and more complex, yet they are plentiful and cheap.
I think the expense is not inherent in the complexity or size but in the production mechanism. Once (if?) we learn how to move beyond chip factories to some inexpensive means of production that will change.
Reply to this
There is nothing cheap about cells that will manufacture exactly what you want them to when you want them to - cell like that are as expensive as magic beans.
Cells are a lot more complex than microchips not less. No one - no scientist or engineer - has yet managed to produce a cell without first having another cell to grow it from - no one yet knows how to - clearly that is not the case with microchips that are designed from the logic circuits upward.
If you don't care what cells make, or how much of it, or if you get to own it, or if you can actually make use it, then sure cells are as cheap to watch as the sunset and more plentiful than the grass. But how cheap is a thoroughbred horse, or a truffle, or a child ? How cheap are drugs produced by cultured cells like Epo?
Cells do not produce inorganic materials they only produce organic materials (ie they are extremely limited in the range of elements from the periodic table that they can "manufacture" with). And, relying on water chemistry as they do, they are also limited in the range at which they will work like cells and do their assembling and replication. Cells rely on the COMBINED liquid AND electrostatic AND particular-small-scale properties of water moelcules in relation to the other biomolecules to do their replicating and to make natural molecular machines out of biomolecules like proteins.
Water is the only molecule that acts just exactly like water. And for cells there is absolutely no substitute for something that acts just exactly like water. Its the water molecules acting like extremely small fingers proding and poking at proteins as they bounce around jostling them to wrap around their water-avoiding bits to take on a certain shape that make life. There are no other "fingers" small enough, and charged or "sticky" or repulsive enough, to replace the water molecules in acting like water molecules. And water molecules don't even act like water molecules - they don't bounce around as separate molecules under Brownian motion - if you change the temperature away from a certain range.
There is a world of difference between cells and microchips.
Reply to this