Joined: Mar. 2005
Let's see if I'm following this. Professor Zhang made use of some properties of atoms and molecules to construct circuit components (logic gates?) for a computer processor, a good bit smaller than the ones currently in common use.
If that makes those atoms and molecules intelligent, then the use of diodes (is that the right word?) on circuit boards would make diodes intelligent, and the use of transistors in earlier computers would make the transistors intelligent, too. Not to mention the beads on an abacus, or the marks on a slide rule.
One diode, or one transistor, by itself? I see no benefit to applying the word "intelligence" to single components that way. Or even to a single neuron by itself, either.
Another thought here: the properties of atoms and molecules make them act in pretty much the same way in the same situation, or at least with fixed odds for each result if quantum stuff is involved. But one of the aspects of intelligence is that beings possessing it vary their behavior at times, sometimes unpredictably.
Granted, an evolving gene pool does share that one property with known intelligence, along with the property of keep a record of things that worked. Although, it doesn't keep a record of things that didn't work, and should therefore be avoided in the future, nor does it have an effective way of doing a massive rework of a feature that has become inefficient or risky. But that's the gene pool, not the individual organisms and especially not the individual genes or the DNA molecules.