My dad tells me he remembers when they thought weather would be predicted better once we had computers, when space stations would generate medical and manufacturing breakthroughs. I remember a professor telling us that with the new 486 computers, we would get some really good econometric models of the macroeconomy soon (all that computing power!). But in fact, just eyeball the satellite picture of clouds and you can do as well as a sophisticated model, and predicting the economy is perhaps no better than it was in 1975 when people would intuit a simple autoregressive formula to GDP growth.

So I was reading The Politically Incorrect Guide to Science by Tom Bethell, and though some assertions I found unconvincing, others were quite fascinating. Take the idea that we thought we would crack the origin of life. Stanley Miller did an experiment showing you could create simple amino acids in an environment that seemed to mimic an early earth atmosphere. That same year, 1953, Crick and Watson discovered the structure of DNA. It seemed the essence of life itself would be revealed. But it didn't work out that way. Miller admitted later in life that his results weren't so clear, because the primordial soup created so many right and left handed organic molecules, the latter being toxic to most cells. Crick and Watson's DNA is proving far more complicated than thought, , in that at first we thought there were 150,000 genes, but now only 30,000, or only twice as many as a nematode.

But here's the problem. Humans have perhaps 300,000 proteins, so the method by which the cell creates protiens from genes is clearly not the original one gene --> one enzyme formula we originally thought. There is something else going on, some interaction among genes, or enzymes, that creates the essence of organisms.

This all highlights that in science, like in math, almost is almost nothing. Until you actually prove some interesting result, potential is very difficult to assess, and best viewed with skepticism.