• Google
    This Blog Web

October 2011

Sun Mon Tue Wed Thu Fri Sat
            1
2 3 4 5 6 7 8
9 10 11 12 13 14 15
16 17 18 19 20 21 22
23 24 25 26 27 28 29
30 31          

RSS Feed

Bookmark and Share

Email Feed



  • Powered by FeedBlitz

« "Molecular Programming" Approaches Molecular Manufacturing | Main | Self-Replicating Nano-Robots Now Possible »

April 28, 2010

TrackBack

TrackBack URL for this entry:
http://www.typepad.com/services/trackback/6a00d83451db8a69e20133ed070c3c970b

Listed below are links to weblogs that reference Scientific Advances Require Blind Luck:

Comments

Feed You can follow this conversation by subscribing to the comment feed for this post.

Dan S

Intresting idea. I have not thought this way before, but now I think you are exactly right.

Not only in science, but everywhere. People seems to reject new ideas based on some empirical "utility function" that includes credibility of the author, previous personal expirience, mood state, time of the day and a number of other factors.

And opinions on new ideas tend to form once and for long.

And yes, telling people that they reject ideas randomly will do no good. If you understand "utility function" of particular person you can exploit you knowledge to make him accept almost any idea you want. Otherwise, you are counting on blind luck.

Tom Craver

I think scientists tend to reject anything that isn't "obvious in retrospect" (i.e. once they hear of it, is it pretty obvious how one gets from the previous or current state of art to the new level?). Incrementally better results from modest changes to existing methods are accepted as worth looking into, reproducing, expanding upon, etc.

Building "anything" out of DNA is much more acceptable once someone has demonstrated building 3D boxes with controllable lids, which was more acceptable after someone demonstrated building interesting tiled objects, which was more acceptable after someone demonstarted linking fragments of synthetic DNA, which was more believable after someone developed the ability to synthesize and replicate arbitrary short DNA segments, etc.

Propose in the 1980's that it should be possible to build machines that build nanomachines that build anything from the atoms up, and with no obvious path in place, scientists will assume that the chain of required developments is long, and any stage of that has a chance of failure, and so the cummulative chance of failure approaches 100%. Which is true, for any single development path one might propose, but ignores the likelihood that there are a huge number of paths to take.

But the chances of any individual researcher betting his career on developing a correct path, in the face of so many wrong paths, is essentially nil. Rather than admit their self-interested conservativism, they throw out a likely sounding objection, that can't be proven wrong until someone else puts in the effort to develop a path that avoids the objection.

The difference between a 2nd rate and 3rd rate scientist is probably intellect - but the difference between a 1st rate and 2nd rate scientist lies in how daring they are.

Chris Phoenix

Tom, the impression I'm getting from the experimentalists here is that nothing is "obvious in retrospect." Any given idea might fail for some small random reason. The good ideas are the ones that take "only" a few years to develop.

So it makes sense to say "scientists reject anything that won't obviously work" but that equates to "Scientists reject anything."

BTW, can anyone tell why it takes so long for the "Post a comment" form to appear?

Chris

Tom Craver

I may be misunderstanding you, but again, I think it depends on how big a jump some development takes.

The "obvious in retrospect" rule of thumb is just meant to categorize the sort of work that is more likely to be taken on - i.e. something that a scientist can reasonably expect other scientists to quickly accept, once they've done the work. One can always find some scientist that will reject any speculation no matter how trivial - but smaller, less risky ideas and steps are more likely to find some scientist willing to take a chance on it, because there is a continuous low level pressure to make SOME sort of advance - publish or perish.

It may seem obvious that there must be some viable path to developing mature molecular manufacturing - but scientists aren't willing to bet their career on finding the right path, and if they feel pressured to bet their careers on that goal, they'll come up with rationalizations for why it isn't worth striving for.

Only when outsiders with a strong motivation put up both the financing AND the pressure for success does goal oriented science proceed at the rapid pace an outsider would naively expect. I.e. an Apollo or Manhattan project.

We came within an inch of getting that with the NNI, but those who politically initiated it didn't really understand what was required or lost their nerve in the face of resistence from scientists, while scientists saw a sweet funding source they could divert to fund ordinary low risk research by sticking "nano" in their research proposal.

Chris Phoenix, CRN

Tom, I think we may be more or less agreeing. Yes, if something is small enough, it will be more likely to be accepted.

There's another factor that we should consider: the development of research communities. A research community may develop a common body of knowledge of what works, and then continue to develop new results and successes within or slightly outside that area.

My impression is that, without a research community, it's rather hard for an idea - even a related idea - to get traction. Even an idea like DNA staples would have found it hard to get traction within the existing DNA community, unless Rothemund had demonstrated it himself first.

Chris

The comments to this entry are closed.