To start, I /am/ a scientist (PhD, molecular biology), and I've also taught. At the moment, I'm pure research, so I've been on multiple sides of this issue.
To really address the suggestion that I as a scientist isn't doing enough to educate the public is just an outright oversimplification of the issues involved.
As I see it, there are at least three major obstacles inherent in 'educating' the public about any particular scientific topic (molecular biology, computers, you name it)
1. Education: There's a lot of work involved in getting to know underlying concepts well enough to properly transfer ideas. I can talk about western blots, ELISAs, single-nucleotide polymorphisms and the effect of multiple genetic anomalies on various types of lung cancer as relates to smoking, but unless you understand why and how those things are important, any conclusions I give you, like 'an increase in these four or five SNPs statistically leads to a higher risk of lung cancer', you either have the choice of accepting what I say on faith, or ignoring it.
2. Apathy: If you don't know, do you want to learn? Most people who ask what I do want a nice easy answer, like "cancer research". They smile, say 'ooh, that must be hard' and go about their day. Any longer than a ten word answer, and I get the classic glazed-eyes look. You can't educate a populace that doesn't care.
3. Propaganda: This is related to 1 & 2, but is mostly a side-effect of 1. I recently got into a rather heated argument with someone over whether or not second-hand smoking is bad for you (yes, it is). Despite all the first-hand knowledge, research, and peer-reviewed data I could provide, the person in question chose to rely on biased politically-motivated think-tank reports and old, outdated, dis-proven studies funded by tobacco corporations instead of listening to my data and realistically weighing the available information. You can do all the outreach you want, but if the people don't want to pay attention, they won't.