If you want to understand what the current research says it means, a great book is Rethinking Autism by Lynn Waterhouse; see my review of it at http://intellectualizing.net/2013/12/17/book-rethinking-auti... and the book on Amazon http://www.amazon.com/Rethinking-Autism-Complexity-Lynn-Wate...

The "disease" analogy is grossly misleading because it makes you think it's a single "thing" with a single cause; the right analogy is to a dimensional trait (such as intelligence), or traits perhaps (such as both social/communication skills on the one hand, and repetitive behaviors on the other). Then think about many different conditions resulting in this trait, much as many different conditions result in intellectual disability (or intellectual superiority). As with intellectual disability, autism diagnosis is more of a semi-arbitrary cutoff on the scale, rather than a binary yes-or-no.

I say "semi-" arbitrary because diagnosis is an attempt to locate "where on this scale people start to really need extra help," but "arbitrary" because there isn't a strong scientific reason to put it at say 99.1 percentile vs 89.9 percentile.

As with variation in intelligence, autism appears to be mostly genetic, with some impact from environment. The underlying condition resulting in autistic traits varies a lot. Sometimes, the condition may have other consequences, such as seizures or intellectual disability. Other times, the condition is best thought of as simply an extreme personality type, or something like that.

Think about how various conditions resulting in ID can have accompanying issues (such as heart problems with Down syndrome). When someone says seizures or gut problems or whatever are "autism," that's like saying heart problems are intellectual disability. It doesn't make any sense. Autism is just the social/communication part, not everything that's up with someone.

If you work in tech, you know a lot of autistic people, you just maybe don't know it, and they often do not know it themselves. Some of them are quite successful and not "obvious"; others are extremely clear to anyone who knows what to look for. But most of them probably do have some level of real disability, or if nothing else did have some level of it as children.

For many of us the disability is primarily in personal life (friends, etc.), or historical childhood difficulty, and we are highly competent at work.

I did not imagine that the diagnostic label applied to me until my son was diagnosed; but now I realize that it does apply or very nearly apply to me and several people I know (but I have no idea how to bring it up). Especially when they are having kids, I wish I'd known earlier that autism can be kind of the way my family is and many people I know are, and that it isn't always this big scary thing.

The tech industry should be much better informed about autism given its prevalence in our coworkers and ourselves. Even the many who aren't diagnosable may be closer to autistic than to average.

When those of us who are around 40 were kids, the diagnostic rate for autism was about 1/1000 and it's now a bit above 1/100, higher for boys. So the delta between those is the number of undiagnosed adults walking around today, who would have been diagnosed by modern criteria. It's a lot of people.

Some number of those got a different dx in the past (such as intellectual disability), but the kinds who are now in tech probably got no diagnosis at all most of the time, other than "weird kid."