1. Anything which solves problem for businesses, which can include programming. Do not, do not,
do not assume that being a programming demigod is required to be a high-class consultant or that is even your primary source of value.
There are any number of folks who consult on things on HN, on everything from optimizing warehouse processes to pen testing Java web applications with obscure wire protocols to SEOing websites for software companies. The common thread among those who make a lot of money is that their customers trust them to make them shedloads of money or avoid equivalent losses.
2) I once read an interview with a wealth management professional on the secret to getting rich off regular labor, and his answer was interesting: "Transactional work for rich people." To elaborate, if you're essentially getting paid a portion of the transaction, you want to work for people whose transactions are huge. Given that it takes virtually the same amount of work to sell a $2 million house and a $200,000 house, get the 7% off the pricier house. Given that it takes essentially the same amount of work to write a SEO strategy for a company that has $20 million a year in revenue and one that has $20,000 a year in revenue, do it for the bigger company.
I get consulting gigs primarily because folks ask me to consult for them. They typically know me from some combination of my blog, my Internet participation at watering holes like HN, or offline networking. (Conferences and HN meetups have been very good to me, and I enjoy the heck out of them.)
By the way: I know this is probably blow-your-mind territory right now, but I wish somebody would have told me when I was a college freshman since it would have saved several years: $100 is not a high consulting rate. People substantially less talented than you command multiples of it, for a variety of reasons. For example, the day after you graduate, you could be billed out at $300 / hour by one of the large consulting firms like McKinsey. That's probably a pathological example, but if you nail what McKinsey nails (giving clients the feeling that corporate money spent on them will ensure their personal career goals), you can also command rates that are substantially in excess of the numbers you are currently contemplating.