My experience in law, however, says otherwise. From what I see and hear, for every high-tech entrepreneur that makes good, there are nine hundred that fail. Another ninety-nine make a frugal living and would have been better off if they had gotten a regular job, leaving that one out of a thousand (at best) who makes it good. And, most times, this one out of a thousand only did so because professional help was at hand.
And the owners of that nine hundred have probably taken out a big loan, bought lots of equipment, and, when their dreams didn't materialize, got stuck with large debts. So they get sued, and, quite often, have to go into hiding. I've seen this happen even with seasoned professionals, and, if they can go down, think of what can happen to university dropouts.
And, while I'm at it, don't think you can pull a fast one. You may have learned in your law course that a person can be protected by incorporating (limited liability). That's true, but as a shareholder in a startup, you would probably be the director and corporate officer, and, under some circumstances, you would be liable in those capacities. Not only that, the creditors will want personal guarantees - they've been in the business a long time, and they know all the tricks. In fact, they may even want guarantees from your family, so that, if you can't pay, they can get their money from other people. In other words, if you go down, you'll drag your loved ones with you.
So I would suggest a slower approach. First, graduate from university or college, if not in computer science and/or engineering, then with enough courses to get a minor. Then get a job in the relevant area. In all other industries, I would say go for the experience, not the cash. But, in the high-tech field, there seem to be a lot of startups (those nine hundred and ninety-nine) that will offer dazzling visions and offers of stock options and a glorious future. The owners will say that the company has no money now, but, when it goes for an IPO, everybody will be rich, so don't worry about a paycheck. Don't believe it. Cash is king, and, if a startup (or any other company) can't put the money up front, then it probably doesn't have adequate financing. Working from a garage to cut costs is one thing; getting free labor is another.
So the best thing is to look for a good job in an established company - as a general rule, the bigger, the better. Work for two years as a programmer or the appropriate entry-level position, so you learn the techniques of the trade. Then work for another two years in design or some other senior level. And finally, after you're familiar with the skills of software engineering, do two more years in management, so you learn how to run a business.
And, while you're at it, look around. Read the trade magazines, mix with the people in the business. Ask questions, find out who's good and who's not, know what's good and what isn't. Check the newsgroups and follow the threads. Always learn.
You can see that I believe in the slow and steady approach. But that's usually the formula for success - aside from the lucky few, there are no quick routes to success. If you can be one of those exceptions, go for it. But, if you fail, you'll get a letter from one of my colleagues, saying that you have unpaid bills, and that, if you don't pay up soon, you'll be sued. And that won't be a pleasant experience, trust me - I've sued a lot of people, and I've seen what it can do to them.
So, if you want to create a startup, go for it. Just make sure you know what you're doing.