For the very smart data science geek who wants to avoid all kinds of social interactions (that is: having no boss, no employee, no colleague, no client, no customer, no contractors, no interaction with vendors etc.), there are a few options. All of them allows you to work from home. Some (when automated) allow you to not work at all. They can generate good complimentary income or (if scalable) great income.

In practice, these opportunities are not real jobs but rather entrepreneurial initiatives that leverage the most sohisticated data science techniques, and they bring a complement of revenue rather than a full income. Also, these "occupations" are usually not fully automated (when they are - and sometimes they indeed are, you are litterally making money when sleeping). Only top talent with strong business acumen is able to succeed due to massive competition.

Here are a few of these jobs or activities:

1. Pay-per-click arbitrage: buy and sell clicks on ad networks (can be fully automated thanks to automated bidding and automated click fraud detection relying on carefully crafted and continuously tested algorithms)
2. Stock trading strategies: competition is so stiff that there are only two ways to succeed: (1) insider trading, e.g. you try to obtain job interviews with small publicly traded companies, then based on information glaned during the interview, perform trades and (2) use trading strategies that professional traders will never use, e.g. stay "all cash" for several years on your trading account, and when the right event occurs, massively trade major indexes for a couple of days, then go dormant for another few years. You need sophisticated statistical models to succeed in this, with good back testing, walk-forward and robustness based on state-of-the-art cross-validation.
3. Sport bets and gaming. Requires very good statistical models (include fraud patterns in your model) and deep domain expertise, and to carefully select which brokers you are going to work with. Horse racing (Australia) is a good choice, depending on the broker.
4. Become a digital publisher of data science (or any) content, and use Google adsense to monetize your websites. Requires great SEO / SEM skills. This can be fully automated thanks to content syndication. Not for the amateur if you want to make a living out of it without working at all - it is not easy to automatically build real, targeted and growing traffic in a purely automated way, but it is feasible. Revenue could also come from your e-newsletter, thanks to Google or other ads.
5. Write data science e-books or reports, and sell them on Kindle (Amazon).
6. Launch a job board where recruiters automatically purchase via credit card and post job ads. Once again, for full automation, the challenge is to automatically grow targeted traffic. Difficult, but not impossible thanks to automated promotion in various social networks, and via Google adwords and careful CPC pricing strategies.

Question: Are men more likely than women to be interested in these "no human interaction" types of activities?