Edit to my previous sentence: I want to explore how to work towards a career in academia, while facing the possibility of having invested many years of your life  f̶o̶r̶ ̶n̶a̶u̶g̶h̶t̶ in something you enjoyed.
2. Open Science
As a PhD student and research assistant I am strongly convinced that practicing Open Science (sharing code & data, ensure reproducibility, create and contribute to FLOSS, ...) does not only benefit science and society, but the individual practitioner as well. Headhunters and job agencies use GitHub to find and target developers fitting their requirements. The stars and forks on GitHub are real endorsements that are  acknowledged and appreciated by both science & industry.
Thus, I believe that while I'm  trying my best to finish my PhD (sooner or later...) going the Open Science way, is the best to keep working on Statement 1 while reducing the negative impact of Statement 2. Which is basically all I want for now.

But Github is... for coders

I have previously presented my own situation and I believe that I am in a very lucky position of having the luxury of doing what I enjoy, while preparing for the worst case. Open Science is giving me an opportunity to practice and demonstrate  skills which are valued in both worlds. But what about those who do not collect and analyse data, write software and computational methods, or whose preprints aren't valued by their supervisor and peers? What about the plan B's of all the grad students in, for instance, philosophy?
I don't know if it's the wrong approach to think in terms of GitHub and data science, or if there is simply no good answers and solutions to these questions. But I went ahead and created a challenge for the upcoming OpenCon doathon, and would love to discuss how we might document and demonstrate research skills in the humanities in order to support a career outside of academia?
If you're interested too, here's the link to the GitHub repository and feel free to add your thoughts and ideas: OpenCon Doathon 2017 - Challenge