Wikimedia Foundation is huge. No, I mean, really huge. So talking about what it does as a whole would require a lot of time - and at the same time, telling that it tries to give everyone “free access to the sum of all human knowledge" would still be true. That's kind of vague description, isn't it?
Yet it gives the right feeling. Wikimedia Foundation really strikes to try to provide knowledge for everyone, and make this knowledge as comprehensive and unbiased as they can. It's not an easy task, and it definitely requires a lot of work from the community too - but there are a lot of projects which are aimed towards this goal. One of them is Abstract Wikipedia.
If you are reading this article, that means that you are quite good at least at one language, English, and that means that finding information on Wikipedia isn't much of a problem for you. That is because English Wiki is huge, containing more than 6.2 million articles.
But that's not the case for everyone. A lot of developing countries still suffer from knowledge gatekeeping. It's hard to learn English, if you are not privileged, and it's almost impossible to access free or relatively cheap knowledge in your native language. And if you suddenly find Wikipedia in your native language - it would be quite small, most likely, due to the fact, that there's probably only handful of people around the Globe who could contribute by creating articles and translating them from other languages. Of course, the autotranslation tool exists - but it lacks an ability to translate to the rarest languages, as there's not enough material to train this tool.
This is where Abstract Wikipedia project comes to the spotlight. The basic underlying idea is that encyclopedic language is quite limited, comparing to normal language, so there exists a finite amount of words and phrases, that would allow to transfer facts and relations between them well enough. So to translate an article to the language, there such article doesn't exist, it's enough to find an article, written using Abstract Wikipedia terms and functions - and translate these terms to the target language.
Of course, it may still sound confusing, so I'll just tell the main idea more directly - it is faster to translate about a thousand words and word combinations than to translate 100 articles directly, one by one. So this project means that even a bunch of wikipedians could fill a Wikipedia on rare language through using Abstract Wikipedia, making knowledge a lot more accessible (which is the goal for the Foundation, as you already know).
It's a legitimate point to ask "and where does the facts come from"? Well, yes, Abstract Wikipedia is more about translating tools to show a relationship between facts - but Wikidata, which exists for 8 years already, is full of facts to fill these gaps.
Abstract Wikipedia project is grand in it's goal. Yet it's out there not only to help in translation, but also to make different functions accessible to everyone. What do I mean by "functions"? Well, right now Wikipedias are already using them a lot. You have someone's birthday date, and you need to automatically update their age. Use a function - Scribunto module, a piece of code, created by someone else and stored in your wiki. You want a page to show a warning, that it's dangerous to edit it, as it would affect a lot of users - use a function, special template, which will do this job for you. They are already here - and there could be more, helping in matrix calculations and strings processing. This is the goal of Wikifunctions, which lies underneath Abstract Wikipedia proposal as a translation tool.
And what is my job, as Wikimedia intern, in this huge project? It's quite simple - at the time of internship Aisha, who's also an intern, and I are the ones, who should fetch all the community functions in all wikis and try to find copies between them. Sometimes the code can be literally copy-pasted, sometimes written from scratch to hold the same functionality. There's no global database now, so when someone wants to use a Scribunto script or a template, еhey must look for it or write it themselves. A global database with these functions would save you time, nerves - and storage space on Wikimedia's servers.
So, while the whole Abstract Wikipedia project is huge, this task is quite small - and yet it's going to make life a bit easier for thousands of people. And even this task is a step forward for the project - an important and helpful one.