>Apertus 1.5, our latest model release, extends the 8B and 70B Apertus 1.0 models with image understanding, an optional thinking mode, a four times longer context window, improved instruction-following, and better tool use.
Do you plan to contribute the data back into OSM? That would be the nice way to go to share with the community and allow you to have better data in the future with less maintenance work.
I would say that rising a kid is going to teach you many things you would never learn at school, and those social skills can make a difference in job interviews. If you're able to take care of a family then you certainly are a good team player.
Well this is not what is happening in practice, Wikipedia / Wikidata, OpenStreetMap, OpenFoodFacts... All provide APIs and even a full dump of their database available to download for free, but no, the stupid bots still DDoS them 24h/24.
5 minutes of wikipedia search would give you plenty examples of complicated software engineering problems that would have a big impact on everyone's life.
/e/ OS with Fairphone is the good choice for that. Don't listen to cromka, /e/ OS is now fully open as the only proprietary app was the map one and they just replaced it. So, 100% free software. It is less secure than Graphene but also leaks less data to advertising companies.
As a (kinda former, as the project is mostly inactive now) diaspora contributor who followed closely the creation of the first decentralized social network (with GNUStatus and Friendica), I'm really interested about this explanation, but it asks more question than it explains.
First of all, why did we move away from blog with RSS with app readers? (We did not fully abandoned that by the way, there are still some excellent RSS aggregators out there such as https://flus.io)
Because Facebook not only it allowed us to react / comment on what we were reading, it also allowed us to follow the reactions of others. The principle behind social network is that you don't follow a website author and so you know when they post, like RSS does, but you follow an account, and so you know every reactions they do. Imagine being informed when someone that you find interesting is commenting on a blog you did not even know it existed? And who themselves are following, so probably you should follow those people too, and discover even more blogs with interesting content!
There were technics to try to do that with blogs back in the days. Planets (public blog aggregators) to read linked content, pingback to know when your blog post is quoted somewhere, etc. But none of this was easy, efficient and complete as Facebook was. So everyone migrated there, because it made content discoverable, which means visibility, which exactly why we are publishing on the web.
But Facebook is centralized. This is a big flaw, we don't want to depend on a corporation to publish online. And so, here come decentralized social networks. I am not going to describe the Federation (the diaspora protocol) here as the comment is already way too long, but it does allow what I described above, at the cost of having an identity tight to an instance ([email protected]), but from which I can easily migrate.
Now, after reading this article on ATProto, it is not clear at all to me how all those interactions described above are implemented. Could you elaborate?
Well I am genuinely interested so I am all for continuing that discussions in details. I am happy to finally meet someone who had a real look and isn't just repeating things read online. So if you have time to share the result of your investigation I'm super interested. But here is not the good place I imagine, where can we continue that discussion?
https://huggingface.co/collections/swiss-ai/apertus-v15