On many aspects, Google is a Good Old Fashioned Artificial Intelligence. It does knowledge base construction with freebase done by humans, I does quality assurance control and curation by humans, it does learning-to-rank machine learning based on algorithms (and (hyper) parameters) curated by humans, also spell checking and keywords suggestion are based on statistical significance mined from keywords co-occurrence in queries done by humans... All that stuff look like brute force and humans.
Why #nlp is all about machine learning stuff ?!
Capitalism had its chance to regulate itself and it didn't and people are pissed and they are only going to get more pissed.
re: free software scavenging / legalese
That is a hell of problem that I just solved while procrastinating doing nothing in my bed. Prolly, it would have been better time spent reading Concurrent ML book, but what for when I can just be a legal LGPL user :)
I hope to make a release this week-end 😃
free software scavenging / legalese
It seems very clear to me that LGPL work can be used as part of a larger work that is not necessarily LGPL or GPL, in my case: #CNPL.
There is thin line regarding language ports, since GPL et al. apply to the code not the algorithm. GPL is not a patent. It is even more blurry in this case, since both language are Scheme.
Mind the fact that I will rework the code even if I intend the keep the gist of it, I will rename a few things and at least drop "work stealing" feature and make it work with R7RS libraries and avoid all around match.
That is to say, I can copy and **rework** guile fibers as long as I release the code using LGPL 🤩
Positive thought of the day 😂
For those interested, there is also a Matrix channel for functional.cafe:
wanna be a green Scheme engineer
The social network of the future: No ads, no corporate surveillance, ethical design, and decentralization! Own your data with Mastodon!