@svpino I get the efficiency from turning from fixed slot meeting to writing (in your own time).
Did you also do this for a remote team, and how did you address the secondary "seeing your teammates" part of such meetings?
@ZiggoWebcare Wat een ontzettend onvriendelijke verkoper die ongevraagd telefonisch glasvezel probeert aan te smeren. Dit is geen plus voor jullie reputatie.
It's that time of the year to get Wrapped... AND ROASTED. ๐ฅ๐ถ
We want to see your Spotify Wrapped side-by-side with our music judgement A.I. Be sure to tag us and use the hashtag #WrappedVsRoasted
Here's one from our team. Get your music judged here: https://t.co/S3FoQAHmDX
Thanks Nora Van Elken for spending 1.012 minutes with me in 2022. I couldnโt stop listening to Sumatra - Lstn Remix. #SpotifyWrapped https://t.co/5nQMxoae6z
Here are the pictures of two different problems.
Two classes. We want to use a neural network to solve these problems.
What's the difference between them?
1 of 8
Alright, fine: it's getting enough traction that I think I need to address this paper as a certified Grumpy Linguist in NLP.
(I generally try to avoid peer reviewing students in public but at this point it's definitely *already* very public, so ๐คทโโ๏ธ.)
https://t.co/mD6zceqnup
In a guest post for @StackOverflow, I take my first stab at writing for a practitioner audience to argue out that "Ethical AI" too often focuses on *how a system is built* rather than *what the system is used for*.
https://t.co/BNYgFG7rXA
#SANCTION ๐ดโน๏ธ Cookies: the CNIL fines GOOGLE a total of 150 million euros and FACEBOOK 60 million euros because the Internet users can't refuse cookies as easily as to accept them ๐ https://t.co/0vqCWydkJp
Journalism AI โ Quotes extraction for modular journalism - An NLP pipline to extract quotes from news articles using NER, add coreferencing information and format the results for an exploratory search tool!
GitHub https://t.co/6pJ5S3P8j9
Blog https://t.co/OQtQbYNMui
@scataco@sebrose@gasparnagy Still working there, and with the Gherkin features :)
Besides using it for other end-to-end tests, what Scato refers to is our use of feature tests for testing ETL processes. Happy to help ๐
From a Data Science perspective this really opens up possibilities to more easily integrate recommendations into the product.
From my engineering perspective this was a nice opportunity to play with Protocol Buffers and GRPC in setting up the service.
(small thread about a change in our recommendation infra)
At Blendle (and thus Cafeyn) we are really close to switching to a GRPC-based recommendation service ๐Our current/previous setup was to prepare recommendations and store them in a DB, allowing backend to fetch them./1
It also eases the effort from backend as they only have one way of getting recommendations, whether these are for homepage blocks, recommendations below an article, or in newsletters./3