I had an amazing indie music discovery service in Google Play Music. I found so many fantastic underplayed artists, and it helped me explore all the small music venues in my city. I've got a wall full of signed albums from artists I discovered with Google Play Music.
YouTube music recommends Britney Spears. It's so awfully wrong about my tastes.
It also randomly inserts YouTube parody videos into my playlist. Why the hell would I want to listen to stuff like this https://youtu.be/-5jWtz3rzco ?
I hate Google so much now. They're like evil 90's Microsoft, but incompetent. They've got their ad monopoly / web destruction engine to sustain them, but they're Dilbert Pointy Haired Boss bad with everything else.
No gamers will be surprised when Stadia gets canned.
It'll be hilarious when they decide to shutter GCP. Remember when it leaked that they were internally threatening to defund it if they couldn't hit growth targets? Imagine all their B2B relationships getting hit as hard as their consumers do.
Dead Comment
E.g. some very active projects are:
* Kaldi (https://github.com/kaldi-asr/kaldi/) obviously, probably the most famous one, and most mature one. For standard hybrid NN-HMM models and also all their more recent lattice-free MMI (LF-MMI) models / training procedure. This is also heavily used in industry (not just research).
* ESPnet (https://github.com/espnet/espnet), for all kind of end-to-end models, like CTC, attention-based encoder-decoder (including Transformer), and transducer models.
* Espresso (https://github.com/freewym/espresso).
* Google Lingvo (https://github.com/tensorflow/lingvo). This is the open source release of Googles internal ASR system, and used by Google in production (their internal version of it, which is not too much different).
* NVIDIA OpenSeq2Seq (https://github.com/NVIDIA/OpenSeq2Seq).
* Facebook Fairseq (https://github.com/pytorch/fairseq). Attention-based encoder-decoder models mostly.
* Facebook wav2letter (https://github.com/facebookresearch/wav2letter). ASG model/training.
* (RETURNN (https://github.com/rwth-i6/returnn) and RASR (https://github.com/rwth-i6/rasr), our own, although this is currently free for academic use only. It is used in production as well. Supports hybrid NN-HMM, CTC, end-to-end attention-based encoder-decoder, transducer, etc.)
And there are much more.
You will also find lots of ready-to-use trained models.
I said I'd prefer to have something smaller with less Christmas lights.
"Policy is everyone has to have the same laptop, and some people need more graphics power than the one you're asking for."
Turns out there'd been some big thing back and forth already about size and performance and this was the compromise that nobody liked.
"But why does everyone have to have the same laptop?" I asked.
"Well, it's just policy. Every desk has a docking station, and we don't want to have to buy different ones."
So I flipped the selected monstrosity over to reveal the extremely lacking docking connector at the bottom. There were some awkward laughs and I got my Thinkpad.
I then proved to management that there was about 5 hours a week being completely wasted for me waiting for my machine to open applications after they crash. So after that we all got new MacBooks that were properly spec'd out for game development. I got to pick the specs :)
Dead Comment