Not sure why you're saying that systolic arrays never went anywhere. They're widely used for matrix operations in many linear algebra accelerators and tensor units (yes, largely in research), but they are literally the core of Google's TPU [1] and AWS EC2 Inf1 instances [2].
As someone with a very midwestern US accent, I got a much better UK score when I used my terrible Cockney accent, about 93% UK, than I did when I used my default accent, at roughly 80% US.
Interestingly, it looks like the model has quite the UK bias for words like "queen", "tea", "chips", and "pub." All words that are used pretty frequently in US English but have a clear cultural affinity to the UK. When I use sentences with these words in my casual midwestern accent, they almost always come up as 80+% UK.
Compare it to the parent article and you'll see what I mean. If you are the center of the watch and you put your hand in your shadow, it will face the direction opposite of the hour hand in the article. If you adjust your other hand to face 12 O'clock relative to that hand, it would be at 6 O'clock if you're going by the article. So the direction is flipped. When you clap your hands together, you'll be facing North instead of South.
I love this trick. You can also do this using your hands if you know roughly what time it is.
For example, if it's sunny out, place your arm in your shadow, parallel to the ground. Keep it there. This arm represents the current time (say, 10 O'clock).
Look at your shadow and point your other arm out, parallel to the ground, and adjust it until the arm's shadow points at "12 O'clock" relative to your first arm (in your shadow).
Clap your hands together, and you're pointing north!
[1] https://cloud.google.com/blog/products/ai-machine-learning/a...
[2] https://aws.amazon.com/blogs/aws/amazon-ecs-now-supports-ec2...