schnitzelstoat
Joined 8 years ago
Joined 8 years ago
A poor model could require many more runs to solve a task, or much longer internal 'thinking' runs - all of which will increase token consumption but are a bad thing for the user.
Ultimately, you end up paying for the usage somehow.
If the authorities want people to accept systems like Flock etc. then they need to severely punish abuse like this.
The universal healthcare is hard to quantify - because most decent companies include health insurance in the US and the quality of the public healthcare provision in Europe can be quite patchy (i.e. location-dependent)
Remember that you are comparing gross salaries too, and taxes are much higher in Germany so the engineer ends up with much less take-home pay.
By the time we get robots capable of folding laundry, we'd already have solved many, many other use-cases.
They probably just used the Germans because they are cheaper.
Nowadays with a wife and kid of my own it'd just be impossible to spend that much time on it.
I was surprised it's recommended so much.
The strongest predictor (by far) for committing a crime is having a prior criminal record.
We are in the age of the replicant now.