GNOME Desktop users can put this in a Bash script in ~/.local/share/nautilus/ for more convincing looking fake PDF scans, accessible from your right-click menu. I do not recall where I copied it from originally to give credit so thanks, random internet person (probably on Stack Exchange). It works perfectly.
ROTATION=$(shuf -n 1 -e '-' '')$(shuf -n 1 -e $(seq 0.05 .5))
for pdf in "$@";
do magick -density 150 $pdf \
-linear-stretch '1.5%x2%' \
-rotate 0.4 \
-attenuate '0.01' \
+noise Multiplicative \
-colorspace 'gray' \
"${pdf%.*}-fakescan.${pdf##*.}"
doneThat seq is probably supposed to be $(seq 0.05 0.05 0.5). Right now it's always 0.05.
Note that you can get random numbers straight from bash with $RANDOM. It's 15 bit (0 to 32767) but good enough here; this would get between 0.05 and 0.5: $(printf "0.%.4d\n" $((500 + RANDOM % 4501)))
> Turns out there was a lot of low-hanging fruit in terms of inference optimization that hadn't been plucked yet.
That does not mean the frontier labs are pricing their APIs to cover their costs yet.
It can both be true that it has gotten cheaper for them to provide inference and that they still are subsidizing inference costs.
In fact, I'd argue that's way more likely given that has been precisely the goto strategy for highly-competitive startups for awhile now. Price low to pump adoption and dominate the market, worry about raising prices for financial sustainability later, burn through investor money until then.
What no one outside of these frontier labs knows right now is how big the gap is between current pricing and eventual pricing.
The evidence is in third party inference costs for open source models.