That's amazing. I remember back in 6th grade, thinking I could take a swing at this... 9 years later, not even a dent made. Maybe one day.
I guess part of it is that if you ask the average person, even if they have ML experience, they might not even know where to start.
It would have been better to go from CPU to GPU for compression/decompression, as it can get a better result for the same price, but on the other hand it's understandable that the prize wants to be compatible with earlier solutions.