The failure is that of the machine learning world to underwrite the prize.
There is no greater potential risk adjusted return to the field.
$100/month paid out over the lifetime of the prize must be compared against the investments in machine learning over that time.