Also, I'm not sure it was very wise of Mike to assume the data was truly incompressible. It's unlikely someone would manage to compress it (certainly for just

Also, I'm not sure it was very wise of Mike to assume the data was truly incompressible. It's unlikely someone would manage to compress it (certainly for just 5k), but if you truly believe you have a source for billions or trillions of bits and that those bits then have exactly 100% as much entropy - well, you're taking a lot on faith. Indeed the fact that the content length was preselected implies some amount of regularity that might be (abused). Oh well.