Send a sailcraft probe[0] to a star ~166.6 light years away with mean 0.2c speed which will take ~833 years. It must be programmed such that once at destination it should shot a laser message back to earth which will take ~166.6 years. Is that calc correct?

[0] https://phys.org/news/2020-05-interstellar-probes-starshot.h...

Delivered and understood? Delivered and have impact? Exactly 1k years? What's your budget?

Just one message? More than one? A stream with a 1K year delay?

Also, why?

Put it on a rocket and set it up for a 1000 year trajectory that would bring it back to Earth (or simply a local orbit that decays over a millennium). I’ve only played KSP so I’m sure there’s some impossibility in this (could a parachute last that long and remain functional?). The medium of the message is another interesting question for this.

Pick a sufficient number of iterations. Recursively hash a random input that many times and use the resulting hash as the encryption secret for the message. Publish the encrypted message, the seed[s], the number of iterations and detailed decryption instructions.

You cannot easily parallelize the solving, but using multiple chain encrypted seeds we can parallelize the generation.

You could mix different hashing functions with different characteristics to hedge against different kinds of computational speedups. Though speedups could also work in our favor, since nobody may want to start solving the challenge today.

For redundancy you could try different less reliable methods to deliver solved seeds at the right time. Say you try to deliver the seed that should have taken 100 years to compute after 100 years somehow.

I don't think too many people have made a lot of fanfare about doing this, so you'd have an advantage in being one of the first.

Here's one important example from history... https://en.wikipedia.org/wiki/Rosetta_Stone

It depends on where it is going to be delivered, you might be able to send a message to WASP-142b that would get there in 1000 years.