3c) Experts in the field, e.g., computer vision, after looking at the dataset, will know how difficult the problem is.
How to know who is an expert? Why would experts spend time evaluating datasets for a cryptocurrency? How to ensure that they evaluate the difficulty of a given challenge fairly and objectively?
Introducing the need for experts to evaluate the nature of the work to be solved (ie. the dataset difficulty) means introducing trusted third parties. This means you lose the core value propositions of cryptocurrencies -- being trustless and permissionless. And since the nature of evaluating the difficulty of a dataset appears to be subjective, you even lose the security aspect of cryptocurrencies.
Don't get me wrong, if you turn this project into a sort of incentivized Folding@home there might be some merit in terms of gamification to it. But from my current understanding of this approach it would make for an awful cryptocurrency.
1) bootstrapping is similar to how bitcoin started, no miners, easy to control 51% of the hash power. Value of coin is low. the pioneers will have to bootstrap the chain by giving out coins, etc. or follow the footsteps of ETH. Need to do research on how these 2 started.
I was referring to the following statement:
Sybill attacks can be discouraged by requiring each dataset that has a defined problem to come with an attached processing fee.
According to this, you have to pay a processing fee for either (a) submitting a dataset or for (b) solving a dataset challenge ie. mine a block. I'm not quite sure whether you had (a) or (b) in mind. Either way this means that currency needs to exist
before the first block can be mined. In essence, this introduces another centralized entity, namely one that has to issue the first units of currency.
2) yes, authentic dataset creation is a tough problem. Dataset contribution can be constrained as follows:
only aggregate datasets are considered, i.e., 100 individual user profile pictures, contributed by 100 users. This will increase the sybil attack cost. The mining problem could be to classify the user pictures into male/female, different races, etc.
3) fake dataset if contributed by a single person, is hard to detect. That's why I suggested the above. Some one posting a dataset and solving it himself would not reap benefits, unless he mobilize an army of miners, as described earlier.
100 individual users? Users of what? Facebook? Google? Some other form of online account? How does relying on the accounts of a third party service protect against a sybill attack? How do you even know that 100 users are 100 different persons?
Identifying 100 individual users (or votes, or however you may call it) is at the core of the sybill attack problem. Not its solution.
How to stop rogue clients from DDoSing the network by flooding it with wrong timestamps and turning 2 minute block intervals into 2 days or weeks or years?
By attaching a cost/fee to each broadcasted result?
So miners are
paying to mine blocks?
Why would anyone pay to mine a block? How would currency be issued in a system where miners pay for mining blocks but don't receive any block rewards? Or does the fee simply get substracted from the block reward? In which case, how is it a fee if miners simply receive reduced block rewards to begin with?
