[prohobo]

Autonomous services (similar to vending machines): See PoolTogether, a service which before blockchain would be considered almost certainly a scam, and would be nearly impossible to audit.

Because it's on Ethereum the contract is publicly auditable, and the rules are simple: insert currency into a pool, and have a chance of winning a %age of total interest earned once a week. No one owns the service. If the site goes down nothing changes. The developers can't just change the rules (backend) on you. They also have a DAO so the users can manage the service themselves.

A similar service is Compound, which recently was found to have a bug that gave away something like $90 million worth of currency to random users. That's the main danger with blockchain services: exploitable bugs.

Besides that, I've been thinking about something like a blockchain based Wikipedia (or Research Gate?), with monetary incentives for research teams to answer questions posed by the community (when they can't get funding by other means) and a codified scientific process for grading the quality of papers, as well as mechanisms to reliably cite related research papers to use as supporting evidence or jumping off points. Think of this: a debunked paper causing a cascade of debunked papers that cited it.

The point of using the blockchain would be to create a decentralized repository of research information with strict rules to maintain integrity of the information submitted (as opposed to relying on a centralized committee that might become corrupt through perverse incentives).

[brightstep]

The first two are just gambling, which is not very different from the speculation the author talks about. Your idea is interesting and speaks to another desire that seems to me to run beneath a lot of interest in crypto, which is the desire to make social change without political action. My perspective is, we don't need a blockchain to reward that kind of behavior. We could decide as a society to create such a system in myriad other ways. The real problem I see is that we're (especially in the US) very much disconnected from the political process, which leads some to look for technological solutions to the problems they see around them.

[prohobo]

The first two are not gambling. There is no mechanism where you can lose your pooled funds. Don't believe me? Read the fucking contracts. How much clearer can this be?

As for the rest of your comment, I think your diagnosis is wrong. This is purely political action, which is why so many people can't see the benefits. They think blockchain tech was supposed to clearly change what's possible - but it doesn't: instead it decentralizes it and removes the requirement of trust, which indirectly enables new possibilities. There is no current system that doesn't degrade over time due to various influences, or is immune from corruption. Blockchain based solutions don't solve the physical aspect of trust, and so would require the buy-in and cooperation of society through politics. What blockchains do is solve digital corruption and codifying rules, which changes the way people interact with systems.

[paulgb]

But it is gambling, it's just that you're gambling your interest foregone rather than the principal.

Back to my original question, though, I don't see what real-world problem it solves to add stochastic variability to interest payments.

[prohobo]

That's not what my point was about. The point was autonomous services, and I also outlined an idea for one that would have real-world benefit (if properly designed). You can argue that there are no "useful" services yet, but not that there will never be any.

[theamk]

Hm... people put money into a pool, and then there is a random chance of payoff? Sounds like plain lottery, something that is sold at every gas station ever. for PoolTogether, there is a twist that this seems to be incorporated with interest-bearing accounts, but presumably you can get the same effect by investing the money manually, and periodically withdrawing interest and using the money to buy the lottery tickets.

And "blockchain based wikipedia" sounds interesting but has a lot of real-world interaction, and so is substantially harder than existing examples which purely interact with blockchain.

Overall, this whole conversation reminds me of the the new programming languages:

- "My new language is very great! Look, it can do HTTP requests!"

- "But existing languages could do HTTP requests as well?"

- "Yes, but my language automatically deserializes the request using Mozart-Rachmaninoff type system, which eliminates the whole classes of the bugs. No other language can do that! I am going to write a new OS in this language now!"