4 ms·

I was thinking about this too, but there is one problem with a purely pay-it-forward type model - there is no 'debt'. That is, how does one earn karma in order

by Cogito 14y ago

I was thinking about this too, but there is one problem with a purely pay-it-forward type model - there is no 'debt'. That is, how does one earn karma in order to buy uptime? You can 'sell' your connection to someone else, but where do they get their karma from? Part of the solution might be to provide 'seed' karma to new users, but then you inevitably will have users who don't spend the karma they gain and so become karma sinks. This all assumes a zero sum economy, so perhaps a different model is a better idea. Keep track of metrics like how much bandwidth someone has shared or used (somehow), and then allow providers to define a measure for who can share their connection based on those metrics. For example, one provider might let anyone connect no matter what, while someone else might only let people who have shared their own connection connect. Determining what to share and to whom is a hard question, and when there are upstream costs to the provider the balance seems like it would be hard to find. The yourkarma model is more or less a loss-leading strategy (we pay you in bandwidth to sign up new users) and it's hard to see that being sustainable long term for the user. It's not completely obvious if the 100mb is for new karma users connecting, or for any connection, but based on their help page[1] it seems to be only new users: Why do I get extra bandwidth when I share my connection? Every time another person uses Karma, our company grows a little bigger. The more the merrier! So when you share your connection, you are essentially helping us and we feel like we should help you. After all, we’re not called Karma for nothing. What I am really interested in are the ad-hoc networking projects [2], that will be able to route data by hopping between networked devices, reducing the need of centralised ISPs. Simply being part of that network adds value, however similar problems will probably arise in any case. [1] https://yourkarma.com/help/82-why-do-i-get-extra-bandwidth-when-i-share-my-connection https://yourkarma.com/help/82-why-do-i-get-extra-bandwidth-w... [2] https://en.wikipedia.org/wiki/Wireless_ad_hoc_network https://en.wikipedia.org/wiki/Wireless_ad_hoc_network

jessaustin 14y ago
> inevitably will have users who don't spend the karma they gain and so become karma sinks. This doesn't seem like a problem to me. A karma sink is a connectivity source. Surely we're trying to optimize for the latter quantity? If it makes you feel bad you can give away prizes or something.
- Cogito 14y ago
  The problem with a sink is that you remove 'currency' form circulation. Without a way of circulating the karma back or a way of generating new karma then you can get a lack of liquidity, and hence make it difficult for people to 'buy' connectivity because they aren't able to obtain any karma.
  - jessaustin 14y ago
    It's apparent to me that there will be multiple ways of "generating new karma". Don't all of these devices come with a certain amount built in? If that's not enough (e.g. once most consumers who would buy one have already done so) then people who find themselves short can buy some more with actual money, or devices that are online can "farm" karma at a certain rate, or whatever. It's not difficult for the mint to print more money.
    - Cogito 14y ago
      Oh, in this particular case I agree, the Karma company generates the karma and furthermore consumes it all as well. What I was referring to was a distributed model where users essentially 'sell' their bandwidth to others for a credit against using someone else's bandwidth in the future. In that situation credit generation is a serious issue (how does someone 'pay' for my bandwidth if they don't have any credits themselves?). Using the term karma in that scenario is a bit confusing I guess :) In an ad-hoc wireless network there is the concern that some users will be selfish with the bandwidth they use and provide. Solving the problem of credit generation will help in that situation and in the sharing of an uplink bandwidth. The main difference is that in an ad-hoc network simply being part of the network adds value, and the bandwidth costs are based on power consumption and actual throughput, rather than an upstream cost.

by Cogito 14y ago

jessaustin 14y ago

> inevitably will have users who don't spend the karma they gain and so become karma sinks. This doesn't seem like a problem to me. A karma sink is a connectivity source. Surely we're trying to optimize for the latter quantity? If it makes you feel bad you can give away prizes or something.

Cogito 14y ago
The problem with a sink is that you remove 'currency' form circulation. Without a way of circulating the karma back or a way of generating new karma then you can get a lack of liquidity, and hence make it difficult for people to 'buy' connectivity because they aren't able to obtain any karma.
- jessaustin 14y ago
  It's apparent to me that there will be multiple ways of "generating new karma". Don't all of these devices come with a certain amount built in? If that's not enough (e.g. once most consumers who would buy one have already done so) then people who find themselves short can buy some more with actual money, or devices that are online can "farm" karma at a certain rate, or whatever. It's not difficult for the mint to print more money.
  - Cogito 14y ago
    Oh, in this particular case I agree, the Karma company generates the karma and furthermore consumes it all as well. What I was referring to was a distributed model where users essentially 'sell' their bandwidth to others for a credit against using someone else's bandwidth in the future. In that situation credit generation is a serious issue (how does someone 'pay' for my bandwidth if they don't have any credits themselves?). Using the term karma in that scenario is a bit confusing I guess :) In an ad-hoc wireless network there is the concern that some users will be selfish with the bandwidth they use and provide. Solving the problem of credit generation will help in that situation and in the sharing of an uplink bandwidth. The main difference is that in an ad-hoc network simply being part of the network adds value, and the bandwidth costs are based on power consumption and actual throughput, rather than an upstream cost.