Holochain Forum

Link with HoloREA / Value Flows

Ah good to hear there is some linking with HoloREA.

Would love to chat some more. Where are you based?

In Belgium.

Hi, greetings from ValueFlows / HoloREA. Happy to be involved with coordinating the projects if it helps.

I wonder if reputation is modelled using the Value Flows language and if/how it could be implemented in HoloREA.

My two cents (or half a cent maybe). I think there is a direct connection between economic activity and reputation. For example, if someone commits to doing something by a certain date, and then does it by that date, or is late, or doesn’t do it, that should affect their reputation. Or if someone’s work is rejected by testing, or by the process that it is input to, then that could affect reputation. On the flip side, people do make economic decisions based on reputation of the agent. But VF does not define reputation directly in terms of reputation types or anything, it sticks to base economic activity. So maybe it is a matter of seeing how the models can mesh?

Hi @lynnfoster, nice to join this conversation.

For example, if someone commits to doing something by a certain date, and then does it by that date, or is late, or doesn’t do it, that should affect their reputation

Yes, that’s exactly how I see it. What we need in HoloREA, imo, is a sort of event-driven listening Holochain component (event processor), that is activated when a ‘vf:EconomicEvent’ and a ‘vf:EconomicResource’ are created (or any other VF object that makes sense). It could then check the according planned objects as well, in any case, it could create a new event when the time window has passed (either positive or negative, reputation up or down). Such event processor should be capable of knowing when to expect a particular VF object and be able to create a reputation event even if that VF object is not created. That would be something to put forward to HoloREA team and @philipbeadle. For me such (REActive) event-driven processing is a natural fit with the Holochain agent-based architecture, but I don’t know if anything like this is available (in RUST/Holochain) or planned. There’s a lot going on out there.

I see a reputation up/down event potentially as a ‘vf:EconomicResource’, a side-product generated by the event processor. That reputation event would then be further pushed to some reputation computation - those zomes @sidsthalekar is mentioning - and the reputation score would then be updated by some other zome. In that sense, I think the reputation score can also be represented as a ‘vf:EconomicResource’, as it is a current-see. Right?

Looking at it even more from a VF angle, we could use a ‘vf:Process’ to represent this entire reputation process, as it clearly is an (new) digital economic activity.

1 Like

I’m sorry to be a bit of a nay-sayer. But I don’t really see a reputation score as an economic resource. An economic resource we have currently defined as “a resource which is useful to people or the ecosystem.” Reputation is an important part of our human relationships and interactions, including economic ones, but I think that is different than being a resource in its own right, something that can be used, consumed, traded, given away, etc. Reputations and economic resources have a lot of different behaviors.

I don’t see that as a problem though, nor does it say that your ideas @yeff about event triggers and listening components are wrong, by any means. Perhaps a “reputation event” triggered by one or more economic events, economic commitments, etc.? And a “reputation score” calculated using these “reputation events”? Just thinking out loud…

It also strikes me that some of the reputation-generating situations we’re talking about could be part of P2P coordination between the agents involved, maybe even before a bad reputation event kicks in. Maybe some shared logic?

Also @bhaugen reminded me of this relevant post by Art, lots more nuance there. http://www.artbrock.com/blog/reputation-orthogonal-exchange

1 Like

Thanks for your thoughts @yeff and @lynnfoster. We’ve been thinking about getting a conversation going with you guys just in the last few weeks.

I would say we’re thinking about reputation as an economy in itself. The basic tenets of this economy are outlined in our AMA thread from last week

@lynnfoster, you called it right. Reputation is not an economic resource as such because it is not bound by materiality. Principle #4 suggests Reputation is non-zero-sum in nature. You can’t spend it, but you can stake it.

Principle #3 suggests that reputation is best experienced when designed as Relative, not absolute. In other words, my perception of you, may differ from @yeff’s perception of you. We don’t need consensus for reputation scores - which is why we think it’s a powerful currency for agent-centric environments. (See image attached)

We reckon the intersection between VF/HoloREA lies in ensuring the underlying data for reputation can be accessed by different DNAs. We’re chatting with @philipbeadle about using bridge calls, access tokens and some of the REA frameworks for this purpose. This will intersect with DPKI/HoloChat to ensure users have to provide consent in order to expose data to different DNAs.

We’ll also build a bunch of reputation zomes for the community - these will specify a basic library of ‘types’ of reputation (likes/upvotes-downvotes/timestamps etc.) which will help entrepreneurs roll out quicker.

Would love to hear your thoughts.

We reckon the intersection between VF/HoloREA lies in ensuring the underlying data for reputation can be accessed by different DNAs.

Makes sense to me.

Also, from the AMA discussion:

This involves providing users a ‘dashboard to navigate the new economy’. Think of it as a control panel through which you can build and harness your reputation wealth to modulate your engagements in the new economy.

We are thinking along these lines too, although not specifically for reputation, and possibly not even specifically for economics. Maybe we should combine efforts, or at least exchange ideas? At least once one of us gets focused on a dashboard; I don’t know when we will be working on this, but to me it seems important.

Dear @lynnfoster , dear @sidsthalekar ,

If an “vf:EconomicResource” is defined as “a resource which is useful to people or the ecosystem” (in VF description) and only “when it is bound by materiality” (in the words of Sid), then I want to challenge you both with a few Sunday morning thoughts.

I agree with the usefulness of an economic resource, but I don’t agree it is only bound by materiality.

Lynn also said about reputation: “but I think that is different than being a resource in its own right, something that can be used, consumed, traded, given away, etc.”

Let’s be clear: reputation is not the “reputation score” which can be used, but not consumed, not traded and not given away. The fact that the reputation score can be “used” in economic activity makes it an economic resource in my view.

I don’t question that both reputation and reputation score are bound to one identity and only has value to that one identity, as Arthur Brock says in his blogpost, and thus trading makes no sense. To stay with Arthur’s example, the electricity is the reputation score, the magnetic field is the reputation, and as we all know the first is an economic resource.

I absolutely agree with Arthur that building a reputation - and the score to derive from that activity - requires a lot of transactions, a chain of relationships. If I have a (verifiable) claim from my diploma as MD or lawyer, that is just a starting point to be able to build these types of reputation chains, but off course it is through many positive interactions with patients/clients that I build my reputation. Only those involved in the interactions are able to make an assertion about me and can add to my reputation score.

No doubt about that and concretely, using the DPKI identity on Holochain is crucial to connect to vf:Agents in HoloREA. In which direction is a privacy concern in case of foaf:Person, but that’s another discussion, but this is very important at least in Europe (with GDPR), and soon in Brazil, California, etc. When a reputation score is bound to the identity of a person, then in GDPR regulation, this reputation score becomes personal data. It always has been, but now it is much stronger protected by law.

The same is true for any type of data that is derived from the behaviour of a person, wherever it resides doesn’t matter, it becomes personal data. Does anybody doubt that personal data is an economic resource? Ask Google or Facebook.

But the score that comes out from building the reputation chain, can be modeled as an economic resource and can “used” in an economic event that is part of a process that generates another economic resource. I’m talking here in VF language. That is the only point that I want to make. And it is totally immaterial.

An example:

I can write an article in some kind of community (eg. Reddit), but if I have not a high enough reputation score on the topic, I can’t publish it and not receive some current-see in return. So the input to this process - of publishing the article - are both the reputation score and the article. The reputation score is like the electricity driving a machine producing an intellectual/digital product. I can definitely model it in VF language :).

This example still has some kind of material ring to it, as it produces a digital product, but many examples can be given in which there is no material or intellectual property. What is oral consulting, coaching, healing, advising more than a bunch of words - in some cases even words are not needed, in case of a guru - thrown in the air, but they are considered for sure as economic “services” in Value Flows and any other economic language.

We can be very creative on the type of “rating” that people give in return of a service, that rating itself is an economic resource, a value expressing the usefulness of the service received.

This “rate” can be a symbol on a website expressing a reputation, but it can also be the 2000 Euro paid to a business lawyer or both combined as outcome of the same process.

It is exciting to me that this sort of value flows can be designed in VF and soon be used on HoloREA …

I think it’s time to make these intellectual services much more “explicit” with VF and HoloREA.

1 Like

Tnx @lynnfoster to remind the article of Art, it is great and really helpful.

Reading it through till the end, I also can see now that a reputation score is a bundle of data - multiple dimensions and meaningful scales - captured about a service/product, like Art illustrates with the example of how he thinks eBay should have done it.

Seeing the detail in this bundle will paint a much more subtle image of an Agent’s reputation.
All super important in economic activity :).

@lynnfoster @yeff maybe it’s time for us all to get on a video call some time? Open to the suggestions above.

1 Like

@yeff - good thoughts! And I’m keen to respond - I’m going to mull over this for a bit since it’s late on a Sunday where I am. Will post soon.

1 Like

Economic resources do not need to be material. Designs, for example, can be economic resources whether or not they are embodied in any material form.

I think differences between reputations and resources that can be exchanged, produced, consumed, etc. are pretty well captured in Art’s article and some of the discussion here. Would be good to enumerate them.

One aspect that might be considered reputational is a skill or expertise that is
used in work, and might qualify someone to do some type of work. I think of that as a potential resource possessed by an agent that can be used for economic work.

Seems to me that a skill exists, and can be used in some work, whether or not the person has any reputation for that skill. Of course, a good reputation will help the person in finding places they can provide their skill in a group context, but that is separate from the skill (though connected).

Does anybody doubt that personal data is an economic resource? Ask Google or Facebook.

That is an interesting point, will think more about it.

1 Like

Re. a call: Bob and I are in CDT, Chicago time US; @yeff I believe is in Belgium so CEST? @sidsthalekar where are you? Bob and I get up early.

I’m in Singapore, my team mate @jethro is currently in London. I think early morning Chicago time might work best for everyone here?
What day works?

Yes, 9am Chicago time is 16pm Brussels time is 15 pm London time and 22pm Singapore time.

Works for me most days, not this Monday.

That works for me. Tues/Wed/Thurs this week are good. Let’s wait for the others to respond

Great to see you here @bhaugen.

There’s no doubt reputation holds potential for economic value. In fact, our aspiration is to build an entire economy underpinned by reputation - it’s something I’ve personally witnessed as a result of my upbringing in less industrialised cultures.

I think the distinction we’re trying to drive is between material and non material forms of value. Reputation obviously falls into the latter. In particular, agent centric environments favour relative reputation, which means peers in a network don’t need to share the same scores, but instead calculate a value locally that is based on their context. That’s why we think the REA protocol could be used for the underlying data, but not necessarily the scores generated from them.

I can share more details over our call :slight_smile: I’m also open to us adhering to a shared vocabulary so we’re not mis-interpreting each other.

Gonna call it a night here in Singapore!

A couple of reputation elements that you can get easily from REA data is who’s done what work and made other contributions to a project or network:
http://nrp.sensorica.co/accounting/agent/326/

Another element would be a record of keeping commitments (or not). Doing what you promised.

Both of those are relatively objective, as opposed to popularity scores.

1 Like

9am good for us Tuesday or Wednesday, Thursday good starting at 9:30am. Could also go earlier on Tuesday if that makes it saner for @sidsthalekar.

1 Like