Comment by zaroth | Hacker News Reader

Comment by zaroth | Hacker News Reader

What I hope actually kills all these products is a decentralized p2p open source system which lets humans find and rate other humans for whatever goods or services they may be selling.

The payment system is getting there (e.g. Bitcoin). The consensus algorithms are being worked out, the distributed trust systems are coming into place (e.g. Stellar). The scalable p2p discovery and mesh is pretty much already working (e.g. BitTorrent).

Why do we need to pay a 10% tax to companies to do this? I mean, as a civilization. To think the algorithms could never exist that would let me find a professional ______ near me which is, a) an actual person, b) someone who actually does this work for a living, and c) has real-life honest reviews from people who have actually paid that person to provide that service... it just doesn't seem like an intractable problem.

If you add up all the expert help, trusted reviews, on-demand this, cloud assistant that, we're talking what, at least 100 billion dollars of rent seeking?

Whether or not such a system has an optional secure anonymity layer (a la OpenBazaar) is just a matter of how you want to boot-strap the initial user base. Needless to say, "50% of the world's drug trade also happens to be conducted on the same code-base" may or may not be a great answer to the "how did you hack this system" question.

I think the "1099 Economy" is actually a great thing, I just think we could do without the gatekeepers.

If you want to see the non-rent-seeking version of a local marketplace, look at Craigslist. Their service is completely free to most participants.

Craigslist has a lot of things going for it, but the "rent seekers" are able to create other advantages that Craigslist can't. They can staff people to bring businesses onboard, get everyone's metadata formatted in the right way, take business photos, handle payment disputes, etc., that they wouldn't have the resources to do without charging a fee on the transaction.

For some, Craigslist is better, for others, the Thumbtack experience is better.

Craigslist is an interesting example. It shows how absolutely little you need to provide real value to a lot of people, but it certainly seems like they aren't even trying to make it better. I think if it was open source you would actually see a lot more interesting and useful features continuously being added. I mean, how long did it take for them to add a map to their rental listings?

The software will eventually exist to properly format the meta-data automatically for the vast majority of cases. Also things like providing pluggable viewports / templates so it doesn't have to be one-size fits all.

There's nothing to prevent add-on services being used in a decentralized market. For example, payment escrow is absolutely essential for many purchases, and payment escrow services themselves would be discovered and ranked through the same code base. But if you don't need it, you shouldn't have to pay for it. Discovery, search, storage, connectivity, etc. is all done using the local compute and storage resources and leveraging network neutrality.

This is an absolutely huge problem space, I've been thinking about it for years, and am hoping to start working on it at least part time later this year. For example, my dad had to move a car last month. He didn't know at the time, but the guy he hired was actually a broker who took 30% of the cost as a finder fee. "Car Transport" is just a simple category in a vast field of services, but you can imagine the interface to collect leads is trivial, and the interface to select vendors, agree on a contract, and commit to the service would be mostly common platform code.

Obviously the biggest problem is getting people to use it. The advantage of an open source and distributed p2p approach is the burn rate is basically zero. I can afford to work on it pretty much for the rest of my life at this point, and it would be reward enough to see people using it and benefiting it. Well, we'll see how it goes. :-)

plus I don't trust Craigslist, but I won't give a moments thought to purchasing anything from Amazon or from a 3rd party through amazon.

And yes, paying an extra 10% is ABSOLUTELY. FUCKING. WORTH IT.

It's exactly like paying for insurance. Insurance is peace of mind.

Amazon is an incredibly powerful band, and brands can do that. But there's also no reason a distributed system can't also have a brand which evokes the same level of trust. In many ways, distributed trust can be stronger and more resilient than centralized trust.

I'm curious, what do you mean by "distributed trust"? A big reason why I find a brand trustworthy is exactly because my trust is consolidated in one entity that produces repeatable, desired results.

Taking Craigslist as an example, Craigslist isn't trustworthy to me precisely because its trust system is distributed. It doesn't police the interactions, moving trust into the realm of peer-to-peer, where definitions of trustworthiness tend to vary greatly. Craigslist in effect has built a well-known brand with overtones of "buyer beware!" because of this distribution of trust.

That's how I perceive Craigslist.

And by no means do I avoid Craigslist. I've purchased plenty of things off of it ranging from coffee tables to what I lovingly call my "man couch" to the occasional video game.

But whereas with Amazon I just purchase with the assumption of quality and SAFETY, with Craigslist I do due diligence first. There's a good reason why I buy things locally that I either go pick up or they deliver. It allows me to inspect the goods before completing the transaction.

Now, I think in general people selling on Craigslist aren't out to specifically hurt you, but they're not out to do right by you either. It's just a completely different arena.

I recently purchased a brand new PS4 off of Amazon. No way would I do that off of Craigslist. Do you know that PS4 died within 5 hours of me having it? Amazon replaced it no questions asked. There's a reason why I would never dream of purchasing something like that through Craigslist, even if many people do it successfully, the cost savings isn't worth the risk (for me personally).

So then you would have the same level of trust that the system works to produce repeatable desired results. E.g. With Bitcoin, if you do trust it, you are trusting that the system works to produce repeatable desired results. The fact that it works through (in theory) a resilient distributed consensus algorithm is an implementation detail.

Amazon happens to achieve their consistent experience through top-down hegemony. I believe that in time a distributed software system will provide an even more reliable and consistent experience in many cases.

For example, Amazon's warehouses and level of automation are often regarded as a key competitive advantage. But when you're buying a good or service from a neighbor who lives within a couple square miles, distribution doesn't get much more local (and scalable) than that!

Craiglist pushes the problem under the rug, they wash their hands of it and say "buyer beware". That was the only possible solution in 1995. 20 years later we are starting to figure out, you can actually do a heck of a lot better, even without a central authority or billions of dollars of infrastructure.

I think that's the hidden message in the nascent "sharing economy". Not only can individuals successfully monetize their assets or go to work for themselves, but once the tech catches up, they can do it directly and in fact way more efficiently than any centralized system could ever dream of.

By the way, I'm not saying there's no place in the world for Amazon. But in particular making connections to qualified service providers are low-hanging fruit for a p2p distributed system, and longer term much, much more.

I don't understand enough about the details of how Bitcoin works to evaluate how apt that is of an analogy it is here. I've heard some very positive as well as a large enough number of surprisingly negative experiences with it making it hard for me to evaluate and thus hard to trust.

I'm having a hard time following your argument about the sharing economy showing us that distributed trust systems work more efficiently. Many of the major players in the sharing economy I've seen (Uber [0], AirBnB[1]) have all moved from open-ended, less regulated distributed trust systems to tightening and more strongly imposing rules as a centralized authority to rebuild damage to their trust due to bad actors.

To avoid thoughts that I'm cherry-picking companies with controversies, my broader point is that the sharing economy isn't really built differently than Amazon or any other marketplace model. Central authorities (TaskRabbit, Exec, AirBnB, Homejoy, etc) inject trust into p2p marketplaces because without their centralized authority, p2p marketplaces wind up like Craigslist or, even worse, Silk Road.

[0] http://money.cnn.com/2014/12/18/technology/uber-background-c... [1] http://www.sfgate.com/business/article/Airbnb-hosts-must-app...

edit: minor wording correction

Agreed. 10% would be worth it. That's like ebay.

20% + card processing fees is not worth it. I think we'll see market forces push down the rates of return on these companies.

<zaroth>: I'm not sure why you were downvoted, so have an upvote.

Not saying I entirely agree with you -- calling it a "tax" is incendiary and no more accurate than saying "why do we need to pay a 10% tax to grocery stores or Amazon?" -- but your broader point is an interesting one.

I'm pessimistic about this application of p2p discovery in the short term, but in the longer term (>7 years) you may well be right. The short/medium-term problem, and this goes back to 1990s discussions among cypherpunks about anonymous reputation systems, is that those "honest reviews" are attractive targets for gaming.

I don't need the karma, I'd much rather have the discussion! :-)

Rent-seeking is probably a better description than tax, in that while it certainly costs a lot to build and maintain these systems, and they do provide some measure of value, the goal here is to build scale, build monopoly, and then extract profit.

The alternative is building the technology in a open / decentralized fashion which doesn't grant any level of centralized control. Open source software provides such incredible economy of scale that I do think such a system could one day reasonably compete with the best a multi-billion-dollar corporation could offer.

10% would be OK. It covers their card processing, advertisement fees, paying some software developers reasonable rates to improve the service.

But 10% is not real. These sites charge 15-30% on top of card processing fees. That's too much.

Agree these guys are parasites. For example, Thumbtack charges landscapers $15 per lead. It could easily take 5-10 leads to get a paying job. And these are not high earning professions.

Reply on news.ycombinator.com