SteemQ - A Decentralized Video Platform for STEEM
We believe that there is a market need for a decentralized video platform for user- generated content.
SteemQ aims to fill that need by creating a next-generation platform on top of the new Blockchain and P2P technologies.
We are bootstrapping a better version of YouTube, with a new social model that creates a richer experience for content creators, curators, and consumers.
Blockchain technology has disrupted the world by providing a mechanism for trust between arbitrary parties. It is the core building block for decentralized public platforms.
The blockchain of choice is STEEM. This allows us to build on top of the same technology that powers Steemit, as well as inherit the benefits of an existing community, currency and platform. All STEEM accounts are automatically SteemQ accounts and vice-versa.
The monetary reward on each video is derived from the stakeholding properties of its voters, while the ranking takes into account reputation-related properties, as well. This mechanism improves curation by granting increased voting rights to users with smaller stakeholding, and high reputation scores. The system aims to empower its users to the maximum extent possible while remaining resilient against Sybil attacks.
Unlike the centralized, privately owned platforms, SteemQ is censorship-resistant. Users will not be subject to content take-downs or other means of censorship as a result of an arbitrary decision within the platform-owning corporation, a faulty algorithm or a false copyright claim.
SteemQ allows its users to take action against spammers, content thieves and other misbehaving agents through downvoting.
No single person has the authority to nuke content by themselves. For every questionable piece of content, a consensus has to be reached by a collaboration of multiple high-reputation downvotes. In such an event, the rankings of the video drops, and the participants in the P2P CDN have the ability to remove the content and thus slow down or eliminate its propagation.
To aid in moderation-related decision making, a clear code of conduct shall be established.
Fast and Efficient
SteemQ's network becomes faster, and more capable with every P2P node it obtains. It is anti-fragile by design - the bigger it gets, the better it runs.
Example: You and your neighbor are both into snowboarding. A new snowboarding compilation has been released on SteemQ, and your neighbor is first to watch it. When you want to watch the same video, it is streamed directly from your neighbor's cache - rather than some data center in a foreign country. This not only allows for lower latency, but also reduces the load on your ISP.
We have to learn the core design patterns that will define SteemQ.
As a starting point, we can look at the two existing and currently quite prominent models: YouTube and Reddit.
YouTube treats each user as a separate entity, and offers an experience that is highly optimized towards passive yet recurring content consumption, which consequently affects the ad impressions. YouTube allows its users to subscribe to individual brands, however, judging by the design of the website, most of the discovery happens through "recommended videos" and autoplay functionalities.
In the center of Reddit's design is the sub-reddit. Each sub-reddit is like a little island, with its own community which is brought together by a common interest.
SteemQ aims to find a balance between both models. We are not afraid to try new things such as, but not limited to:
- a personalized feed that is based on subscriptions to individuals as well as sub-communities.
- a new content propagation system that replaces the multi-layer new/trending/hot model within the sub-communities
We believe that money should not be the primary driver of people's behavior. We would like to remove money and its influence on people's decision-making in regards to voting. Voting on contributions of other people should reflect the voter's experience with said contributions in the purest form possible.
For example, if a video made you laugh, or if you have just discovered something totally awesome, you should be able to express your reaction with an upvote.
If the information on monetary reward is presented, it might influence your perception, and change your behavior. Perhaps you might feel like the video already earned too much, and refrain from voting despite the fact it made you laugh in the first place. Perhaps it could even trigger emotions of envy and frustration, or amplify negative feelings in the realm of the scarcity mindset.
Monetary rewards cannot be perceived as "fair" by everyone, because:
- interpretation of value is subjective
- rewards correlate to the size and stakeholding power of content's target audience
- haphazard events
For example, a video explaining a 'binominal options pricing model' may be very valuable to a small group of individuals, and completely devoid of value to the rest of the world.
A funny cat prank video might be valuable to a large audience, but completely worthless to our mathematics enthusiasts.
We can conclude with reasonable certainty that more often than not, the cat video will outperform the geeky math video, making our hard-working mathematician dissatisfied with his reward. Perhaps a whale who likes math will discover our options pricing video, making it earn more than the cat video. In this event, the producer of the latter will be dissatisfied.
Most often, however, we will see two or more funny cat videos being submitted at the same time, yet one of them might make it, while the rest of them don't. This is the best example of subjectivity and luck at play.
This is why SteemQ will remove the monetary information from the view, and replace it with a ranking score. This score aims to a more democratic reflection of community feedback, as well as act sa a more emotionally-neutral way to keep score.
The monetary component will still be there, both on the blockchain as well as on the SteemQ website itself, available for anyone who wants to take a look. However, it will be decoupled from the main view, for the reasons outlined above.
The rewards system could be tweaked to improve curation as well as further democratize influencing factors. This could be achieved through proxies, or a modification of the weight function itself. We would love to experiment with this, whether the idea is simple, such as a linear system with minimums, decreased exponent, range limited sigmoid function, or a more complex mathematical model.
The SteemQ prototype uses IPFS as a core building block of the content distribution system. IPFS is a great tool that does a few things really well. It provides a robust layer for managing, transporting, referencing, deduplicating, versioning and ensuring the integrity of the content. We can secure the multihashes by storing them on the immutable blockchain (ie. STEEM's posts become immutable after first reward payout. STEEM's transfers are much faster, being immutable and permanent within seconds, when the block is confirmed).
This way, we can simultaneously guarantee the ownership and integrity of the content.
We still have several hard problems to solve, in regard to the content distribution itself.
1.) Intelligent distribution and sharding. The amount of storage required to host video files is mind-boggling. We need to develop a system that can distribute the content between thousands of global nodes/gateways in such a way, it is always available at the best possible latency in any geographic location.
2.) Redundancy guarantees. We need to develop a system that can track the state of content replication so that a sufficient number of replicas exist at all times to prevent data loss.
3.) Incentives. We can get pretty far with a P2P model, where every user of the site automatically re-seeds the recently watched videos in the background. This however is not sufficient to guarantee 1. and 2. We need to adopt a mining model that provides content availability and its distribution (bandwidth). New content could pay for hosting with a percentage of its rewards pool, while the old content hosting could be paid for via a smart contract.
What has been done so far:
- SteemQ concept design
- POC app for video uploading and playback
- minimal integration with STEEM's blockchain
- minimal integration with IPFS
Goals and Milestones:
- 1 month: Have a working prototype w/ p2p content distribution
- 3 months: SteemQ website, Limited Alpha Release
- 6 months: Platform enters Beta, Public Release
100% of the funds raised with this announcement will go towards SteemQ development.
Likewise, all consequent SteemQ-related posts will have their rewards sent to @steemq-funds, which acts as a funding source for the SteemQ project.
If you would like support the project with a donation, feel free to send STEEM/SBD to @steemq-funds.
Don't miss out on the next post - follow me.
SteemQ | steemtools | steem.li | witness