Business models for video that can actually work.
The future of online video is unfolding before our eyes. Last month, Livepeer went live on the mainnet with its pioneering distributed network of transcoders. It was a big step for the team, and a huge one for the ecosystem - we’ll start to witness more and more ‘migrations to mainnet’ as the year progresses. The omnipresence of Big Tubes and Flixes may be on the verge of being challenged by their blockchain-inspired counterparts soon. Paratii aims to be an important player in this transition.
This ANN post is longer and more informative than most. Reading through it, you should have a pretty good understanding of what Paratii is all about. If you don’t want to, here’s the
TL;DR:Paratii is a peer-to-peer peer-to-peer curation protocol and modular streaming pipeline that opens up ways of earning a living from video without depending on middlemen. Through a
content management portal, an
embeddable player and the underlying
protocol, Paratii aims to help disintermediate online video and unlock the value of the “creative middle class”.
Why video? Who’s the “creative middle class”?2.1 billion people watch online videos every month. The medium is responsible for ~80% of global internet traffic,
and grows 25% annually. To keep up with demand, creators, publishers & media companies run to the self-publishing platforms (e.g. Youtube) or video-as-a-service providers (e.g. Kaltura) that dominate the market.
Ultimately, Google and AWS are the
de facto private backend for most videos in the web. Both “silent overlords” have the final word on everything from infrastructure-access pricing and privacy policies to algorithmic recommendations. Their service bundles are extremely convenient for customers, although they come with the
corollaries we are already tired of discussing.
Only around 3% of actively monetised Youtubers
make earnings above the US poverty line, today. From 2006 to 2017, the top 3% went from garnering 60% to over 90% of total views in the platform, leaving just 10% of the audience to the remaining 97%, raising questions about the elitist nature of the incentivisation and recommendation policies in place.
Approximately 99% of monetised creators today have less than 2.2M followers, and they arguably make up a lot of the value produced and offered by the platform. We call these the “creative middle class”.
Paratii’s existence addresses the lack of open registries and public monetisation tooling for the online video industry. To claim the ownership of a feature-length movie and go on to distribute and monetise it digitally, in Brazil, for example, a film studio undergoes a months-long, expensive process through a few public institutions. On the other extreme, YouTube allows one to self-publish a video and activate monetisation in a few seconds - the downside being the requirement to waive away 55% of future earnings by default, from the start. The “creative middle class” needs an alternative in between - safe enough, easy enough, fair enough.
Wait… I’ve heard of that Youtube-in-a-blockchain thing beforeWe know. Sometimes it seems like the concept is more hyped up than stablecoins, decentralised exchanges and non-fungible tokens… combined.
It’s no secret that there are
many distinct efforts currently aiming to provide decentralized alternatives to tubes and flixes. The ‘boom’ was iminent for those
watching the space from the beginning. There are a few elements that make Paratii stand out from everything else that’s out there in the ecosystem right now:
- It does not have its own blockchain; it’s also not a specific dApp (the player can be seen as an application, though, just like our alpha portal) - it’s a protocol for distributed curation and an open streaming pipeline built partly on top of an existing blockchain (Ethereum).
- The liberty to choose your own model for monetization. Paratii is not against ads, like DTube, or against micro-tipping, or against paywalls - the only things we’re opposed to are imposing a single model instead of allowing the creator to choose, and taxing transactions.
- The embeddable player that serves as an ephemeral node in the underlying p2p network (wherever it is).
- Everyone who uses the player can be rewarded by the player itself without having to be tied to a single platform or environment, as it’s made to be freely embedded on the web. This is made possible through an in-player wallet.
- Less capital concentration. By forgoing a traditional ICO and prioritizing distributing tokens through minting once the network is live, we avoid giving control over to whales or turning what was supposed to be a democratic system into an oligarchy of sorts.
- Fully open source.
- A specific decentralized curation framework derived from TCRs.
- First and only of its kind in Brazil (+200 million people market, second biggest one for YouTube, after the US).
It should still be noted that practically no video platforms
today are truly 100% decentralized yet, except maybe for torrents. In the blockchain space, we reserve two “honorable mentions” to projects ahead of most:
Livepeer was the first distributed transcoding network to ever go live in the Ethereum mainnet. It is pioneering a service model that suits a specific part of the video streaming pipeline, and we’re excited to follow its evolution as we also experiment with ways to integrate it into Paratii’s flow.
DTube is a Youtube-like portal that monetizes on top of the Steem blockchain - it’s operational, and if you’re just looking to a place where you can escape youtube and make some bucks, we recommend you give it a shot.
With Paratii, we’re trying to achieve a model that permits sustainable revenue for small-to-mid-sized creators, puts in place a truly distributed curation framework, is less prone to valueless tokenisation, and also can tailor to publisher-grade needs.
On the needs of PublishersThe
de facto benefits of distributed-ness in the context of media streaming are largely overhyped.
Storage today can be treated as a commodity: why would a publisher want to store its stuff on clunky, fully-visible IPFS, if AWS costs are negligible and allow for much broader control? However, there is a sector decentralisation can prove itself invaluable:
Peer-to-peer assisted streaming can offload CDN traffic and reduce distribution costs by up to 98%.
We have already released a
JS library to give ETH and IPFS superpowers to developers willing to build video-driven decentralisable applications. We’re also working towards setting up our player so that any interested publisher can simply embed it, pull files from a given source, and save on distribution costs by unlocking the unused bandwidth of their audiences for streaming.
Beyond the hype: this is what p2p-assisted streaming can do to improve traditional CDN-based distribution (from Peer5). Target audiencesFirst and foremost are the small-to-mid-sized
content creators who find themselves dissatisfied with the monetization and revenue sharing options available through current platforms. Replacing rent-seeking intermediaries with public infrastructure permits 100% of transacted revenue from their videos to go directly to them, with no one else taking a slice of the pie, while also allowing for new monetisation models.
Developers can build web or mobile based video-driven applications with paratii.js, our simple but powerful open source library for decentralised streaming.
Our open source technology will also benefit
publishers who wish to use video streaming on their website. They can already easily embed the Paratii player and pull their content from its original source into our pipeline for reducing distribution costs with the assistance of peer-to-peer, and will soon be able to experiment with novel forms of monetization.
Advertisers will also eventually also benefit from increased transparency
by bidding on cheaper, crypto-verified impressions on their ads while inheriting the workflows and practices already used by the industry today.
Most of the Paratii team is located in
Brazil, where we have the support of
Bossa Nova Group, the country’s second largest film production company. Technical know-how from industry leaders in online video (including the engineer responsible for serving live footage from the 2014 World Cup to hundreds of millions) and studio production experts give us familiarity with the needs of our potential markets, while influence in Brazilian media grants a path to growth and adoption domestically; then, internationally.
Technical StructureParatii nodes ‘talk’ a simple peer-to-peer message-based protocol that facilitates passing jobs around and entering types of data-exchange agreements. The system presently relies on IPFS for distributed file storage + delivery, and Ethereum for identity and financial logic execution. The code is running on a permissioned testnet,
in test mode. Paratii is comprised of several different bits of software that interoperate:
Paratii.JS: a library that abstracts the on-demand streaming pipeline with methods and convenience functions for developers to simply
put videos, and
get a playable URL for them, while the whole thing is processed (distributedly!) behind the scenes.
Reading its documentation is a good way to get a grasp of our current architecture.
Paratii Player: a customised Clappr player that can be embedded with a single line of code into apps, webpages, tweets, Medium articles and other social media outlets which leverage oEmbed or Embed.ly (about ~half the internet). It has a built-in wallet and serves as a node in the underlying peer-to-peer network every time it plays.
Paratii Portal: a custom React interface on Paratii.js, making use of the Paratii Player. It is the main interface for creators to open an account, upload & manage videos and for their audiences to browse through content. You can try a demo right now at
https://paratii.video/join. We expect different interfaces (e.g. for niche-content) and applications to push the boundaries of user experience, and envision this portal as evolving towards a creator-focused content management tool.
Video streaming is complex. We’re headed towards fully-fledged decentralisation, but that means breaking oligopolies in a range of closely related services, and it’ll happen gradually. Using IPFS, for instance, incurs particularly challenging issues when it comes to indexing and querying data efficiently. Besides, relying 100% on browser-to-browser communication for streaming is risky, and the safest approach is one that takes into account gateway nodes too.
The
Paratii Portal is ‘where’ token distribution is going to occur (see the token section, below). It’ll happen organically, and, instead of doing an ICO, we want to simply program the network to “boost” reward distribution during a “Token Launch” period before falling towards a predicted rate. This way, early adopters (those who use the platform, upload content, and help curate it) are the truly most benefited group.
Did you just say “no ICO”?This may be unusual, but yes.
We aim to launch our token in Q4 2018, along with a public, mainnet beta of the Paratii Portal. We’ve managed to stay relatively lean, and are privately funded so far. We believe it’s more important to prioritize rewarding the early community than having dozens of millions in cash.
The PTI TokenYouTube pays over 10.000 people to manually review over 2 million flagged videos a month. They do this (what’s the cost like? ~U$20 million/month?) because it’s unaffordable to host illegal or copyright-infringing content. For permissionless file sharing systems (e.g. torrent), it’s a tough problem to tackle. Roughly speaking, the PTI token serves this purpose.
It fuels a stake-based curation machinery (
read more about it here), whose design borrows heavily from that of TCRs (Token Curated Registries). Instead of paying an army of people to do the job, Paratii employs a challenge-and-vote system to determine what gets to stay on the system and whether it abides by the ruleset that is agreed upon. Every user earns tokens upon registration, and publishing a video somewhere requires staking a security deposit behind it.
Newly minted tokens go mostly to active stakers, meaning that if you have a lot of videos published, you’ll get a fair share of rewards, so that, in the long-term, creators who behave should be free to publish as much as they wish, and “bad guys”, who lose stakes and rights to rewards, are disempowered.
Curators, in turn, can stake PTI in order to challenge a video and kickstart a voting round, having the original video’s stake distributed between the challenger and voters (if the video is deemed
rejected) or the challenger’s stake (if the video is
accepted).
This is meant to be abstracted into a separate specific interface, while normal users will likely only see warnings or notifications if anything seems to be wrong with their stuff.
The PTI token may or may not be used as currency for tipping creators, purchasing pay-per-view videos, bidding on media, or subscribing to “Flix-like” services within the system. This depends on the evolution of stablecoins along the next couple months, and on further mechanism design iteration.
The Paratii Foundation’s share of the outstanding token supply (12% in total, distributed over three years through a locked smart contract) will start relatively strong and dilute itself as more tokens are minted. In essence, this means that the power we have to curate and curb infringing content is gradually reduced as the network matures and more people have more PTI.
To get there in good health, we must educate audiences in respect to cryptoeconomic mechanisms (what we like to call
crypto-citizenship), incentivise proper community diligence over videos, help creators experiment with novel monetisation models and welcome any party willing to help us #BUIDL a better future for video and the people who make it.
Current StateWe started assembling as a team in late 2016, and have released 2 versions of our embeddable player and a developer preview of the Paratii.JS library.
We've also opened the alpha portal for public testing, and creators have already started uploading their work and sharing it on Twitter, Telegram and Reddit, where the player can be natively embedded.
In the coming weeks, we will be taking action to further expand the
Early Uploader program: creators who are testing the platform, filling it with original content and pushing its infrastructure to the edge. Promotional materials that explain the concept and structure of Paratii to content creators are already in circulation and we’re working hard to expand our user network.
Our player is in version 0.0.2, with 0.0.3 coming shortly and including features like in-player account management, live view counts and tipping. The portal is getting basic discovery and browsing features implemented soon.
Team, Advisors and RoadmapBounty and Incentives Program (coming soon):Our bounty and incentives program will be focused on three main areas:
1. Content CreationThe success of any video platform rests on the breadth and depth of its content, and Paratii will be no exception. People who apply to the platform as content creators and post their work (be it exclusive or not) on Paratii will be rewarded during the token launch. Rewards will be weighted by views and whether the content is original and/or exclusive to the platform.
If you wish to join Paratii as a content creator, you may begin experimenting at
http://paratii.video/join.
2. Promotion and GrowthThis is where the more standard bounty program actions that involve social outreach and participation come in. Joining our group on Telegram, liking and following us on social media and sharing the platform through a referral program are all part of this category, as well as translation of our material and promotion of Paratii in the media (e.g.
translating our website to Chinese).
3. Technical ContributionsThese are rewards awarded for reporting bugs, making pull requests to fix them and contributing in the documentation of our code and libraries.
More details on the bounty and incentives program and quantification are forthcoming. Feel free to join our group on Telegram to stay up to date on the latest news.
This is an early field with lots of experimentation happening, so don’t hesitate to ask questions, make suggestions or point to stuff we can help clarify.
We’re always happy to answer, and to discuss questions to which we don’t necessarily have answers yet.
