My Blockchain Startup Journey — Reflections of a Tech Entrepreneur

I am a person who loves technology, gadgets, and all things geeky. I am also the type of person that wants to change the world.

Part 1: My Experience Creating PPLive/PPTV

When I was in college back in 2004, my friend and later partner, Bill Yao, informed me that there was no place on campus to smoothly watch NBA games over the Internet. He was wondering if we could develop a real-time live video software using P2P transmission technology so we could watch the NBA. Not exactly the most common type of thinking when faced with the inability to watch basketball. The idea behind it was similar to the then-popular BitTorrent software. I eagerly agreed to work with Bill, both so we could watch the NBA and because deep down I could not resist the temptation of this technical challenge. From here, the humble beginnings of PPLive, later PPTV, were formed. Bill served as the Founder and CEO, while I become the Chief Technical Architect and focused on the development and advancement of P2P transmission technology.

The first thing PPLive achieved was real-time streaming via peer to peer transmission technology. If you’re not quite clear why this is, it’s the ability to upload and download data at the same time, which makes live streaming possible.

In the beginning, I led the team to build a live broadcast platform with P2P. With many people sharing the same data during a live broadcast, we managed to achieve:

• Achieved a bandwidth saving ratio of more than 99.9% during simultaneous streaming of popular content
• Only 10Mb of bandwidth released to support 10 million simultaneous active users
• Average start time: 1.2 seconds
• Average interruption count: 1.6 seconds per hour
• Total network delays from a broadcast source: 90 seconds (max)

After we developed and reached a level of advancement that guaranteed a high Quality of Service (QoS) for P2P live broadcasts, we changed our focus towards P2P Video on Demand (VOD) as it was becoming bigger and bigger on a daily basis. P2P VOD and live broadcasts are different from each other. The former needs to use hard disk storage and memory cache, whereas the latter is better suited for delivering popular content. PPTV’s VOD too achieved an excellent QoS:

• A bandwidth saving ration of 90%
• Average start time: 1.5 seconds
• Average interruption count: 2.2 seconds per half an hour
• Average playtime when seeking a playback position: 0.9 seconds

Subsequently, I led the team to port the P2P kernel to the embedded system. We streamlined the P2P transmission protocol and made a number of optimizations in terms of performance. The read and write of an embedded hard disk are not used for ordinary mechanical hard disks. Therefore we created a lot of protection measures to make the P2P core run on embedded devices. Afterward we quickly launched an iOS client, an Android mobile phone client, and finally an Android set-top box client, all of which supports P2P.

Amid the growing popularity of smartphones, video creation became easier than ever before leading to the rise of user-generated content. I led a team to create a cloud-based broadcast service, which gave each user a certain amount of free storage space. On this platform users could freely upload, download, and share video content, In fact, it was very similar to network hard drives such as Google Drive. The difficulty found in sharable network hard drives is similar to that of VOD, the major difference being that the amount of videos greatly exceeds that of VOD. We have put a great amount of work onto the optimization of this particular area.

In 2013, PPTV was sold to Suning Yunshang for US$420 million. A year later, I waved goodbye to video and P2P technology, which I had been doing for over ten years. To pursue a bigger dream, I started a new path with a new startup of mine.

Part 2: My Experience With JDO

After the success of PPTV, I still maintained my interest in technology and all things geeky. However, this time I chose to venture to a new industry unrelated to video and P2P. I began to devote myself to intelligent hardware and artificial intelligence. In 2014, I co-founded, with Cloud Wong, JDO — which produced car accessories that implement AI technology and are connected to the Internet. Over almost four years, I served as the CTO at JDO, and accumulated in-depth technical experience in intelligent hardware, embedded software, artificial intelligence, and machine learning.

The automotive industry is very closed. For everything we do, we needed a long business negotiation cycle, and the product development cycle is limited by each iteration of cars launched. In an environment where we must cater to other businesses, I could not entirely follow my ideas to implement innovation where it was needed. JDO, as a business, was doing very well in the automotive field, where we received orders from many internationally renowned car manufacturers. However, in order to pursue a bigger dream and to change the world, I decided to leave JDO.

Part 3: My Thoughts on Sharing Storage

Because I was working on artificial intelligence every day, I completed a number of neural network education and training programs and I have established cooperation and communication with many other AI companies. To do neural network training, high-end NVIDIA GPU graphics cards are required; such graphics cards are also costly and most of the time they lay idle. Always asking myself different questions, I wonder is there an opportunity to build a shared GPU graphics platform that encourages everyone to rent out unused GPU resources? In other words, when I need it, I pay to use other people’s unused GPU resources; when others need it, they pay to use my remaining GPU resources. The benefits are numerous but thinking of startups, it would mean they don’t need to buy so many GPUs at once. Their costs will be reduced and companies that do have GPU resources can use it to make money.

I ran with this idea of sharing a GPU graphics platform, made a simple Minimum Viable Product (MVP) to verify my thoughts, and let several cooperative AI companies use this MVP. However, I found that it was barely used. After I learned about the situation, I realized that although AI companies like lower prices by sharing GPUs, they were more concerned about the security of data than with price. An AI company’s data collection often requires a high cost to collect data by competitors. More importantly, the neural network does not have an advanced nor efficient way to split into multiple computer calculations, encrypt the data in time, and then have the data decrypted before entering the neural network. On a technical level, there is currently no good way to decentralize the neural network model, likewise, there is no good way to send cryptographic data directly into the neural network calculation. Due to this obstacle plus other reasons, I gave up on this particular idea.

But when one door closes, another opens — P2P storage suddenly came to mind. Is it feasible to only share hard disk and bandwidth, and not GPU? Three things came to mind:

1. As we are entering the Internet of Things era, a large number of households have idle computers which not fully used.
2. A large number of households purchase bandwidth on a monthly basis but do not fully use it.
3. A large number of IoT devices have storage capabilities.

If these vast amounts of idle resources can be fully utilized, then society will reap the benefits.

Whereas with PPTV, users voluntarily shared content without incentives, times have changed and the rise of sharing culture necessitates a type of incentive model. I envisioned a platform where users pay a fee, while other users provide services to earn income. A type of digital Uber or Airbnb, but instead of drivers and landlord providing the service to passengers and rentees, it would be unused hard disk space and bandwidth.

Sharing storage would provide a number of key benefits to users:

1. The cost is lower because, in terms of cost structure, it is simply the reuse of idle resources.
2. P2P can speed up the transmission of data, and PPTV’s success is the best proof of that. With regards to shared storage, because storage nodes are located everywhere on the network, the data would be stored in the nearest storage node so it can achieve the fastest transmission.
3. Decentralized storage is actually more private. Because big companies don’t necessarily guarantee privacy, almost all big companies have exposed data leaks. Big companies today rely heavily on big data and may use user data for AI training. Decentralized storage double-encrypt data through the user’s key and developer key, and then slice it into several segments. Different segments are stored on different machines. Moreover, the sharing platform itself does not store data, which completely prevents data breaches. Only the user’s key can open the data and those holding fragments of the file have no way of accessing it. Furthermore in the case of hackers, hacking into one computer to access data is useless, because the data is stored across many different computers. This is a contrast to centralized storage which is vulnerable to hackers as all data is stored in one location.

Although shared storage offers several benefits, it also brings many challenges.

1. How do you guarantee the stability of nodes holding onto data? Some nodes won’t always be online to transmit data. How do you offer a stable service based on unstable nodes?
2. How do you guarantee the quality of the storage nodes? For example, the hard disk of some nodes may be old or can easily break. What if the bandwidth quality of some nodes are terrible? How does a storage platform deal with nodes that experience frequent power outages due to their location?
3. How do you deal with bad players or malicious nodes? When people have something to gain (in this case payment from a service) there will always be those that try and game the system.

Although these problems are challenging, I believe my decade of experience with international P2P projects, and variations of these problems occurring and being resolved with PPTV, I am very confident that my team and I will be able to solve all of them.

But with the successful creation and implementation of such a system, why is it that we believe that our income distribution will be fair. Well, this is the value of blockchain technology.

Part 4: My History with Blockchain Technology

Starting in 2010, I was exposed to the now infamous Bitcoin. Given my interest and experience with P2P technology, I quickly read through Bitcoin’s code: block, block hash, nonce, and mining algorithms, all of which I quickly understood. However, at the time I could only recognize the greatness of Bitcoin’s technology. The application of that technology for business eluded me.

Later, Ethereum was born. In my opinion, in addition to optimizing Bitcoin’s block performance and mining algorithm, Ethereum’s greatest improvement is that Bitcoin can only execute simple instructions OPCODE into complex logic code, whereas virtual machine code and can be programmed in solidity high-level language. The Ethereum blockchain can be used not only for digital currencies but also for writing a variety of smart contracts that work in many scenarios.

Finally, after carefully studying the technical mechanism of the open-source Hyperledger Fabric chain from IBM, I finally found the value of the blockchain itself.

From a technical point of view, blockchain is essentially a distributed and fully synchronized database. Since the chain is scattered on different computers, a dispute resolution method is needed, and consensus algorithms are employed to resolve any disputes between the distributed ledgers. From a sociological perspective, the decentralization and consensus of blockchains can generate enormous social value. The value is that the blockchain can build real trust. Furthermore, the value is that the public blockchain can build true public trust.

With regard to sharing storage, I think it is wrong to store data on the chain. The function of the blockchain for shared storage is not to solve the problems of data storage itself, but rather to solve trust problems. The data stored in the chain should be trust-related data, such as user assets, storage contracts, proofs, rewards, and penalties.

Part 5: Refining My Thinking

With blockchain technology, an effective third-party node certification mechanism can be established to monitor cheating or dishonest nodes by consensus. With blockchain technology, the rules of sharing cannot be changed according to one’s will; in that fact, there is fairness and transparency.

Because of fairness and transparency, everyone can safely invest resources in the platform. Likewise, everyone will be willing to invest funds and build a large, reliable, and competitive platform that provides services for incentives. Lastly, there will be a truly transparent marketplace that drives everyone to look for better and more affordable resources in order to provide users with better and affordable services.

Because of fairness and transparency, we can build an economic model that is implemented purely by computer programs. For bad players and malicious nodes, there is now a mechanism in place that will ensure collateral invested in the network will be reduced to 0, thus disincentivizing them from cheating. For nodes with stability issues that impact the platform, there is now a mechanism that will deter these types of nodes from offering a weakened service. This, of course, helps incentivize nodes to provide a stable service. That said, we do understand that nodes may occasionally fail which is why there can be a warning penalty to serve as motivation to upgrade to more stable hardware. We believe that with the right incentives to get the right nodes to participate and the right penality to disincentive the wrong nodes, the entire platform will be stabilized and QoS will be guaranteed.

What needs to be done is:

• Create a sharing storage network that utilizes P2P and blockchain technologies
• Use incentives to stimulate sharing
• Use the blockchain to ensure fair and transparent incentives

The ultimate value of this is a platform that is based on affordability, speed, and privacy. This network will be easy to use and is friendly for developers wishing to create a variety of applications in numerous computer languages.

Part 6: The Launch of PPIO

In the days before PPTV, it was partner Bill that approached me with the idea. This time around, I was the one who approached Bill and told him that I wanted to help change the world. Together, we decided to launch this new project in order to utilize the world’s unused storage and bandwidth in order to it better. This is PPIO — a decentralized data storage and delivery platform for developers that values speed, affordability, and privacy.

To learn more about what we’re doing, visit our official home page or read our articles on Medium that detail the thought we’re putting into the platform. Or you can follow us on Twitter or speak to the team over at our Discord.