Raven Protocol AMA Recap

Raven Protocol AMA Recap

CryptoDiffer team

Good day everyone! We are glad to meet here

Sherman Lee(@sherm8n ), Co-founder of Raven Protocol and Rahul Vishwakarma (@curious_rv) Co-founder of Raven Protocol

Let’s CryptoDiffer x Raven Protocol AMA get started!

 

Sherman Lee
Co-Founder

Hi everyone. We are so excited to be here for the AMA!

Let’s do this!

CryptoDiffer team

We are glad to have both oof you with us today!

Q1. Can you introduce yourself to the community? What is your background and how did you join Raven Protocol?

 

Sherman Lee
Co-Founder

Hello CryptoDiffer community! I am Sherman, one of the co-founders of Raven Protocol. Thank you so much for having us here today. It means a lot as we’ve seen CryptoDiffer cover many projects early and well before the hype. Great community to find the latest news in the blockchain industry. My background is in Computer Science and I’ve been doing machine learning for 10 years now. One problem unraveled after another and here we are building a distribution framework for faster AI training.

 

Rahul Vishwakarma
Co-Founder

Hey everyone, I am Rahul, also a co-founder of Raven. I have over a decade worth of of experience in AI/ML, and previously open-sourced an NLP library. I founded  Mate Labs  where I built a ML platform for business professionals who can train models without writing a single line of code. User growth was coming fast and cloud compute costs started soaring. Thus, I joined hands with Sherman to solve the same problem we both had.

Glad to be here

 

CryptoDiffer team

Q2: Can you briefly describe what is Raven Protocol in 3-5 sentences?  What technology stands behind Raven Protocol and why it’s better than the existing one?

Sherman Lee
Co-Founder

Raven is creating a network of compute nodes that utilize idle compute power for the purposes of AI training where speed is the key. AI companies will be able to train models better and faster.

We developed a completely new approach to distribution that speeds up a training run of 1M images and brings it down to a few hours. We solve latency by chunking the data into really small pieces (bytes), maintaining its identity, and then distributing it across the host of devices with a call to action: gradient calculations.

 

Rahul Vishwakarma
Co-Founder

Other solutions require high-end compute power. Our approach has no dependency on the system specs of each compute node in the network. Thus, we can utilize idle computer power on normal desktops, laptops, and mobile devices allowing anyone in the world to contribute to the Raven Protocol network. This will bring costs down to a fraction of what you need to pay for traditional cloud services. Most importantly, this means Raven will create the first truly distributed and scalable solution to AI training by speeding up the training process.

Our consensus mechanism is something we call Proof-of-Calculation. Proof-of-Calculation will be the primary guideline for the regulation and distribution of incentives to the compute nodes in the network. Following are the two prime deciders for the incentive distribution:

Speed: Depending upon how fast a node can perform gradient calculations (in a neural network) and return it back to the Gradient Collector.

Redundancy: The 3 fastest redundant calculation will only qualify for receiving the incentive. This will make sure that the gradients that are getting returned are genuine and of the highest quality.

 

Cryptodiffer Community

Q3: What are the major milestones you have achieved so far?

 

Rahul Vishwakarma
Co-Founder

We’re building a distribution framework from scratch. Thus, we needed to first complete the foundational implementations. This felt a lot like building an operating system from scratch. Things we take for granted such as file systems, memory management, and scheduling had to be built from scratch. We needed to build the foundational implementations of the distribution framework. Both the Raven team and the community executed on that successfully.

Foundational Implementations 

Development of functions such as Back-Propagation (Back-prop) and several activation functions that are used to train vanilla Neural Network models have been completed

Convolution Layer implementation

RNN unit implementation

In 2020, we are ready to leverage all the work we’ve done as building blocks 

Development

Open-sourcing the first set of Function Implementation on Github

Laying the foundation for a group of open-source contributors around the world, and kick-starting the evangelist community

Strengthening the CNN and RNN implementation

Staking and launching a liquidity pool

Opening a dev center in Bangalore, India

Business

We have partnerships in place. As mentioned in an AMA, we can’t announce it yet, but we’re working with companies on very interesting use-cases like building trust models in Africa and chain analysis in China.

We will start doing active pilots after initial tests are done on the test net.

3 companies have agreed to be our Pilot customers

We are actively speaking with exchanges to expand the reach and utility of RAVEN

 

Cryptodiffer Community

What is your long-term vision about the industry which RAVEN is working at? Are you afraid some day there will be another project with more innovative technology can replace RAVEN?

Rahul Vishwakarma
Co-Founder

To proliferate ML into society. We’re not afraid that someone will come and steal the limelight from us, we will actually anyone and everyone who wants to contribute.

 

Cryptodiffer Community

Raven Protocol is working on Deep Learning, Artificial Intelligence which are helpful for Programmers & developers ! But, whats the main use-case of Raven protocol to Normal users or Investors like me?

As raven protocol is working on Blockchain! So, Whats the use of Blockchain technology in Deep Learning , artificial intelligence, etc?

Rahul Vishwakarma
Co-Founder

To be honest, Raven is being built for Data Scientists and developers, right now for Non-Devs we have Mate Labs, where we help Nondevs build Predictive Analytics and AI models.

Cryptodiffer Community

What is the importance of developers and the community in the RAVEN ecosystem and how easy and convenient it is to connect and apply RAVEN technology in the existing enterprise infrastructure?

Rahul Vishwakarma
Co-Founder

Developers will be building the back-bone and application on top of Raven Infrastructure. leveraging the technology to train deep neural nets faster and efficiently as ever. Currently the right market for Raven will be for early-stage companies and SMB market. Enterprises usually have infrastructure almost sorted.

 

Cryptodiffer Community

As I see there are many crypto projects came that also works on AI Technology with Blockchain! But the question is – what role Blockchain plays in AI world? Also, Can you explain the benefits of Smart AI Contracts?

Sherman Lee
Co-Founder

AI/ML definitely needs to be decentralized. Blockchain will transform the AI industry. It enables the sharing of data, research algorithms, and compute resources as people are incentivized to provide the resources they have to others in the community.

Cryptodiffer Community

AI/ML is a very complex area which requires more competent hands. With 10 engineers currently at Raven protocol, how were you able to achieve such huge milestones? and don’t you consider hiring additional AI/ML researchers to assist you with this complex work of yours?

Rahul Vishwakarma
Co-Founder

Oh we are always hiring. Please send the CVs of competent AI/ML engineers if you know one.

 

Cryptodiffer Community

What approach does Raven adopt in eliminating latency and scalability issues?

Rahul Vishwakarma
Co-Founder

1. Apart from having individual contributors we would also be building relationships with server farms to build redundancy in the system.

 

Cryptodiffer Community

If AI continues to augment the work of humans, my fear is this, in the next 20 years and beyond.

1. Won’t AI displace humans completely in various work places since it can perform same tasks as humans in a more accurate way?

2. Don’t you think AI won’t pose a major threat to human world that people would start working against AI/ML companies or projects like Raven?

Sherman Lee
Co-Founder

Augmenting work is the key word. It helps us humans become more efficient in the tasks we want to accomplish. Like analyzing large amounts of data. AI will eliminate some jobs for sure. However, it will also replace those jobs with new ones. People will be hired to manage the training of AI models instead of doing manual work. AI could definitely pose a threat humanity. Many people are working on the ethics of this behind the scenes.

 

Cryptodiffer Community

What is RAVEN PROTOCOL plan to compete with AWS, GOOGLE CLOUD, GOLEM ? and what’s the latest RAVEN PROTOCOL plan in the near future?

Sherman Lee
Co-Founder

Amazon AWS, Google Cloud, Golem, etc., offer general cloud compute. We focus specifically on the speed of AI training. Thus, building our own distribution framework allows us to bring down the AI training time down. The community will play a critical role in pulling the power out of the hands of mega-corps and putting it into the hands of the community.

Cryptodiffer Community

What is your business model? Who are your potential customers and which markets are you targeting on?

Rahul Vishwakarma
Co-Founder

Business model: A percentage of all the transactions that’ll take place on the platform. SMBs and Startups

Cryptodiffer Community

What areas of the world are Raven focussing on for your services and what is your strategy about that? Focussing small and rolling out bit by bit? Or are you rolling out worldwide? Can you tell a little bit about that please?

What do you do to protect users against Cybercrime and to secure data?

Rahul Vishwakarma
Co-Founder

Focussing on small and staying lean. As simple as it gets. Harder to execute. 🙂

Cryptodiffer Community

What are the advantages Raven has been achieved since the migration to Binance Chain?

To what extent have been the effects of the recent bearish market on RAVEN platform? Have the recent poor market conditions hindered or stunted the platform’s progression and goals to be achieved in the nearest future ?

Rahul Vishwakarma
Co-Founder

One of the biggest advantages of building on Binance Chain is help from the Binance family. The Binance Chain team itself and the projects who have migrated help each other. In the past month, we received so much advice from teams like Thorchain and Matic. That support is invaluable. The market conditions are tough however it has not stopped us from executing. We learned to stay lean, get a lot done, and in a short amount of time.

Cryptodiffer Community

Adoption is one of the important factor that all sustainable blockchain projects should focus to be more attractive in the investors’ eyes. What Raven Protocol has done and plans to do to achieve real world adoption?

Sherman Lee
Co-Founder

Those that have been following Raven from the beginning know that our core focus is on real adoption. Thus, we started from use cases that customers actually have. And we are building to solve those real use cases. I know this is different from many blockchain projects who are building amazing technology, but looking for people who need their solution. For Raven, we already found those who need us!

Cryptodiffer Community

How does the Distributed Training ensure the consistency around all the Node in Network?

How does the Dynamic Graph Computation help Raven?

Rahul Vishwakarma
Co-Founder

We always have redundancy of 2, where we make sure the weights are being calculated are correct. Dynamic graph ensure that. What we mean by it is, that the load per unit is allocated dynamically to ensure speedy training

Cryptodiffer Community

How much time will it take to sync all the node with new data? How you are going to mitigate the latency issue?

How does the Dynamic Graph Computation help Raven?

Sherman Lee
Co-Founder

Conventional distribution methods transfers large chunks of data between nodes. Raven breaks it down into bytes so it eliminates the latency issue.

We can achieve this feat because we take both Data and Model Parallelization approaches to form a different model of distribution.

A major difference between static and dynamic computation graph is that in the former, the model optimization is preset, and the data substitutes the placeholder tensors. Whereas, in the latter the nodes in a network are executed without a need for any placeholder tensors. The benefit of a dynamic graph is its concurrency, and it is robust enough to handle the contributor addition or deletion, making the whole Raven training sustainable.

Raven is thus capable of eliminating the latency and scalability issues, with both the approaches. Hence, distributing the training of any deeper neural network and their larger datasets, by getting rid of the added dependency on the Model replication. Data is also sharded in smaller snippets.

In fact, the Model is intact at the Master Node, and the heavy lifting is distributed in the tiniest snippets of data subsets over the network of contributors. The resultant gradients, after the calculations that happen at the node/contributor ends, are sent back to the Master Node.

This creates a ton of difference, as it is easier for calculations to pass through from machine to machine, rather than creating multiple replicas of a complicated Model.

Cryptodiffer Community

Q1: $RAVEN is in the top 10 the addresses with a balance above 100 USD per asset on Binance Chain. How did Raven achieve this in a short time? What is the recipe for success? How motivates its believers to hold $Raven long-term ?

Q2: When you looking back on 2019 and before, what did Raven achieve? Which STEPS are you most PROUD OF? Can you tell us about your important steps in 2020 and what difficulties and challanges do you think the Raven Protocol will face in 2020 and the coming years?

Sherman Lee
Co-Founder

1) Hard work to be honest. It was not an overnight success. We built this thing from the ground up. People who were part of that journey from the beginning continued sharing our successes and visions to ppl in their own community. This is what motivates long-term believers. They see us execute on what we wanted to accomplish and it gets them even more excited!

Cryptodiffer Community

2. What are the benefits and adoption that Raven Protocol really brings to the community? What is the motivation for the project development? 

3. What are your major goals to achieve in the next 3-4 years? Where can we Raven Protocol ecosystem in this period? What are your plans to expand and gain more adoption?

Sherman Lee
Co-Founder

2. It allows ppl train AI models better and faster. The motivation came out of our own problems doing AI training. It took a long time and was expensive 🙂 3. Working with real customers and gaining adoption through expanding use cases. Similar growth path to many enterprise software companies.

Cryptodiffer Community

1. The computational speed, on which Raven focuses, depends on the number of network nodes involved. What are the main incentives for a network member / node holder in terms of monetization and prospects in general? At what expense should the Raven network outgrow mega-corps involved in this area and hold the lead?

2. AI algorithms are touching many aspects of human rights. As a result, many countries are already trying to legislate the use of AI. Im curious what do you think in this matter. Will the introduction of restrictions hinder the development of technology?

Rahul Vishwakarma
Co-Founder

1. The incentivization will be proportional to the number of calculations a node makes. 

2. Regulations never hinder the development. It is always better to ensure nothing bad is happening.

 

Cryptodiffer Community

Does Raven Protocol stacking?

What are the conditions to be able to run Node and get Raven Protocol reward?

What is your plan to fund the marketing campaign for Raven Protocol that will be needed to gain visibility and build a large community user base?

Why should we buy this token? What is the advantage of holding this token for long and short term period? Can you give me 3 key reasons why I should invest Raven Protocol long term?

Sherman Lee
Co-Founder

We did run an experimental proof of steak program. We are working on launching another iteration of that soon. We will be open-sourcing some Function Implementations to grow the community of developers and evangelists. 1) You should buy this utility token for access to the Raven Protocol network 🙂 2) If you believe that AI will be an important technology, we are building a framework to speed up AI training. 3) If you want to be part of the journey in seeing us reach global scale

Cryptodiffer Community

Explain the utility and Real Life Use case of RAVEN token.

Sherman Lee
Co-Founder

Think we explained part of this in Q1-3, but…

Raven is creating a network of compute nodes that utilize idle compute power for the purposes of AI training where speed is the key. A native token is the key to bootstrapping a nascent network.

We want to incentivize and reward people all over the world to contribute their compute power to our network. Additionally, we will reward token holders for running masternodes which will be responsible for orchestrating the training of various deep neural networks.

Our consensus mechanism is something we call Proof-of-Calculation. Proof-of-Calculation will be the primary guideline for the regulation and distribution of incentives to the compute nodes in the network. Following are the two prime deciders for the incentive distribution:

Speed: Depending upon how fast a node can perform gradient calculations (in a neural network) and return it back to the Gradient Collector.

Redundancy: The 3 fastest redundant calculation will only qualify for receiving the incentive. This will make sure that the gradients that are getting returned are genuine and of the highest quality.

 

Cryptodiffer Community

Raven Protocol looks to be set to revolutionize the AI space. How do you visualize this revolution will impact and forever change the numerous industries of the world that already or have yet to utilize advanced AI?

Rahul Vishwakarma
Co-Founder

By bringing ML to the masses, which will help discover more use-cases for general audience.

 

Cryptodiffer Community

Where do you and the   raven protocol team see the world of blockchain being in years to come, and how will Team change the landscape of the cryptocurrency space in years to come?

Sherman Lee
Co-Founder

Blockchain is an important technology. Raven certainly will push the industry forward as we are one of few companies who has a focus on working with real customers with revenue.

 

Cryptodiffer Community

What do you think is the biggest problem @raven_protocol will solve that no other project is solving now, and why is the problem important to solve?

Sherman Lee
Co-Founder

Speed of AI. it will unlock creations that were just never possible before due to sheer volume of data

 

Cryptodiffer Community

RAVEN have definitely achieved a lot. Could you give us a summary on what RAVEN did in 2019?

Cryptodiffer Community

1)Which are the aspect are considered by the Proof of Calculation consensus mechanish to incentivize and reward people for contributing their CPU power to the network? 

2) Raven protocol always try to accelerate the development and adoption of AI technologies and Deep/machine learning. Which are you plans to contribute the improvement for the AI technologies to the big markets? 

3)Do you have plans to create a stablecoin wallet with AI? 

4)One of your objectives of 2020 is the strengthening of the CNN and RNN implementation, so which method or strategy you will use to perform to do it?

Sherman Lee
Co-Founder

We have thought about the stable coin problem a lot for AI services. Thinking long-term, we want to allow people to chain together services like data tagging, cleansing, training, etc. A stable coin or atomic swap could be a good way to bridge together all these services

 

Cryptodiffer Community

How would you simplify and explain what raven protocol is to someone who does not understand AI, Blockchain or Cryptocurrency?

Sherman Lee
Co-Founder

One of our community members had the perfect explanation to his 5 year.. 

“While creating logo’s for #RavenFuture my 5yr old decides to color them in.

She asked what was it..

how do you explain the complexities of machine learning to a 5yr old…

So i said…

Me: know how daddy loves computers?

Kid: yea

Me: well this helps daddys computer be better…”

 

Cryptodiffer Community

What are the main markets that Raven Protocol is focusing now? What is your strategy to access to these markets?

How many wallets that can support Raven Protocol until now? And do you have any plan to build your own wallet?

Security in every transaction is something that everyone needs. What steps have Raven Protocol taken so that security can be maintained?

Sherman Lee
Co-Founder

We are focusing on China, India, and Africa for user adoption. The AI industry is growing fast in these markets. We want to make sure we are in there early.

You can find us on Trust Wallet, Atomic Wallet, and Binance DEX itself. 

Security is something binance chain focuses on a lot. There’s no smart contract so that helps with security 😉

 

Cryptodiffer Community

How the demand of your token will increase in the future? or what steps would you take to increase the demand?

Sherman Lee
Co-Founder

The demand for token comes from customers themselves. They will need to use RAVEN to access the network

 

Cryptodiffer Community

1. One of the best facts of raven is than raven doesn’t depend on the architecture of each compute node in the network, and can utilize idle computer power on normal desktops, laptops, and mobile devices.But the obtained reward  still depend of the  device contribution or just for the online time? 

2. What can be cosidered your competitive advantage over your competitors? specifically  with the centralized industry leaders such as Amazon AWS, Google Cloud, Microsoft and Facebook? 

3. which companies have you working or have you applied AI/ML?

4. Apart of China, India and Africa, which others continents or contries you will try to enter? maybe Europe or South America?

Sherman Lee
Co-Founder

1. It depends on the calculation done by your device

2. We focus on speed of AI training

3. We will be announcing the partnerships in due time

 

Sherman Lee
Co-Founder

Thank you so much Maks for being a wonderful host and inviting us for the AMA. Thank you to everyone in the Crypto Differ community for the great questions and spending the last 1.5 hours with us

 

Cryptodiffer TEAM

Thanks a lot for your time and great answers! 

Wish you luck with Raven Protocol and looking forward to see more updates from you!

Spread the love
  •  
  •  
  •  
  •  
  •  
  •  
  •  
  •  
  •  
  •  
CryptoDiffer Admin
query2212@gmail.com

Leave a Reply

Your email address will not be published.