Decentralized Machine Learning ICO Analysis: Making Machine Learning More Accessible For Everyone
A decentralized economy for democratizing the machine learning industry and providing unhindered access to private data without compromising data owners’ privacy.
Artificial Intelligence and Machine Learning technology has come a really long way and is set to become a part of the coming era’s foundations, but the problem is that this technology is currently not available to everyone. Machine Learning requires a significant amount of computational power and access to a variety of data in order for it to become developed enough to produce practical and usable applications. Currently, machine learning technology has open access to public data only, people simply are not ready to share private data with developing parties for a number of reasons.
This limited access to private data is currently a major obstacle in the path to machine learning development, Decentralised Machine Learning (DML) is a crypto venture that is being worked on to knock down this barrier. DML is going to take the unique qualities of blockchain technology to allow machine learning developers to gain open access to private data without compromising the data owner’s privacy and by incentivizing the sharing of private data through the DML project’s native token (DML token).
The goal of DML is to make private data easier to access and to make machine learning less resource heavy by providing people with a decentralized ecosystem in which they will be able to share data, computing power and algorithms, making the machine learning industry fairer for everyone.
Why is DML Relevant?
Due to machine learning development requiring access to large amounts of data and computational power, there are very few technology companies out there that can actually afford to work with this technology. In its current state, the machine learning industry is pretty much an oligopoly, making progress and development very slow since there is very little pressure from competition.
There are two factors that affect the development of machine learning technology; the quality of data being used in the machine learning process and the quality of algorithms being used to create applications and for carrying out data analysis. Right now the few technology giants that do have access to private data either have to go snooping through people’s smartphones and computers (which is unethical) or they have to provide people with some sort of incentive in order to persuade them to share private data. Both of these methods provide companies with limited access to private data, greatly reducing the overall quality of the data that they use in their development process.
The overall quality of machine learning algorithms leaves a lot of things to be desired, the machine learning community simply is not open enough at the moment to make combined learning possible; parties with the expertise to produce these algorithms lack innovation since the overall industry is in a stagnant state.
The DML project is planning on addressing all of these problems by providing the machine learning industry with a platform on which data owners, algorithm developers, and technology companies shall be able to interact with one another to buy and sell access to data, computational power and develop/share algorithms with one another through crowdsourcing.
The main goal of DML is to provide open access to private data in an ethical manner that does not compromise people’s privacy and to level out the machine learning playing field so that everyone has an equal opportunity to participate in this growing industry.
The DML development team believes that the machine learning market has an immense amount of potential not only for tech companies and developers but also for regular people who have private data. Private data is in great demand within the machine learning industry, meaning that people have the opportunity to take private data and sell it to companies that are in need of it. The DML project is going to provide people with the means to sell their private data to tech companies in a manner that will keep their privacy intact.
A really promising thing about DML’s solution is that it will not require people to hand over copies of their data to companies at all, they will only have to let companies access their smartphones, computers and other devices on which their private data is stored. To keep every data owner’s privacy intact, DML will only let companies access private data by running their machine learning algorithm on the devices, this will ensure that companies shall only gain access to the output of the algorithm and the data owner’s privacy will not be compromised in any way.
The results from these algorithms will be used to form unbiased and precise crowdsourced predictions and analytics. DML’s solution is also going to address issues with computational power as well since algorithms will be getting processing power from multiple data owner devices simultaneously.
The blockchain based DML ecosystem will be able to protect people’s privacy while allowing unrestricted access to private data, it shall provide developers with access to more processing power by borrowing idle processing power from devices that shall be a part of the ecosystem. It is also going to create a marketplace backed by a community of developers where parties shall be able to buy and sell algorithms and bring innovation to the machine learning industry.
Democratized access to private data and computational power along with the creation of a machine learning algorithm marketplace will encourage more people to start participating in the machine learning industry, giving the industry a much-needed surge of competition.
The DML community will play a vital role in bringing developers together and letting them discuss, share and innovate. DML developers will have an opportunity to work as freelancers and provide people with ready to use algorithms, fine-tuning and other services to earn revenue.
DML will ensure the privacy of data owners by eliminating the need for data extraction and by giving data owner’s complete control over their data’s accessibility. Data owners shall be able to define what kind of data they want to make accessible through the DML Protocol and what kind of data shall be off-limits, and once shared data has been analyzed by an algorithm, only its results shall be transferred to the other party through a secure node.
The DML Marketplace
This marketplace will have developers and machine learning tech companies selling ML algorithms to customers such as research bodies, corporate agencies, governmental and private organizations and individuals. Basically, anyone wanting to get analytical predictions about any kind of data will be able to come to this marketplace and buy algorithms that meet their needs. Once they have the algorithms that they want (bought or made by themselves), they will be able to run them on devices that are available on the platform, using private data and idle processing power provided by the devices.
The process of pairing algorithms with devices carrying the right kind of private data shall be managed by the DML Protocol, this will ensure that customers get accurate and relevant results every time. The DML protocol will connect developers and algorithm runners with various decentralized nodes that will be connected to multiple data owner devices, this process will ensure that every algorithm gets access to a huge amount of data (and an ample amount of processing power) from several data owners at a time in a highly stable and efficient manner.
Customers wanting to gain access to private data and idle processing power will need to pay in DML tokens, the DML Protocol will calculate the availability of required data and other factors to come up with a price that shall be presented to the customer. Smart contracts will be used to ensure fair and unbiased governance between parties and will also enable the DML marketplace to allow peer-to-peer transactions without the need for any middlemen.
The DML Token
General Information: there shall be a total of 330,000,000 DML in circulation, out of which 54.2% shall be distributed amongst the public (36% in ICO, 9.9% as rewards for data owners, community contributors and node developers, 8.3% as a bonus reward for the developer community). 2.5% of the total tokens shall be set aside for marketing and PR partners, 8.8% shall be kept for business and research partners, 15% shall be kept in reserve and the remaining 19.5% shall be kept by the advisory team, development team, and early contributors.
Purpose: the DML tokens will have several roles to play, they shall be used to pay data owners who will share their private data and processing power on the DML platform, incentivizing their participation. These tokens will also be used to incentivize the development of the various nodes which are going to be at the very core of the DML Protocol’s operations, motivating all the decentralized nodes on the platform will ensure that the platform is always responsive and has the potential for scalability.
Anyone wanting to buy algorithms on the marketplace will need DML tokens, the tokens shall also be used to reward developers who contribute to the fine-tuning of existing algorithms. An interesting use of the DML tokens will be to form a user growth pool that will be used to reward early contributors, developers and data owners who become part of the DML ecosystem in the early days will be able to benefit from this bonus. This step will help in attracting people to DML and accelerating the platform’s early adoption.
Value: the DML tokens will be an essential part of the DML ecosystem since they are going to be used in a variety of ways, while I do not have a lot of information available about their ICO yet, their numerous purposes will definitely increase the token’s demand. Also, since the tokens are an integral part of the DML ecosystem, their value shall appreciate as the platform will grow.
There is not a lot of competition for DML at the moment, one project that is trying to accomplish the same as DML is Steamr; a data marketplace where people can go and trade their data with parties that need it. Steamr sounds quite similar to DML, however, Steamr’s solution requires people to actually have to upload their data onto the marketplace, meaning that once it is sold to someone they can go through their data. This results in Steamr being unsuitable for the selling of private data since it does not take any measures to ensure data privacy.
DML is providing a complete solution to machine learning development that not only provides access to private data but also lets developers monetize their algorithms, offer training and fine-tuning services and providing access to idle processing power.
The Development Team
3 out of 6 core team members have experience with blockchain technology, the project is also being backed and supported by Kyokan Labs; a blockchain consultancy firm.
The team behind the DML project consists of 6 people in total, all of whom have a plethora of experience with software development, business development, and machine learning technology. All of the core team members have plenty of prior experience and a majority of them have experience with other start-up projects as well. One blockchain developer (Jacky Chan) is a co-founder of Kyokan Labs.
The advisory team that is working hand in hand with the DML development team has reputable individuals who are well-versed in the field of machine learning, the advisory team also has an experienced blockchain advisor (Roderik van der Graaf).
What makes the DML ecosystem an interesting venture is the fact that the project proposes a very well-rounded solution that takes several steps to ensure the anonymity of parties that will be contributing private data to the platform. Another attractive feature of DML is that it does not just stop at creating a secure and private marketplace where data owners can monetize private data, it also caters to the problems with computational power and tries to eliminate the stagnancy that is currently plaguing the machine learning industry.
DML offers data owners the ability to generate revenue by sharing private data and processing power, it lets developers produce algorithms as freelancers and sell them through a marketplace and also plans on creating a community through which developers shall be able to innovate and improve existing algorithms through collective intelligence. The project is also planning on making its native currency quite important by giving it a number of uses.
Overall, DML is definitely innovative, its biggest plus is that it is going to venture into a market that no one else has tried to dominate so far, and the solution that they are proposing is well-thought enough to actually make people want to share their private data and processing power. Its numerous steps to encourage and monetize participation are also worth noting.
· The core team members have experience with business development and have participated in founding other projects before DML.
· 3 out of 6 core team members are experienced with blockchain technology and have worked on blockchain related projects.
· The team has similar skillsets and is being supported by 3 additional members from Kyokan Labs.
· Have not worked at any notable companies, however, they have their own start-ups and plenty of industry experience.
· The advisory team has past experience in the project industry.
Investors & Partners
· Backed by Kyokan Labs; a blockchain consultancy firm
· Makes good use of blockchain technology for creating a secure and efficient way of buying and selling private data, processing power and algorithms.
· What DML is trying to accomplish cannot be done without the help of decentralized blockchain technology; the use of nodes to connect algorithms with multiple data owner devices at a time and the operation of a trust-less marketplace both rely heavily on the features of blockchain technology.
· The DML project has its own currency in order to make transactions faster and less costly, making the DML ecosystem more efficient.
· The DML token will be used as a currency and utility, it shall also play a role in providing participants with an incentive to take an active part in the DML ecosystem.
· Is going to start off on its own private blockchain (meaning that it will be less susceptible to network overloading issues).
· Hard cap is around $14m
· There’s a lockup period of 3 years for the team
· The project’s telegram has around 10,000 members and has active discussions in which the development team takes part.
· The project has an active presence on Twitter, Reddit and GitHub.
· No mention of the team members raising capital before this project.
· The project is still in an early stage with its first prototype slated for release somewhere around September 2018.
· Only 36% tokens are available for sale
· Token price is around $0.13 according to the ETH rate on 18th March ($491)
Overall Score: 8.3/10
Token Sale Details
Token sale date — March 2018
Whitelist — Open (https://decentralizedml.com/main_whitelist_form)
Hard cap — $14m
Price per token — $0.13
Total supply — 330M
Available for sale — 36%
Maximum market cap at ICO — $38.89m