Effect.AI Rating Review
|Start ICO||24 Mar 2018|
|End ICO||28 Mar 2018|
We have reviewed the project's progress as of February 2019 and would like to note that official telegram group contains several messages on questions about the project progress, however the team answers vaguely and unclear. Public beta was supposed to be launched in mid-February, however, it did not happen. Initially, the platform was going to be launched in November 2018. Therefore, the project's rating has been decreased to Risky-.
Effect.ai tokens, in our opinion, will be interesting as a long-term investment. The potential of the project’s intended market is quite large; however, the project team lacks a proven track record with other successful projects, and has not announced any partners or advisors.
Based on the project’s roadmap, profitability of Effect.ai tokens will not be high at the initial stage, since the project is planning to focus on the development of the platform and the framework (Q1 2019 and Q4 2019 respectively). However, after completion of development, we think the team will move to increase profitability for their tokens.
General Information about the Project and ICO
Ticker symbol: EFX
Project type: Artificial Intelligence
Extended company description:
Effect.ai is an open, decentralized network that provides services for the AI market. The platform will operate on smart contracts developed on the NEO blockchain. The goal will be to provide all necessary services for working with AI: searching for developers to write AI applications, buying, selling and exchanging AI applications.
Use of the Effect.ai platform will not require payment of a commission, thus offering a low barrier to entry.
The platform’s improvement strategy comprises three separate phases: Mechanical Turk, Smart Market and Effect network.
Company: Effect.ai BV
Details of token offering
Public token offering start: March 2018 (public crowdsale start date is in the process of determination)
Public token offering end date: N/A
Token sale duration (weeks): 18 days
Token emission: 650,000,000 EFX
Tokens available (Pre-Sale + Public token offering): 260,000,000 EFX
1 EFX price: 0.07 USD
Accepted payment: NEO, GAS
Token limitation: 650,000,000
Hard cap: €14,820,000
Soft cap: €4,280,000
The open token sale will take place in 2 stages:
- Whitelist (a date for registration has not yet been announced).
- Token offering: Sale start date March 2018, end of token offering 18 days after start date.
Bonus: 10% for the first 5,200,000 tokens (2% of token sale)
Description of the Services and Scope of the Project
There will be two principal categories of user:
- Customers: submit requests for analysis of information or order an AI application.
- Users: fulfill requests for customers, who in turn remunerate them.
The Effect.ai platform is built on the basis of the NEO blockchain and AI technologies which enable users to create applications based on them and process information.
The platform includes 3 phases of development:
1) The Effect Mechanical Turk Platform.
2) The Effect AI Smart Market.
3) Effect Power.
Let us consider these phases more closely:
1) Effect Mechanical Turk is a decentralized P2P marketplace, where customers can find developers to implement ideas for their application or data retrieval. The functionality is similar to Amazon Mechanical Turk, Fiverr, Crowdsource and Guru.com. A hash-based distributed file vault will be used for data storage where each media resource can refer to one hash. Applicants will also be able to provide data sets via traditional channels such as Amazon S3, Google Cloud Storage and FTP.
In addition, developers will have a rating that depends on the quality of tasks performed, while only the task owner will be able to view the rating. Users are also awarded Honor Tokens (HNR) for honest tasks. They enable receiving a higher reward in the future for work, or paying less tax in the Galaxy Pool (see below).
2) The Effect AI Smart Market is a decentralized market which enables AI algorithms to exchange services. The network will maintain a registry of available applications. This registry will be enriched with a semantic ontology that describes applications, as well as a technical diagram of its inputs and outputs.
Application endpoints are transmitted over HTTP. Data is exchanged in JSON format and must strictly confirm a specific RDF.
3) Effect Power is a decentralized distributed computing platform that will run popular systems for in-depth training. An efficient decentralized computing engine is based on popular training networks such as Caffe, MXNet and Tensor. In the future, it is planned to do this through partnerships with such projects as Golem with their supercomputer.
The platform also includes elements such as the Galaxy Pool.
The Galaxy Pool consists of EFX tokens, NEO, GAS, BTC, LTC, etc. This will ensure the liquidity of EFX tokens as well as fast currency exchange. A mechanism of "labeling" of coins was invented to protect the pool from traders and speculation. Any coin bought in the Galaxy Pool will be initially labeled and cannot be placed back in the pool. The labeled coin is erased (converted into a regular EFX token) when it is used to pay for services.
In addition to the main EFX token, the Honor Tokens (HNR) token is issued. Users receive these tokens for their efforts and high-quality task performance. These tokens cannot be sold or bought; they are also disposed of over time. These tokens enable users to take more highly paid assignments or reduce payments for Galaxy Pool services.
Users can also make changes to:
- smart contracts: rate in Galaxy Pool and transactions with taxes on services.
- service contracts (including their creation), such as new types of task and changes to existing contracts required in the future.
A proposition is accepted only if the majority of users from the council created, voted for it. Voting mechanisms and the formation of the council are under development, according to the team.
Phase 1. The platform will function as follows:
1) A customer registers his HIT on the platform (HIT - Human Intelligence Task).
2) A smart contract is created.
3) Data is provided.
4) The HIT is executed and confirmed by the contract system.
5) The customer accepts work and pays for the service provided with EFX token.
1) An application owner is registered on the exchange, specifying the degree of development of his application and payment for its use.
2) After this the application can be used with smart contracts. The contract will have to transfer the necessary funds to the contract owner in order to obtain an authorization token that enables it to interact with the application.
3) Algorithms can interact with each other through the application registry.
At the last stage, algorithms will be executed globally, i.e. processed on different computer nodes within the network.
Based on the white paper, we see that the project solves several problems at once and provides the following options:
- Access to AI technologies and a large amount of data.
- The exchange of AI applications, making it possible to improve existing applications.
- Makes the purchase and sale of AI applications affordable.
According to the RoadMap, a beta version of the product will appear in Q2 2018. Since it is supposed to use AI on demand through a smart contract, we think that the project should concentrate on its training options and increasing the number of tools for its use. An obvious advantage is making changes in the platform’s smart and service contracts. However, we consider the issue of Honor Tokens (HNR) superfluous; EFX tokens could be used for the same purpose.
Description of the market niche for the project:
Effect.ai is oriented to the AI algorithm market, so it is advisable to analyze this niche.
General description of the market sector, segment and project niche:
According to Statista, revenue from the use of AI was $2.4 billion in 2017; growth to $59.7 billion is expected by 2025, so CAGR will be 57.2%.
This growth rate is confirmed by the analytical agency Technavio. According to their forecast, the global market for artificial intelligence will grow with an impressive CAGR of above 50% over the next 3 years.
PWC and McKinsey also state in their report that AI will have a significant impact on the market and will become an integral part of research. According to Gartner, 85% of interactions with customers in retail trade will be managed by artificial intelligence by 2020. However, artificial intelligence will be popular not only for retail trade. AI is also used by financial companies such as banks and investment firms; many companies use AI to increase investment efficiency and fraud detection. In fact, 98% of corporate leaders consider AI a necessary element in their business [AI Transforming Business: Corporate CxO Perspectives].
Current dynamics of the main market segment, a forecast of future dynamics, domination of major players, as well as general trends in adjacent segments
Based on the analysis presented above, we can say with certainty that the AI application market will develop dynamically, and this type of service will remain strongly in demand.
However, the team considers a slightly different angle in its review, concerning growth in world GDP due to the introduction of AI.
Analysis of existing problems in the industry, reflecting the current needs of this market segment, its issues and problems.
The white paper highlights the following problems for this market:
1) Data processing: Tasks performed by AI on data processing require an in-depth approach and consequent cost in money and time.
2) Divergent tasks: an obstacle for the development of a complex algorithm is the need to interact with aspects of the world outside the current domain, which increases the work required.
3) Computing costs: the development of AI algorithms is a task that requires computing resources. This requires a technical infrastructure capable of handling terabytes of data.
The chosen market segment is clearly promising. High growth rates in the AI applications market indicate an opportunity to occupy a market niche. However, both the RoadMap and White Paper lack information about the marketing campaign or scaling strategy.
A lack of agreements on the use of the platform also makes us wonder whether the company will be able to gain a foothold in the market.
The idea of creating a platform for writing and exchanging AI applications will be of interest since the market has shown steady growth in recent years in this direction (paragraph 4 of this review). However, as will be shown in paragraph 5, companies with a strong reputation and steady customer flow are already operating. This niche can be thought of as oligopolistic; in this regard, the company can only count on a small share in the market.
Competitive Analysis and Competitive Features of the Project
To solve the above problems, Effect.ai proposes the implementation of a protocol that decentralizes the global artificial intelligence market, which according to the team reduces barriers to entry, stimulates market growth and significantly reduces the cost of using such technology.
According to McKinsey, innovations in the field of digitization, analytics, artificial intelligence and automation create opportunities for increasing productivity for business and the economy, even when they will affect employment patterns and the future of work in general.
Below we present some industries where the application of these technologies is possible in the near future [McKinsey]:
The concept of using AI is now quite popular in the target market. Here are some key players in this area:
- Amazon Mechanical Turk and Amazon AI - this platform enables users to find AI developers. The MTurk web service enables companies to provide software access to this market and find a variety of workforce participants on demand.
- Fiverr - a global online marketplace that offers tasks and services, including writing AI applications, starting at $5 for work completed.
- Onespace – an "all-in-one" platform for the automation of workflow with an embedded network of freelance talent.
Effect.ai does provide services at a lower price in comparison to the competition, and in a wider range of countries, but the companies above are already well-established.
For example, Amazon Mechanical Turk involves about 6.5 million people on average; Fiverr has a 40 million-strong audience, so it will be extremely difficult to compete with these giants. [https://be1.ru/stat/fiverr.com], [https://be1.ru/stat/mturk.com]
As well as Amazon, other large corporations such as Microsoft, IBM and Google.are actively interested in this area. [https://www.microsoft.com/en-us/ai/], [https://www.ibm.com/watson/], [https://ai.google]
In the sphere of blockchain-based AI projects, we have compared Effect.ai with existing companies and identified the following direct competitors (including companies entering the market):
- Graph Grail - a platform that provides the opportunity to buy and sell AI applications, conclude smart contracts, as well as create and train neural networks.
- Startcrowd - a platform that enables users to find developers for their AI applications or to offer their services in this market.
- SingularityNET - a protocol developed for the exchange of data and coordination between AI processes, buying, selling and exchanging AI applications as well as for automating processes.
Let us consider each of these in more detail:
- Graph Grail is quite a successful project, currently at the stage of its token offering. In the pre-sale stage a hard cap of $200,000 was reached within 15 hours. The project specializes not only in AI but also in writing and selling smart contracts.
- SingularityNET has already completed its token offering, collecting $36,000,000 and thus reaching hard cap. Popularity of the project is growing rapidly: over the last month traffic on the website has grown by 300,000 visitors. In general, SingularityNET can be defined as the main competitor, as functionality of Effect.ai and SingularityNET are very similar. The only difference with SingularityNET is the presence of a humanoid robot for processing information.
- Startcrowd is slightly different from these projects as it positions itself as a social network which also offers help in finding AI developers. However, Effect.ai offers a wider range of services for its users in the form of a marketplace, and the ability to process information. [https://be1.ru/stat/singularitynet.io]
In general, Effect.ai differs from its competitors in its low service costs. The project is mainly focused on creating its own development tools and their commercialization.
We consider this a positive moment, as the project has a focused business model for its commercialization, in contrast to some competitors who plan to immediately operate using several models.
In general, the market is already occupied by large companies. However, the rapid development of AI technology enables us to say that the project is potentially able to occupy a market share.
Team and Stakeholders
The team consists of 8 people. Information about the main team members based on LinkedIn is given below:
- Chris Dawe (LinkedIn) - CEO. No data on previous experience.
- Jesse Eisses (LinkedIn) - Leader in the field of blockchain. Previously intern at 3DUniversum, developer at Itsavirus, MyAdbooker.
- Laurens Verspeek (LinkedIn) - Head of development. Developer at Itsavirus.
The team has 5 more people.
The whole team has indicated their affiliation with the project. According to the website, the team also has a design specialist, developers, community manager and a management specialist.
However, according to Linkedin, none of the team members have previously taken part in a token offering. The team also lacks AI specialists.
Most of the team also works for Itsavirus (this company is more specialized in e-commerce).
Having analyzed open sources, we did not find any involvement for any of the team in any suspicious activities in the past.
The main governing body responsible for the development of the platform is the Effect.AI Council (EC). EC is a group of people who have the right to vote on propositions for improving the platform. This council can be left at any time, and new members can join if they are approved by the majority. Board members are responsible for the continuous application of changes in the network to adapt it to current market conditions.
The parent company: According to the official documentation, Effect.ai does not have a parent company conducting its token offering. The team have been assembled for the implementation of a specific project.
The project also lacks information on early stage investors, key partners and advisors (experts). The team has not yet disclosed this information. Perhaps it will be available closer to the start of the public token offering.
The project team is from the Netherlands and has a young average age; many of its employees do not have much experience. This is the first joint project for this team, as far as we can judge from LinkedIn information.
We consider it necessary to announce key partners and advisors before the start of the public token offering; this will be an added advantage which could increase attention from the blockchain community and attract potential investors.
General characteristics of the EFX token:
- Token type: EFX is a utility token
- EFX is a platform currency.
The token is used internally for the following purposes:
- Payment for developers’ services
- Commission payment in the Galaxy Pool (The Effect Network will maintain a central pool of tokens to provide liquidity, encourage adoption and stabilize network fees). [https://effect.ai/download/effect_lightpaper.pdf]
- Buying and selling AI applications on the marketplace.
Token type: NEP-5
EFX is a utility token, not subject to SEC regulations.
The sale is implemented through a White List and KYC procedure. Residents of the United States cannot take part in the token offering due to legislative restrictions.
The total volume of tokens supplied will not exceed 650,000,000 EFX. The maximum number of tokens that will be available for sale will be 40% of total supply, i.e. 260,000,000 tokens. 20% is reserved for future sales in September 2019, 16% is for the Galaxy Pool and 2% for the bounty program. Effect.ai reserves 15% for the team and 7% for advisors and partners.
The team is not planning any additional token emission.
Analysis of Factors Affecting Future Token Value
Long-term, medium-term and short-term factors affecting the project’s value are given below.
1. Scalability of implemented technologies in the long term:
- We assume that AI technology will be strengthened in the market in the long term. Thus the project can be considered promising in terms of scalability of the technology.
2. Evaluation of the project’s development plans and the interest of developers and companies in the project in the medium term:
- We assume that success for the project will directly depend on the number of clients and partners on the platform. In this regard, it is necessary to conclude more partnership agreements, as well as conduct an extensive marketing campaign for development.
3. Evaluation of the real utility of the token in the coming months, and the popularity of the project:
- The token will not have a use in the immediate future since the project does not have a publicly usable platform.
- The platform’s release is scheduled for Q2 2018.
Conclusion: According to forecasts, the AI market will develop dynamically over the next 5 years; this will entail not only the launch of a large number of new platforms but also capacity strengthening for the market’s biggest players. In this regard, Effect.ai must be prepared for competition. The company should focus on its marketing campaign, recruiting partners and developing the product itself, as some competitors are already firmly entrenched in the market and offer more complex services. (p.5)
4. Evaluation of the Roadmap
Comment: The roadmap is presented with a plan for platform development; the platform will be built on the NEO blockchain.
Analysis of Investment Risks
In this section we consider the project’s weak points which could potentially have a negative impact on its implementation. We highlight the following risks for Effect.ai:
- If the project fails to recruit a sufficient number of partners (solvent users) this would prohibit its further development.
- The project has a rather long development period planned for the platform and the framework, which may lead to a decrease in user attention.
- There may be difficulties with launching the platform (no public tests have yet been conducted).
- Since the project is counting on the development of tools for the platform by third parties, we see a risk of lack of interest from IT professionals.
Risks for the investor:
- Only 40% of the total number of tokens will be available for traders, which means that the majority (15% + 7% + 20% + 16% = 58%) of tokens will remain on the platform (16% - Galaxy pool and 20% - reserved for the future token offering scheduled for September 2019) and will be in the hands of the company (15%) and unannounced advisors and partners (7%). According to industry analysts, this has already happened with Ripple (XRP); although the presence of the blockchain implies decentralization, the project will have a completely centralized approach.
- Lack of agreements on cooperation, advisors and insufficient popularity for the project may lead to returns on investment being below target at the initial stage.
- Implementation of the project (this risk arises from an internal contradiction between the team and the objectives of the project. However, these factors are not yet clearly evident).
The developers may be contacted using the links below: