In 2005, Venky Harinarayan believed that a platform could be created to exploit the fact that humans can easily perform certain tasks that were difficult for computers. Harinarayan, who worked under Jeff Bezos during Amazon’s early days, predicted that there was a business to be built around connecting those who wanted research done with those who were willing to do it. He created the marketplace in 2005, and called it the . Mechanical Turk What Harinarayan did, in effect, was popularize and legitimize the micro-task economy by crowdsourcing raw, human-based intelligence to perform tasks that computers could not efficiently perform on their own. In the case of Amazon’s mechanical turk, the seller (turker) would be a human fulfilling the micro task that the buyer (requester) has listed. Amazon charges an outrageous 20% fee for the introduction as a third party marketplace, and an additional 5% for requesting workers that have a good track record. With more than half of the Amazon ‘turkers’ reportedly earning less than $5 an hour, the platform as the “sweatshop of online crowdsourced labor.” has been described Given the recent popularity of decentralization and disintermediation of marketplaces via the , and the inherent problems with the Mechanical Turk model, several startups are creating decentralized and open-sourced networks of turkers and requesters to complete micro tasks more efficiently and at a fraction of the cost. In the review that follows, we compare 3 of these platforms, each in the process of launching their Initial Coin Offering ( ). The three platforms are: , and . blockchain ICO Eventum Trive Effect The structure for the evaluation of the 3 ICO’s will be as follows: Value proposition (clear case for blockchain, real utility for the token) Use cases (feasibility, limitations) Quality of white paper (viable token model and demand generation roadmap, thoroughness, well-structured) MVP (stage, traction, proof of concept) Team composition (right people for the job) In order to rank the quality of each ICO, a scoring matrix will break down each parameter by the factors mentioned above. Each factor is weighed on a scale from 1–5 where 5 is the strongest and 1 is the weakest. I. Value Proposition Perhaps the most important qualifiers in measuring any ICO are a) recognizing why the blockchain infrastructure is to the proposed business model and b) understanding how its specific token value for participating stakeholders rather than just storing units of value like fiat currencies. essential creates Their decentralized solution will allow real-time information across a myriad of use cases to be collected validated across a network of users via “wisdom of the crowd” logic. This methodology takes the mechanical turk to the next level by using the crowd to develop consensus on highly variable human inputs. This allows Eventum to efficiently export data and establish critical mass agreements with high throughput, reducing the excess time and cost originating from a central source. The Eventum token (EVT), which powers the Eventum blockchain, rewards platform adopters who support those tasks in addition to data processing, platform governance, etc. In return, users are scored by the quality of their interaction with the platform and as its adoption rises, token circulation will also increase. All user and token interaction is maintained and updated on the Ethereum smart contact. Eventum: and Score = 9/10 Like Eventum, Trive too deploys the “wisdom of the crowd” phenomenon to investigate nearly all streams of media content in the war against fake news. Network reviewers will screen any piece of literature and independently rank its trustworthiness before it is stored on the blockchain. An operation of this scale requires hundreds or thousands of participants, all of whom will be instantly rewarded with a Trive Coin (TRV) for their efforts they correctly identify factual content that aligns with the results of the consensus algorithm. The independent review process takes place at the user’s convenience but consensus development takes place over a much longer duration. Trive: if Score = 8/10 Effect utilizes crowdsourcing logic in order to harvest human-based information that feeds AI algorithms. Unlike Eventum and Trive however, the means for crowdsourcing is only used for the capturing of human inputs as the core business case does not require establishing consensus. The process cuts away at the burdensome costs, storage and time required to capture this same data by a single feed. The NEP-5 token is distributed to users after the completion of a task and its value is linear to the growth rate of the community of users. Effect: Score = 7/10 II. Use Cases and Applications What makes Eventum unique is their ability to cover a infinite number of applications across this use case. Specifically, Eventum’s value arises from the collection of real-time data, and depending on the end user, can support applications like fake news identification, content moderation and image analysis. While these are Eventum’s strongest use cases, they are exploring opportunities in more dynamic instances such as emotional recognition and video analysis. Eventum: nearly Score = 8/10 Trive operates one main use case which is to identify, research and validate media content to limit the amount of misinformation that is pushed out to the public. Like Effect, Trive utilizes the Mechanical Turk technique established by the industry but challenges exist in crowdsourcing for fake news, especially when it comes to authenticating the work of opinionated or controversial pieces. Trive: Score = 7/10 Currently, Effect’s primary use case is in AI, particularly machine learning algorithm development. Their business case does not deviate from the original definition of the Mechanical Turk but like Eventum, they too are exploring other use cases such as selling algorithms-as-a-service and distributing computational power to run requested algorithms. Effect: Score = 6/10 III. Team When evaluating an ICO startup team, the five things I look for are impressive talent, extensive experience, a history of working together, a strong advisory board and proper legal counsel. : Well-rounded team of 8, all based out of the same office and led by Martin Mikeln, an electrical engineer and software developer with an extensive resume of building software and hardware projects, chatbots and raising $100k+ on a . The team successfully integrated decentralized architecture into their working prototypes and is gaining traction with users via their live demonstrations. In addition, they’ve recently onboarded full-time experts in and . Eventum Kickstarter campaign Peter Jamnik Julien Coustaury Score = 4/5 : Team of 10 led by (Columbia University MBA) and , both previous entrepreneurs and blockchain development consultants based out of Miami, Florida. They have 6 listed advisors who have very little information about them available online with the exception of economist Jeffrey Tucker, a staunch advocate for Bitcoin. Trive David Mondrus Phil Reale Score = 3/5 : Rather young and inexperienced team of 8, led by , a serial entrepreneur with over 15 years experience leading teams in both Europe and North America. The team is made of recent college graduates with little experience in blockchain, but they are well-versed in AI. It is unclear whether the company has any advisors at this time. Effect Chris Dawe Score = 2/5 IV. White paper Before jumping into this section, it is critical to define what to look for in a white paper. While there is no definitive format or guideline, the most comprehensive and successful white papers include: Problem & solution Competitive landscape Technical explanation of the token economy and use case(s) Timeline for development, phases of token release Payment processing ($ flow) ICO Strategy (price, market cap, incentives, distribution, marketing) Team behind the project Legal aspects of the tokens Token deployment and future plans The does a good job highlighting existing limitations in how real-time information is gathered while also anticipating future gaps in how that data is captured. In particular, it underscores the use cases where decentralized technology can address this pain point and positions Eventum as the market’s solution to this problem. The paper backs up its claim with technical detail of the product’s infrastructure, user interoperability and EVT’s functional value. Eventum: Eventum white paper Score = 13/15 Much like the Eventum white paper, the also has a heavy emphasis on technical and architectural detail of its proposed blockchain, particularly on the Trive Storage Module. The piece highlights the intricacies of how users interact with the platform, how they are incentivized to perform tasks that drive platform adoption and how they will be compensated and ranked for their efforts. The piece paints a picture of the emerging need for misinformation management but due to the nascency of the industry, there isn’t much market data that formalizes its magnitude. Trive: Trive white paper Score = 12/15 The outlines a strong case for a decentralized solution to AI and machine learning applications and abstractly proposes the Effect Network solution in 3 stages. However, the piece leans heavily on theory-based solutions and doesn’t quite detail system hierarchy and functionality. It’s also important to note that Effect’s white paper is still a draft and will be modified in the future. Effect: Effect white paper Score = 8/15 V. MVP Having a working prototype, or better yet, alpha or beta versions of the product demonstrates strong commitment to executing on the product roadmap. Characteristics of successful test projects include having a demo led by developers, having an interface that lets you navigate the platform, accessing fully transparent open source code and establishing a community of quality test users. has already attracted a number of users who’ve been able to verify use case proof of concept. Eventum also is the only platform of the 3 to have the proof of concept demonstrated on a live, functional blockchain! Just last week, Eventum held a that featured the platform’s ability to crowdsource and identify fake news in just under 2 minutes! For this event, 78 users split 4 Ethers (1M EVT) for their work in collecting, processing and validating data — all of which was confirmed by validation nodes and stored on the Ethereum smart contract. There are 3 more on the alpha scheduled for February that will further validate proof of concept for other use cases like content moderation and imaging analysis. Eventum: Eventum’s alpha live event live events Score = 14/15 Eventum posted the live event’s results to their alpha. The screenshot below showcases analytics detailing the tester interaction in validating “Fake News.” Eventum also posted the URLs to the smart contract that stored each tester’s transactional activity during the “Fake News” live event. More detailed access to the smart contract requires the installation of a metamask plugin. The , a javascript application that is still in the MVP stage, archives a set of media content that can be annotated and scored by reviewers following the installation of a Chrome extension. The proof of concept is on display, but the time efficiency of the process needs further articulation. Additionally, the dissemination of TRV, the decentralized infrastructure and the underlying storage of data on the blockchain has not yet been realized at this time. Trive: Trive prototype Score = 10/15 Trive posts all scored and annotated content on their MVP. The screenshot below is a sample of what input gets placed by a sample reviewer. At this time, Effect does not have a working prototype in production. We will update the post when their MVP does go live. Effect: Score = 3/15 VI. Conclusion The matrix below breaks down the overall scoring for each ICO, including how each factor was incorporated into each parameter. is the furthest along in articulation, deployment and demonstration of their technology. They have an experienced team with a legitimate alpha prototype where platform testers are rewarded for their interaction across a variety of use cases. On top of that, all of this is tracked, and can be viewed open source, on the Ethereum smart contract. Eventum Score = 87.3/100 attempts to tackle the problem of content misinformation via crowdsourced fact-checking. The MVP is still in progress, but the proof of concept is visible. However, limitations arise when this concept is rolled out in a real-time format and onto opinionated content with varying political commentary as articles come under more scrutiny depending on a reviewer’s political leaning. Trive Score = 70.9/100 makes a strong case for algorithm training and development through the collection of raw, human data. Though limiting in scope, I personally think there are great academic use cases for this tool such as helping social scientists model out how social dynamics factor into the “groupthink” phenomenon. Effect still needs work on developing a prototype, but a lot of the groundwork is established. Effect Score = 49.1/100
