Effect.AI September Update
“The limits of my language mean the limits of my world.”
— Ludwig Wittgenstein
BULLETS
- Project X revealed
- TEN Protocol
- Rosette.AI - Rosette Referral Campaign
- Chris in Georgia, UN Update!
- Q4 overview
- Effect DAC/Governance
- Effect Dashboard 2.0
- Smart Market
- Requestors, Requestors, Requestors! - Meet the new and upcoming requestors
Project X revealed
Over the past several weeks the Effect.AI team has gone almost totally out of the public eye to focus all efforts on what has been dubbed Project X. Today they are more than excited to unveil exactly what Project X entails and bring you along on the journey.
Project X was born from members of the Effect.AI community. These community members wrote a very detailed and comprehensive business plan for their ideas. The core team of Effect.AI have always stated that they would support well thought out community initiative. The quick overview of the plan was to create the very first application on top of the technology that the Effect team has built. A true example of TEN Protocol as described in the Effect.AI whitepaper.
With an eagerness to show the world a working application on top of the TEN Protocol the core team quickly joined forces and jumped right into development. The main issue was that the TEN Protocol was set to be developed after the full release of the Smart Market. After a short deliberation, the core team was too excited by the proposal to pass up the opportunity to work with the community to make this a reality. So Project X, in a nutshell, is the first development phase of the TEN Protocol and a commercially global application on top of that Protocol called Rosette. Let’s get to the details!…
TEN Protocol
“The Effect Network (TEN) is an open protocol that enables anyone to build an automation pipeline faster than ever before. The two most commonly used ways to automate processes is human power and algorithms. TEN seamlessly allows anyone to build out a workflow that can include specifically qualified human agents at scale and algorithms to automate almost any digital process. Both the human agents and algorithms are offered as easily interchangeable modules in a marketplace. This gives you the power to access an extensive library of algorithms and various skilled human agents. Not just for data scientist or computer programmer. With an intuitive drag and drop UX the protocol allows anyone to build automation into their process. As automation continues to be the main priority for human evolution, we see TEN as being an essential part of the world of tomorrow.”
-Chris Dawe
Project X is a combination of the development of the TEN Protocol and the first-ever “3rd party” application built exclusively upon TEN called Rosette.AI. Yes, you heard correct… The TEN Protocol is here! If you have been following the project you would probably know that the Effect.AI team proposed releasing the 3 phases of the Effect Network and then tying all 3 platforms together in an open protocol called TEN or The Effect Network. The issue the team had with this roadmap was that it greatly limits the usage and adoption of the first 2 phases of The Effect Network, Effect Force and Effect Smart Market. The TEN Protocol allows 3rd party service providers to easily build on TEN and connect their products. They simply need a front end and the TEN takes care of the rest. Jesse Eisses, Co-founder and CTO has just released a blog detailing various aspects of TEN and where it may fit in to the Automated future. Check out his blog here.
Without further ado, the very first application to be built on TEN, in association with members of T.E.A.M., developed by Effect.AI, introducing…
Rosette.AI
We are at the forefront of the AI revolution. AI is all around us and it’s constantly getting smarter. Although several cutting edge algorithms have been deployed and are helping humans in fascinating ways we have a long way to go. Take translation of text for instance. We all know Google translate by now and some of us rely on it on a daily basis to communicate in our day to day lives. But if you use Google translate, you have probably noticed it does not do such a great job. For example, the below messages are translated from Dutch to English and vice versa by one of the most advanced algorithms on the market:
Now imagine you have a company that you have worked night and day to get off the ground. You have spent countless hours on branding and telling your story with the content on your website. Your website is the first line of communicating your business and its proposition. You enter a new market in which your customers speak another language and have a different culture. Would you use Google Translate to get your story across to this new frontier? Yeah… I don’t think so… Enter Rosette.AI
What is Rosette?
Rosette is one of the most sophisticated translation pipelines on the market. Using a combination of Human and Machine Intelligence to translate text faster and more accurately than ever before.
By utilizing Rosette, businesses can tap into new global markets to broaden their customer base. A company can realize a competitive edge when they can connect to their audience in languages other than their native language. The language of the customer is what matters. With Rosette.ai it’s easier for a business to open new markets, explore, and develop them faster with higher efficiency and lower costs. Rosette, powered by Effect Force workers, is able to offer translation services of the highest quality, fast, and at a great price in comparison to competitors.
When a customer pays for the Rosette service they are able to pay in a variety of traditional and alternative ways. For an additional 10% off, customers pay in EFX. Whether a customer pays with Paypal, VISA or Bitcoin the funds will be converted to EFX on exchanges to pay the translation workForce on Effect Force. Check out the designs so far for the payment process below. Notice the Referral code section… More on that further in the newsletter.
A recent event further demonstrates the use case for Rosette.AI. Mid-August, Sebastiaan van der Lans from WordProof put out a call to the community asking for assistance with translating the WordProof Timestamp. With more and more timestamps being created every hour, and websites all over the world securing their content, WordProof recognized the need to make the timestamp plugin more inclusive. Members of the Effect.AI community stepped up and within 48 hours the timestamp was translated into 15 new languages! Sebastiaan gave Effect.AI and Effect Force great praise during a webinar he was a featured guest on. You can watch it here.
Visit https://rosette.ai and leave your email address to keep up to date with the launch and for your chance to win $100 in crypto.
Rosette Referral Program
Rosette will be a global community effort and we invite you all to take part in its success!
The Rosette referral program is focused on bringing in customers that are in need of translations of their WordPress website or WordPress Plug-in. Rosette will soon offer most types of document translation but to Kick-Off, the focus is on a precise but massive part of the market. WordPress websites account for 34% of all websites on the web today. That’s 75,000,000 websites and in 2018 there were 204,000,000,000 words published on WordPress sites… Rosette charges by the word, so you do the math!
When the referred customer pays for a text-to-text translation service of at least 500 words, you (as the referrer) will receive a referral reward of 10% of the purchase order after referral discount. The customer will be greatly incentivized to include your referral code at the check out as it will also give them a 10% discount! A Win-Win.
Along with the launch of Rosette and the referral campaign, there will also be a potentially HUGE bonus to those who are participating in the referral campaign. 5% of the total revenue generated by Rosette in the first 90 days of operation will go into a prize pool. Those funds will be distributed to the top 10 referring participants. A digital leader board will appear 30 days after the contest has started. Good luck and stay focused:)
A very special thanks to community members Rochelle and Wisse for the detailed business plan and research of Rosette.AI and helping make it a reality.
If you or someone you know has a business model of a service that could use Effect Force and the TEN Protocol please contact TENfund@effect.ai to explore funding and technical support of the project.
Chris in Georgia, UN Update
The biggest step for the Effect Social Impact HUBs is here!
The United Nations in collaboration with the city of Rustavi in the country of Georgia has substantially funded Effect.AI’s efforts to build out and open the very first Effect Force worker center.
Creating safe work environments for Effect Force workers to learn, grow and develop skills within a physical location has been one of the most important parts of the Effect.AI vision since the start of the project. Understanding that such a venture would pose difficult to do alone, the team reached out to the UN in Georgia back in October of last year to explore the idea of collaborating on this initiative. The team proposed a pilot in collaboration with one of the poorest cities in the county. The pilot took place over several months to show that Effect Force could train and teach citizens to work on the platform and earn a living. The pilot headed by Chris Dawe Effect.AI CEO and Sam Spall Head of Strategic Partnerships was described as one of the most structured and successful pilots by the head of UN Georgia.
The process of utilizing the partnership and acquiring funding can take a long time when working with one of the most impactful organizations worldwide, nevertheless, the team is ecstatic to be moving forward to show how powerful the impact of Effect Force, blockchain and AI can be on the lives of Georgian citizens.
The mayor of Rustavi has granted Effect.AI a location in the center of town within the most historic building in the city. Several decades ago it was the location where the town would gather for markets, theater and events. With the economic collapse and political unrest over the years, Rustavi, like all of Georgia has seen better days. With new positive outlooks for the future of the country, Effect.AI is eager to help turn this once-thriving city center back into a place of hope, prosperity and community involvement.
Chris will be in Rustavi from October 13th-18th to coordinate the build-out of the first of many Effect Impact HUBs. Watch out for daily video blogs from the trip on Effect.AI’s social media pages.
Q4 overview
DAC
As the TEN vision has become ever so clear to the core team, it is import to create a decentralized network behind the initiatives of TEN. V.1 of the Effect DAC should be operational with a few key features to kick-off. A worker proposal system with various topics the core team and community would like to see in the evolution of the platform, the EFX and NFX tokens and governance of the Network will go live. A key feature of the worker proposal system will be the ability to vote on these community proposals and follow your favorite contributors. The initial focus of using the DAC’s proposal system will be to onboard AI and Blockchain academics to join our research and development goals.
Along with the worker proposal system will come the structure and members of the governing board of the DAC. The structure is not solidified as of yet but we do know that the board will comprise of several Effect.AI community members and members of the core team to guide the path of The Effect Network.
The team will work closely with third parties on the creation, development and roadmap of the Effect DAC and we’ll keep the community updated as this exciting step moves into reality.
Effect Dashboard 2.0 release
Dashboard 2.0 will be upgraded with a few new features to give much more functionality to users. Currently, the Effect Dashboard allows users to send and receive EFX, stake EFX, claim NFX rewards for staking, and a Stake AGE tracker. 2.0 will see the worker proposal system portal, voting, governance features and possibly a couple of features we were saving for 3.0. Check out the vision of the future of Effect Dashboard.
Smart Market Beta
The Smart Market has been the most interesting and exciting part of the last few months. As algorithms will continue to be key components in the automation pipeline of productivity we have been building several components and functionalities of the back-end of the Smart Market. This includes the deployment of a few generic and self-made algorithms within the Effect Force platform. Using a combination of humans and algorithms in the same automated productivity pipeline will become the norm and Effect.AI and the TEN initiative is ahead of the game in this new world. Utilizing campaigns from Effect Force, we have deployed validation algorithms and non-proprietary algorithms to aid in the quality of the human annotations performed on the platform. Though simple and only slightly helpful it showcases how companies can tap into and benefit from both Effect Force and the Smart Market in the same pipeline. Another fantastic example of AI and Humans working together is the Rosette translation pipeline. Find the infographic in the Rosette.AI section above in this newsletter.
We’re extremely eager to have a live and functioning Smart Market within the TEN protocol by the end of Q4. Various updates will come along the way.
Requestors, Requestors, Requestors!
We will continue to focus our efforts on getting new and exciting requestors onto the Force. The attention and interest from some of Europe’s largest companies continue to pour in. As some of you know, we are deep in talks with a few of these parties and look forward to sharing when we get these deals done.
Meet the past, present and future requestors
Over the last year we’ve had a variety of requesters utilise the services of Effect Force. From Lobster Media to Obi4wan to Ortec. At the beginning, we focused on covering all annotation types, much like an MTurk. However, we’ve noticed during this project that focus is key and our strengths lie with our worldwide workforce which has the ability to speak a tonne of different languages. Let us take you on a little journey of the requesters that have been and are to come!
The Past:
Computer Vision — the starting blocks
Lobster Media
Annotation task — identify whether an image was considered a selfie or not
Why — to improve the accuracy of detecting selfies in a photo database
Annotation type — image tagging
Workforce — 400,000 images tagged in 2 days
Results — algorithm performing at 95% accuracy
Altoros
Annotation task — define the level of damage to car parts
Why — to automate the damage level detection for insurance claims
Annotation type — bounding box
Workforce — 15,000 images processed
Results — algorithm performing at 90% accuracy
Sentiment Analysis — human accuracy needed!
Obi4wan
Annotation task — detecting positive, negative and neutral sentiment from online content
Why — to gain access to a wider workforce that can deliver quicker results
Annotation type — sentiment labeling
Workforce — 10,000 articles processed
Results — Effect Force workforce outperforms Obi4wan’s current solution to manual sentiment labeling
The Present:
Natural Language Processing — we have 30+ languages with the workforce
Ortec
Annotation task — detecting impact data within each sentence in a Disaster Relief report
Why — to improve the process of fund distribution for disaster relief cases
Annotation type — sentence classification
Workforce — 11,000 sentences processed from 1,000 DREF reports
Results — Algorithm now at 80%, reducing the time spent on the funding review process by 50%
Lincoln Labs
Annotation task — identifying what stimuli categories can be seen in an image e.g. water, damage, environment, vehicle, infrastructure and from there identify the stimuli within those categories
Why — to automate the process of distributing resources required for disaster relief operations as well as speed up the time to the disaster
Annotation type — image labeling
Workforce — 14,500 images labeled generating over 100,000 tasks
Results — desired results are to create a robust image training data set that trains an algorithm which speeds up and automates the disaster relief process for emergency services
The Future:
coming up — continuing with NLP
Nina
Annotation task — transcribing flemish audio
Why — to train a customer service voice bot in flemish
Annotation type — voice transcription
Workforce — 136 hours of audio, equating to 1000+ hours of work on Effect Force
Results — desired results are to have the voice bot understanding 90% of conversations
Adocia
Annotation task — extracting critical information from invoices
Why — to streamline the invoice distribution and payment process
Annotation type — invoice transcription / document extraction
Workforce — 3,000 monthly invoices to be transcribed on Effect Force moving up to 10,000
Results — desired results are to have all invoices automatically pushed to Effect Force to be transcribed by a worker (determined by language) and then pushed back to Adocia’s invoicing system in less than 12 hours.
New Effect Force Website
As the Effect.AI team has identified their go to market strategy for Natural Language Processing (NLP) the need for a fresh new website was a must! The new Effect Force website will clearly communicate the value proposition of using a Human Intelligence network in automated flows. Watch out for the new site in October.
Thanks for reading!
Web | Telegram | Twitter | Facebook | Youtube | Github | Reddit | Linkedin| Medium | Steemit
