Humans have the ability to learn and understand things then be able to deal with new and difficult situations. This ability is also known as Intelligence or more precisely Human Intelligence. With this intelligence humans crafted equipments(machines) to make tasks that had been earlier understood and completed, easier and less time-consuming. These equipments could only perform tasks with the direct help of humans and could not perform predictive analysis or take decisions on its own, in essence, the machines could not think.

‎The realization of the enormous benefits and possibilities that would open up if machines could think propelled humans to start thinking of machines making decisions independently under supervised or unsupervised learning.

Machines Think!

Ever watched a science-fiction movie where robots could communicate with humans, complete chores unmonitored or even dance to music beats? If yes, permit me to say that those movies still remain science, but no longer fictional.

‎ Artificial Intelligence (AI) is simply the theory and science of machines learning and understanding data from events and then being able to also deal with new and difficult situations. The process through which they learn and make judgements is not so different from how humans do so too. Humans learn by being able to receive, store and analyze sensory inputs(data) to make an optimal decision.

‎Getting to these optimal decisions is not a smooth sail most of the times because our data storage(memory) may fail us or we may even handle some variables wrongly, making humans prone to errors. However for machines, through some simple set of written AI algorithms such as SVM, Naive Bayes, Decision Trees, e.t.c. allow stored sensory inputs(data) to be analyzed for optimal decisions with less proneness to errors at a lesser time.

()

Machine Learning as the name implies, basically involves training a machine to be able to make decisions for itself. A machine is trained with a particular dataset call the training data and after then it is tested with another dataset called the testing data which it had previously not been exposed to. The accuracy of predictions and judgments have been impressive, surpassing human intelligence with minimal errors. Trained machines or like they are popularly called Artificial Intelligence have been applied in a wide range of fields including businesses, customer services, automobile, Health and medicine, music, robotics e.t.c. Some major uses of Artificial Intelligence include image recognition, object identification, classification as well as detection.

This may all sound easy to do, but if you think of how difficult it is to train humans, you may change your mind about the difficulty level😅 because it is indeed difficult to train machines. Intelligent machines' intelligence depends enormously on the training dataset. The bigger the dataset it is trained with, the more intelligent it will be. This automatically raises the problem of data collection, pure data with little or no outliers.

‎Collecting data is not the only problem associated with AI. Being a fairly new technology pulling a huge chunk of industries, it is also pulling a larger chunk of criticism from some particular regions who fear that machines will take over the human race in the nearest future. Machine trainers in these regions are often discouraged as anyone wouldn't patronise their products. Hence a marketplace would be suitable to promote AI Development and further healthy competitions to release better products which would be constantly on-demand by businesses or requesters.
‎Another commonly overlooked problem is the computational power needed to store and analyze a huge amount of data. Getting super computers is usually the last thing on new and small AI services providers.

These limitations have grossly affected the growth of AI in some regions especially in African countries specifically my country Nigeria, where there are majorly traditional-machine services consumers who do not really see the need for AI and only a few developers producing few products but no marketplace for it. Also due to the small sizes of the present businesses with little capital who render AI services it would prove costly to pay for datasets (which to a large extent could be forged judging from the positions of African countries in the corruption perception index which shows low transparency) and good computational power.

To understand how these AI industry per se can be enlarged and made active, one needs to know its components are and bring them closer by creating secure links or connections(a network) between them. The components of this industry include;
• AI developers
• ‎Datasets
• ‎AI services market
• Computational power

Connecting components together produces a system and like in mechanical systems, the closer the components are, the lesser the entropic and frictional losses which would consequently reduce the efficiency of the particular system. It clearly proves that, the AI industry has been inefficient hitherto and that although the industry currently worths about $2.42 billion, it can be improved to feel more of its effect in our daily living.

That long awaited improvement is finally here.😎

EFFECT.AI - Feel The AI Effect!

The internet has undoubtedly has been the greatest technology which has helped to increase proximity between places, point of goods production and consumption, people, businesses and customers and it's fundamental definition still remains "a network of computers" and it has led to a seamless and efficient distribution of natural and human resources, consequently improving humans' standard of living.

‎How about about applying these network advantages to artificial resources as well?🤔💭

This is what the Effect network aims to do.

The Effect network, Effect.ai, is a decentralized Blockchain based ecosystem where all the earlier stated components of the AI industry meet. It restructures the already existing AI economy by pulling closer the pillars of this economy to aid cheaper, faster and easier interactions between the economic components.

The development of this network proceeds in three(3) phases through which it innovatively solves the problems of datasets, AI start-ups determent, prerequisite computational power, no market for AI products. These three phases does not only solve these AI-related problems but it also helps improve the current state of the world economy.

‎How about about applying these network advantages to artificial resources as well?🤔💭

This is what the Effect network aims to do.

Phase 1 || Effect Mechanical Turk

A mechanical turk(M-Turk) is an online crowdfunding platform with an on-demand global workforce for completing human intelligence tasks(HITs) to get monetary rewards in return. For instance, say you've got a ton of pictures and you need to tag a particular object in all of them, as a requester, you load the tasks into an M-Turk and workers also called Turkers conpletes these tasks for the pre-specified rewards. The problems associated with these kind of platforms until now were inequivalent compensation for completed tasks, reward fraud by the requesters e.t.c

‎The first phase of the development of the Effect network is bringing M-Turk to the blockchain. The loaded tasks are targeted by developers to help get training datasets from a workforce not just limited to the developers' environment without any intermediary costs. Multiple HITs submissions helps spot more accurate datasets, hence solving the problem of forged datasets. Also, as regards the problem of reward fraud and little compensation, it is practically impossible because the blockchain technology provides transparency and with the use of smart contracts, the tasks and rewards ascribe to it are loaded into a smart contract visible to both requester and worker and at the submission and approval of the completed task, the rewards are released to the turker.

‎The implications of this phase goes beyond developing the network and increasing user growth. It goes as far as reducing the social vices of the world's economy. For most mechanical turks whose least workforce age is 18, it provides employment for a large number of people who either complete tasks for pleasure or for livelihood earning an average of a dollar per hour($1/hr). Some of these tasks include podcasts transcribing, image classification, data labelling e.t.c.
‎Effect M-Turk helps us to see how humans and AI collaborates to ensure survival of both races, as humans help machines with some tasks they are unable to do, Human Intelligence Tasks(HITs) and machines help humans to accomplish tasks beyond our abilities as well.

Phase 2 || Effect Smart Marketplace

With the rise of users and AI developers from the first phase, it is only natural that the second phase of the network development would be to include a market where AI developers can sell or rent their products to businesses globally. The exponential growth of AI activities in the network through phase 1 data exchange interactions will attract businesses to seek out AI developers for their services. The marketplace is designed such that every AI algorithm has its own wallet to make payments easy and controllable by the developer. It also allows for collaboration, communication and exchange of services between different algorithms.

‎A market for a particular product either leads to rise in supply of that product or a rise in its price. Effect smart market which lets AI developers meet with individuals or businesses who have recognized the benefits of AI functionality globally raising the developers above their environments or critics. This increases the reputation or worth of AI developers and soon enough, we would have children having dreams of becoming an AI developer rather than becoming a doctor or some other profession and who knows how long it will take before "AI development" will be included in high school curriculums. 🤔😁

Phase 3 || Effect Power

This is the sustainability phase and it involves providing a decentralized and distributed computational power all running on some deep-learning frameworks like Caffe, MXNet and Tensor flow to support AI developers rendering services to a large amount of customers in order to avoid weakness of the network like downtimes which maybe be due to hacking of computational power failure.

‎This is essential to small businesses built around rendering AI services who may not at present have enough capital to acquire desired computational power. Effect power, made available to these businesses at a low cost encourages infant AI start-ups to grow until they become independent with more than enough required computational power.

‎The Effect power support can be envisaged to lead to a grossly reduced mortality rate and an increased natality rate of AI start-ups, hopefully until a balanced is reached to prevent overpopulation of the network.😁

Not so Artificial Dreams!

As a computer engineering student, Jordan has always been fascinated about the truth that machines could be trained to think for themselves. Jordan specifically had dreams to build a machine which could recognize faces.

‎Although still a student who was not being taught what he felt was needed to build his dreams, he was eager to learn how to build this face recognition machine. So he did some research online with his low spec laptop, discovered some machine learning courses and enrolled for these courses. Balancing taking these courses with school work was not so easy especially because he was drawn towards AI development more than he was school.

Months past and Jordan was becoming an expert AI dev. He already started his face recognition AI algorithm and just before he finished it, he realized that he needed a lot of faces, which means a lot of pictures and also sorting of these pictures. Being just a student, a nerdy one for that matter, he felt other students wouldn't share their selfies with him and getting pictures from Facebook or other social media was not an option since it would seem like an intrusion of the user's privacy. Jordan was almost crushed, he felt already that his dreams of building an artificial intelligence was only artificial.😢

‎During the past months, he had met other AI enthusiasts, so he decided to consult the expertise of one of them who enlightened him about the Effect network and the phases through which it was developed, Effect.ai was already up and running.

‎Jordan loaded two simple tasks on the Effect M-Turk

Classify Images uploaded as identifiable or unidentifiable.
‎Label uploaded images( Jordan uploaded images of famous people).
In less than a week, Jordan received some pretty good submissions, he also made room for some double submissions to confirm the accuracy of submissions. A month's job had been completed in less than a week with Jordan's school work unaffected.

‎Jordan proceeded to train thoroughly his AI with the identifiable set of labelled images he had gotten. He also tested it and it produced good results with minimal errors. The productivity of this face recognition machine inspired Jordan to build a business around this face detector he had built. Effect smart market made this an easy-peasy task. Jordan listed his AI algorithm and soon he started receiving offers from a whole lot of businesses, getting paid right from his little corner in the world.

‎As Jordan's customer base grew, he needed much more computational power to handle the available data. At the moment, he had spent all his earnings on school and probably his girlfriend 😅. It dawned on him that if he didn't do something quick about this expanding customer base, he would lose all of them. Just at the verge of giving up, he recalled the Effect power which could help him sustain his AI business long enough to gain grounds again and also improve the strength of his AI services against cyber terrorists.

Jordan became well known in the smart market for his reputable services, owning a lot of other AI services company, rewarding lots of Turkers in Effect M-Turk. His dreams were not so artificial after all.

The Effect Network and it's Reliability

The effect network aims to make liquidation of tokens acquired in the network easy by maintaining a pool of tokens and stabilizing network fees by some set of rules which are subject to change by a governing platform eventually leading to a large adoption of the network because of its reduced volatility.

‎The effect network ensures stability for the requestors and turkers by making sure the pool(Galaxy pool) is protected from external trade conditions. Also the pool tokens(Galaxy tokens) is jealously guarded by the network and it makes sure tokens are available only to the requestors and workers. Also, it can only be converted to regular tradeable tokens by spending it on an Effect smart contract which will could be converted to any NEP-5 token, NEO, GAS, BTC or LTC. In essence, it can be said that Galaxy tokens are utility tokens and not securities which are free from the SEC regulations, unlocking completely the essence of the Blockchain technology - Decentralization.

The Effect network is powered by NEO(formerly called AntShares), a cryptocurrency and a blockchain which enables the development of smart contracts and digital assets. NEO as a blockchain which uses a delegated Byzantine Fault Tolerance(dBFT) consensus system allows for up to 10,000 transactions per second and as a cryptocurrency is rated the sixth coin in coinmarketcap, with a market capitalization of $4.3 billion and an average trading volume of $244 million every 24 hours.
‎This proves to an extent the reliability of the Effect network and gives prospective requestors and workers a glimpse of the potentials of this network in all ramifications from its transactions speed down to its smart contracts.

‎The advisors and team led by Chris Dawe comprises of experts from all of "walks of the Blockchain" coming together to bring home this innovative project that will ultimately transform the already existing AI ecosystem into a more effective and collaborative one. Some of the advisors include Charlie Shrem, Tony Tran, Sally Eaves and Stephen Deurloo and having these experts at the helm of affairs convinces me to a significant extent that this project in safe hands.

Token Sale

Maximum supply of tokens are 650 million with only 40% available for token sale and the others distributed as shown in the chart below.

Commencing on the 24th of march, 11am CET, the token sale ends 18days after its commencement and a soft and hard cap of €4,280,000 and €14,820,000 respectively, leaving a very short window of contribution. The minimum contribution is €50 and maximum is pegged at €25,000. Accepted currencies are GAS and NEO.

It is speculated that AI could contribute up to $15.7 trillion to the global economy by 2030 but this speculation might tend to fail if the AI industry is not placed on the right vehicle to get it there. The increase in proximity of AI developers to data, a market and power is just that vehicle that would increase the efficiency of AI and ultimately effect our lives.

For additional information, see video below or visit one of the information channels;

#Information Channels & Resources
#Effect.ai website
#Effect.ai whitepaper
#Effect.ai lightpaper
#Effect.ai Youtube
#Effect.ai Medium
#Effect.ai Facebook
#Effect.ai Twitter
#Effect.ai Telegram
#Effect.ai Explanation Video