Effect.ai — What it is and where it’s going.

Mark Westerweel
Feb 14

Artificial Intelligence. For a lot of people, the first thing that comes to mind may be a frightening image of Skynet taking over or HAL-9000 going haywire. Mechanical demons, powered by supercomputers will transform the world into a terrifying place, if not — destroy it altogether.

Honestly, I think we should not dismiss the warnings of Steven Hawking and Elon Musk for there is truth in it. A.I. is becoming a powerful technological entity and for the first time in history, given enough time, we might be seeing something that can mirror our own intelligence, and eventually most likely go well beyond it. In my opinion, it can be the tipping point between equality or elitism

The potent Artifical Intelligence of the future is being developed today as part of the Fourth Industrial Revolution, and the way I see it, is something that should be kept in check and regulated to prevent its development getting out of hand, or monopolized by a corporate entity or a nation with a state capitalist strategy like China.

Do we really want corporations like Facebook, Amazon or Alphabet(Google) to capitalize on an entity that can think for us, outlines our next actions and can guide us? I certainly wouldn’t want it! We would become data-producing cattle to an oligarchy, and I fear for the day if that came true.

Or how about China? It has already introduced a social credit score to discriminate citizens. If China becomes the dominant algorithmic society by leading the Fourth Revolution and imposes its privacy policy (which is no privacy) and standards on the rest of the world by A.I. algorithm. One would be restricted and confined by default. Any camera would recognize us and by an algorithm, our lives are determined. I’d say that’s worse than Skynet. Chance is, though, the way things are going now, with the high cost of developing A.I. that one of these major companies or possibly a nation will develop the technology behind closed doors and it will not be for the good.

Luckily, little over 10 years ago, Satoshi Nakamoto released Bitcoin and the broadcasting network it uses gave us the possibility to distribute data in a decentralized way. Nakamoto gave us possible freedom from fractional reserve banking and institutional currency. We could now bypass banks or governments so Anna could send funds to Bob because Anna could now trust a system without the need for a third party. A few years later, smart contracts were developed and now Anna could send and receive transactions under certain conditions. If we skip a bit more, we get to the point where Anna can send transactions in the form of microtasks and receive compensation by smart contract. Beauty is, we can all be like Anna. No exceptions made. Gotta love blockchain and decentralization.

This is where a great project is coming into all of this — Effect.ai . It’s an open, decentralized network that provides services in the Artificial Intelligence market (and it’s Dutch! Yeah, sorry, proud it’s coming from my country). What this essentially means, is that this will be the place to go to if you want to develop your Artificial Intelligence in an open platform that doesn’t have the disadvantages of a centralized platform (with a vertical structure, you never high up the chain). The ecosystem can develop more organically in this way (see phases later on).

So far, computers can do all kinds of amazing things and even became really good at beating humans with chess, go and Starcraft, but still can’t tell the difference between a cat and a dog — it will still need human intelligence to develop its algorithms. Cool thing, we can help!

There has been some controversy regarding the use of this human resource since Google’s Project Maven AI (together with the Pentagon) used human intelligence to differentiate between human targets and surrounding objects. Without workers knowing what they were contributing to for as little as $1 an hour.

Effect.ai will be clear about the Requestors as well as going for more than fair compensation. The distribution of labor will also distribute compensation for the tasks done by the workers. Since there will be no discrimination for becoming a Worker(unlike Amazon’s Mechanical Turk), anyone who qualifies — read: has a basic level of an available language or passes the data labeling qualification- can earn a basic level of income (around $8)once the platform goes live. I have lived for over a year in Nepal and know this would make a huge impact, some worked their ass off for $40 a month.

At the moment Effect.ai is in its first phase, called Effect Force. From the lightpaper:

Effect Force is a private, decentralized, marketplace for work that (for now) requires human intelligence. It is based on centralized business models like Amazon Mechanical Turk, Fiverr, OneSpace and Guru.com. It’s a worldwide workforce on demand. A key difference with these centralized models is that the Effect Force is peer to peer, meaning supply and demand are connected directly and more efficiently.

Interested parties (Requestors)can knock on effect.ai’s door and request training for their algorithms, which require the human intelligence of the Workers. Workers receive fair compensation for their HITS (human intelligence tasks)and are rewarded with EFX, Effect.ai’s utility token, by smart contract. Workers that deliver good quality work are getting a higher reputation score (ERS) to promote good behavior and qualify for better-paid tasks. The ERS gradually expires over time,

These microtasks can be transcribing audio files or data labeling, which is basically a more advanced form of the ReCaptcha we have to do before logging into any site nowadays. I tried some, and they don’t take too long. I can imagine myself doing some while traveling by train.

At some point, these tasks could be done by bots or Artificial Intelligence. This will actually be allowed on the Effect Network and creators of fully functioning algorithms will receive EFX for their services. This leads to The Effect Network’s second phase — the Effect.ai Smart Market.

The first phase shows why a distributed ledger comes in handy, to settle transactions between Requestors and Workers or creators of said algorithms.

The second phase makes use of this technology as well. It’s here where algorithms can be found for their services on the Smart Market. Application owners can register on this platform by specifying a public endpoint for their application following the Effect.ai data interchange format. In this way, the application can be invoked through smart contracts. The caller of the contract needs to transfer funds to the owner of the contract to get an authorization token that allows interaction with the application. In this way, it’s not the workers getting compensation, but agents supplying the AI algorithms will receive EFX from the Requesters

There are two vital components of the Effect Smart Market, namely the application registry and the endpoints.

The application registry will keep full track of the applications, with descriptions about its origins and a technical schema of its inputs and outputs. Algorithms can work interoperably through this registration and it aims for standardization of data exchange formats. If applications work interoperably, more interactions can be explored for more financial gain.

The effect.ai lightpaper on the security of the endpoints and phase 3;

Access to services and algorithms made available by users of the Effect Smart Market will be managed through endpoints. Application endpoints on the Market communicate over the HTTP protocol. Data is exchanged in JSON format and should strongly confirm the defined RDF schema. Requests signed with the private key of the buyer will be accepted by the endpoint. Issuing authorization tokens and checking their validity can be done by public APIs that hold a partial index of the blockchain. It is possible a small fee will be required for using this service.

Phase 3
Phase 1 and 2 of The Effect Network decentralized the data gathering and usage of AI algorithms. Up to this point, the algorithms themselves still run on centralized servers. In the final ‘Effect Power’ phase of the network, the actual computation will be distributed so that the algorithms run globally without a single point of failure. To achieve this, The Effect Network uses the fact that most machine learning algorithms have a rigid structure and operate on sets of weights. These types of algorithms are relatively easy to distribute. The Effect Network decentralized compute engine is based on popular Deep Learning (DL) networks like Caffe15, MXNet16, and TensorFlow17 where the network structure can be defined as a declarative graph and weights are stored as matrices of real numbers. These matrices can be distributed over a decentralized file system and be processed at different compute nodes on the network.
More detail on this phase will be provided as the project progresses.

Effect.ai started out on NEO and still is number 1 dapp on that blockchain network, but NEO can’t handle the volume of transactions needed for a dapp like Effect.ai, so it’s migrating from NEO and will soon share details about their new home (there are rumors of EOS- which would be HUGE, knowing the American base).

EFX is mostly traded on Switcheo . More exchanges coming soon!

I know I will be writing more on Effect.ai but thought this is lengthy enough for an introduction.

Website: https://effect.ai

Force: https://beta.effect.ai

Github: https://github.com/effectai

Telegram: https://t.me/effectai

Twitter: https://twitter.com/effectaix

Facebook: https://www.facebook.com/effectai/

Medium: https://medium.com/@effectai

Reddit: https://reddit.com/r/effectai

LinkedIN: https://www.linkedin.com/company/effectai/

Bitcointalk: https://bitcointalk.org/index.php?topic=2737469.0

Funfact:

The name Mechanical Turk comes from the first automaton, the first chess-playing machine, but turned out to be a fraud.