Startups

Cortex – Upgrading Blockchain with AI

Blockchain has emerged as a formidable technology having numerous disruptive possibilities for almost all of the industries out there. As it has been the scenario, entrepreneurs and programmers are coming with their own versions of blockchains. It almost seems that every industry is being populated with out-of-the-box blockchain one step at a time.

Speaking of the AI (artificial intelligence), although there has been some work on the AI sector in the centralized world, in case of the blockchain, one of the works that is geared to completely change the scenario for the AI implementation is the Cortex.

Cortex in its simplest description

To state it simply, and for the sake of the average Joe like me, Cortex is the next generation technology that strives to add the power of AI to the blockchain; thus, making the technology more flexible. It has features that target the smart contracts of a blockchain and upgrade them to the AI Smart Contracts, which will serve as a repository of features to allow programmers contribute to the system by introducing AI models necessary for the programming of AI DApps.

Thus, summing up everything, Cortex is all about:

  • Upgrading the capacity of the existing smart contracts by introducing AI to them,
  • Letting the AI experts come up with AI models,
  • Letting the users create AI DApps based on those models, and
  • Incentivizing all the model contributors – the ones that necessarily make the network viable.

By the way, AI DApps are nothing but applications that have Artificial Intelligence running at their core (simple, no?)

Anyways, you might ask a question here.

Why was there a need for something like Cortex?

The need for a tech like Cortex was not just a need, but a grave one (like stressing on the “grave). This is because the existing architecture of the blockchain (Bitcoin and Ethereum) do not allow major computational activities like the ones required by the Artificial Intelligence. In simpler words, the smart contracts that run the networks don’t have the required power to cope with the processes occurring in the core of an AI DApp. Therefore, the researchers and pioneers at the Cortex thought of coming up with something that could use GPU computational powers rather than CPU and that could have an extraordinary capacity for letting the smart contracts decide on their own.

This is how Cortex came into being.

But this is not the stop, however. Cortex is expecting to take their concept globally by allowing AI experts from around the world upload their models on the SDK of the Cortex public chain.

What would that do?

That would inevitably place the Cortex in the center of the vortex of practical usage of the AI. And this is particularly where incentives come into play.

The Token of the Cortex

Although the default coin of the Cortex chain is Cortex Coin or the CRTXC, the basic pricing unit is Endomorphin. It is the total computational spent on the submission of a data model governed by the AI Smart Contract during the inference of the model (fancy name for usage of the model by any user).

The price of the Endomorphins will be governed by the total AI smart contracts executed by the users and the number of transactions taking place. As it is stated in the whitepaper,

It depends on the gaming between miners and people who initiate transactions or execute AI smart contracts.

The amount of Endomorphin spent will be equivalent to a certain amount of CRTXC, and the conversion rate between the two will be a factor of two things:

  1. The total miner’s price of packaging of the transactions in a certain block, and
  2. The number of smart contracts attracted by the model providers.

Of course, the final cap on this conversion rate will be set by the Cortex itself.

Talking about the usage of the Cortex Coin, it will be used to incentivize model providers and the miners. But the most awe-inspiring feature of the Cortex Coin is that it eliminates the chances of any false model submission or over-crowding as a result. As model submissions do not come free, anyone wishing to have their model on the chain would have to pay for the storage. This also would, in turn, encourage model submitter to submit better and stable models.

The core of the Cortex

Talking about the Cortex Coin and the Endomorphins is all sweet, but talking about the rest of the features of the chain is even sweeter.

Behold.

  • Cortex Virtual Machine

This is the inference engine of the Cortex and it is compatible with the EVM or the Ethereum Virtual Machine. It is slightly same as the EVM, only that it has inferring capacities.

  • Model Submission Framework

The model submission framework acts as a medium between the model submissions and the computational power providers; thus, allowing greater collaboration between the users or the participants of the chain.

  • Cortex Inference Consensus

The consensus mechanism on the Cortex chain is created by agreeing on the results of the inferences unanimously.

  • Cortex Intelligence Inference Framework

I have already explained parts of it. This framework lets the researchers around the world submit their AI models and also lets the users make necessary inferences through the AI smart contracts to use the models.

Use cases of Cortex

The promising core of the Cortex enviably ushers us to delve into some of the use cases of the AI blockchain.

  • Financial Services could use decentralized AI blockchain to calculate the credit score of a user or according to his financial data take an apt financial decision.
  • It could help in the creation of numerous AI assistants, chatbots, and the sort.
  • Information services industry could see a revolution in the form of “image search engines”, automatic summarization, or even personalized recommendation systems.
  • In the intelligent decision-making process that could lead to automatic driving, etc.

While these are some of the use cases, Cortex states its goal to be “the leader in the field of machine learning on the blockchain.” In its future projects, Cortex aims to develop efficient data privacy systems, improvement in block size and TPS, and mainstreaming the “AI chips on the mobile devices”.

Tags

Noor Najam

Contributes guest posts to BlockPublisher with industry analysis & news in the blockchain space. Email: editor.opinions@blockpublisher.com

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.