Aldo Agostinelli

Decentralized networks and AI: according to the experts, within 10 years, the latter will be relying on the former and AI and Machine Learning will be the foundations the development of deep tech businesses will be based on. Read it In Italian.

Current figures seem to confirm such forecast: according to KPMG in 2017 venture capitals supporting AI have doubled compared to 2016, reaching up to 12 billion dollars; at the same time, the Blockchain has cast the foundations for decentralized internet, gathering over 5 billion dollars through ICO (Coindesk).

However, according to Sherman Lee, expert of AI and partner of Zeroth.AI, for such transition to turn into reality, three different elements have to be taken into account: Big Data, algorithms and IT resources.

Basically, what they are saying is that if we want to witness to the birth of  a decentralized and smart digital world, we first need to build the related high- tech infrastructures and several different protocol levels (AI Will Take Over Blockchain In 2018, But It Won’t Happen Without These 3 Key Areas).

With reference to data, Ocean Protocol is building a decentralized data exchange system  to collect information concerning the AI development and allow businesses and consumers to make use of an ad hoc protocol. Provided everything turns out well, in the next few years even smaller businesses will be able to access Big Data, and several projects like this one will be devoted to specific data sets.

And since global players are not interested in sharing their data, objects implying the use of the Internet of Things ( IoT ), will play a more relevant role, i.e. collect the info needed for the evolution of AI. IoTeX, for instance, can create a blockchain working onto IoT devices for this purpose.

However, if data are essential for AI and Machine Learning, these latter also need the algorithms to process them which will have to be centralized, as well. Not an easy task, anyway. Nevertheless, there is already someone who is working on it:  SingularityNET uses AI to make algorithms communicate, while Deep Brain Chain uses AI to make algorithms work on inactive GPUs. Cortex, instead, is a project which allows users to write machine learning software and to communicate with each other and exchange data within a safe and protected environment, thus facilitating cooperation within the AI community.

Finally, the power of calculation: some smart techniques are needed to exploit the power of AI and Machine Learning. The Golem project has led to the creation of a marketplace for calculation power, while Raven Protocol has developed a technique to distribute deep learning calculations  to any device in the world.

This way, Blockchain based AI may soon become a reality available not only to the giants of the digital world.

What do you think about  the fact of having the AI on a  decentralized log? Will it boost its development? Tweet @agostinellialdo.

To find out more about the digital world, you can read my new book: “People Are Media” 

If you liked this post, you may also like Simplifying big data may be useful, but without overkilling

Aldo Agostinelli