Back

 Industry News Details

 
This Startup Is Dreaming of a Global Brain on Blockchain Posted on : Feb 11 - 2018

Artificial intelligence is one of the most celebrated and hyped concepts today. From science fiction to the nightly news, AI has been making record headlines, whether it will become the most dangerous technology around or the one most likely to save humanity. And it isn’t the only technology riding the hype wave. Cryptocurrencies and blockchain, once the territory of dark net denizens and righteous Redditors, are topics of conversation in even the most venerable and stodgy of publications.

What happens when you combine the two? Yes, a buzzword Armageddon. But also a fascinating vision of a future in which AI and blockchain converge.

Decentralized blockchain technology may hold the key to tapping AI’s potential in a democratized way, one that opens participation in the AI economy of the future to more than just a tiny handful of the largest technology behemoths. And blockchain may even be used to not only develop AI but also potentially to function as the white matter linking today’s narrow AI into a global brain.

AI for All

Ben Goertzel, a leading AI researcher, thinks the key to steering us towards a future in which AI is the best thing to happen to humanity is to develop it in an open, democratic, distributed way using blockchain—which is why he’s developing SingularityNET.

SingularityNET is a decentralized platform for AI. Goertzel and his team aim to build a blockchain-based infrastructure to enable various kinds of AI algorithms—from image recognition to natural language processing—to flexibly interact with one another in real time. The system will also be a way to track which algorithms are being used and to compensate developers accordingly.

Instead of humans manually stringing together algorithms, as the system develops, they’ll be able to communicate data and coordinate processing with one another. In the system’s initial incarnation, a user who has a task to complete using AI, training a robot to dance, for example, would send that task to the system, which would then parcel it out to various algorithms specializing in the different skills required to complete the task. The developers whose algorithms are used to complete the task will be compensated by the system with the tokens the user spends to get the task completed. View More