Bitcoins launch in January 2009 plus its rise as world impact throughout the last 9 years prove that blockchain technology is changing the world. Since 1991 when the idea of blockchain technology was first described by Stuart Haber and W. Scott Stornetta, and most significantly since the launch of Ethereum in 2015, this technology is showing a huge impact on global economics. Ether sold over 7 million ETH in 12 hours, at a price of 1 BTC = 2000 ETH. It has later increased in value, up to 1 BTC ~ 7 ETH.
We want the KaaS blockchain to be more than just a store of value or a solution for renting computing power. Kaasy, the artificial intelligence agent we are developing on top of the knowledge blockchain, will also have the ability to share knowledge with spin-off units bi-directionally, both providing new skills and gaining new skills from them. If someone trains a narrow AI to execute a task, say identify any star on the sky from an image taken without GPS coordinates, when this skill is integrated into Kaasy, it will be combinable with previous skills. Later, someone else can spawn another narrow AI, with the already-integrated ability to identify the GPS coordinates of the camera that took a specific image, when given the date and time of the picture. The date and time might even be read from the picture itself, if someone is looking at their watch and there is a calendar or newspaper on display somewhere.
The goal is to have skills on demand, to assist people in whatever task they have to accomplish at some moment. This will increase the productivity of anyone using the network, freeing up their time to do more meaningful tasks that the mundane, already automated ones.
The system also offers an instantly-scaling solution. Today, if a company needs 50 employees for a specific task, they have to train at least 50 people, depending on employee churn rates. Using Kaasy, they get a compiled AI with the prerequisite skills, they train only the top-most skill, and then they can replicate as many as they need. Should one of the spawns generate some new knowledge piece, due maybe to a difference in the quality of the data it was fed, they can all be updated, and gain the ability to react correctly in the future. If the task can be parallelized and executed by 500, 5000 or 50000 units, the time to completion can be reduced significantly.
The end goal is to develop the infrastructure (both software and hardware) where each person can have a personal AI assistant, tailored to their activities, with private knowledge saved privately, yet with inter-compatible skill sets that can be replicated from one AI to another. For the ones of you that read Manna by Marshall Brain, these AI assistants can be integrated into the computers each person had at the end of the book. Seamless integration, allowing skills to be shared between person and AI.
We have put together all our ideas, intentions and predictions for the kaasy.ai project into a downloadable PDF Whitepaper. Please take the time to read the document and then contact us for any questions you may have.
40 Bloomsbury Way, Lower Ground Floor, London, WC1A 2SE