The Botchain benefits any person who interacts with intelligent agents that are supported by the its network. Many users of intelligent agents and other products powered by artificial intelligence are unaware of the security risks that will need to be solved, especially as those systems become increasingly intelligent.
Intelligent assistants that automate work for humans are a new, exploding category of software. While early bots were mostly scripted and did not actually have any intelligence built behind them, new tools and machine learning approaches to natural language processing are now bringing real A.I. to these agents.
As bots get more autonomous, they will sometimes make subpar decisions, much like humans do. Today, two apps talk API to API, but what happens when these API endpoints are suddenly intelligent and adaptive? What happens when bots negotiate with each other, and autonomously enter into smart contracts on behalf of your company? Bot-to-human and bot-to-bot communication creates multiple problems with existing models of workflow and agent governance. For example, if a bot makes a simple decision on your behalf when talking to another bot, there is no independent, third party, immutable trail of that conversation. What if you need to audit it? What if you need to understand what happened so you can re-train your agent model to make a different decision next time? How do you ensure some version of this error isn’t happening to your agents? Botchain solves these problems.
We already have several committed bot partners who are interested in using the Botchain to increase their security, compliance, and all around product offering to their end users. We will announce them before the ICO.
We don’t believe that bot marketplaces should be controlled the way that Apple and Google dominate the mobile app marketplaces, with their cumbersome centralized approval processes. A decentralized solution will allow for more innovation. Botchain supports the ability to add skills and modules for bots to use with no centralized control, just agreement between trusted nodes.
Trust, Identity and Discovery are also major issues for bots at the moment. The Botchain provides security in allowing bots to validate that they are in fact representing the owner they say they are, and that they have the ability to engage in the transactions they are proposing. With every transaction, the Botchain will record a digital certificate that is a hash of all the inputs that went into it - the model “brain” of the bot, the version, the timestamp, the task, the owner, the situation, etc. If you have to validate or audit a bot, these digital certificates will make that possible.
For enterprises to adopt any new technology at scale, it has to meet certain compliance standards. By making bot tasks and communications auditable and inspectable, we take a major step in that direction. A distributed ledger facilitates compliance and makes tampering difficult.
Botchain also expands the ability of smart contracts to support artificial intelligence models, so that probabilistic outcomes can be combined into contracts and workflows.
Botchain is currently in private beta. For developers that would like to participate on the network, sign up here.
Direct press inquiries to [email protected].
General questions can be addressed to [email protected].