How to build trust in AI with blockchain technology
Machine learning and AI are all about data. We give AI-powered systems massive amounts of data to ingest and analyze, and then we trust the results. But I want to know how we can ensure the integrity of those systems, from the algorithms to the data sources.
Solving the AI data-integrity issue
I believe the issue of data integrity is a huge challenge for the AI community, both for users and developers. And I don’t think we’re looking at it closely enough. AI results are only as good as the data on which they’re based, and I think we should ask how and whether we are verifying the authenticity of that data. We already have problems with fake news, fake photos and fake identification.
We don’t need these kinds of issues where people are going to be making critical decisions based on artificial intelligence. For example, as those of us at Kilroy Blockchain were developing our IBM Watson AI-enabled app called Riley, which allows people who are blind and visually impaired to use their smartphone to get a description of their surroundings, we wanted to be sure that no one could come in and hack the dataset. That’s when I really started thinking about the data authenticity problem, and I realized no one had yet gotten out in front of it. I thought to myself – this is a train heading for a cliff!
Creating new synergies with blockchain and AI
I’ve come to believe that we can use blockchain technology to provide that “trail of truth” for the data in AI-enabled systems. My true aha-moment came to me at a blockchain hackathon in Austin, where I saw business teams with great ideas, but they lacked technical skills. They were trying to solve a number of related business problems, but they needed a platform to connect their applications to so that they could focus on their business ideas. That’s when I decided that Kilroy could solve the problem by providing a blockchain platform-as-a-service offering!
Now we’re providing our blockchain solution to enterprises that need to authenticate data before feeding it into an AI system. Blockchain helps prevent anyone from hacking a system and altering the data, because all transactions involving that data are recorded on a distributed, immutable ledger. I also believe blockchain and AI go hand in hand, and that’s why Kilroy is working closely with IBM on other projects, which means that we don’t have to reinvent technology solutions where IBM has the expertise.
Our Riley (RealLife Adventure) app was born when I coached a team of dragon boaters who were all blind or visually impaired. They had trouble navigating their surroundings. Riley uses Watson Content Hub as well as Watson Visual Recognition and Watson Text to Speech to describe real-world objects to the user. Our Riley app won the IBM Watson Build Challenge for North America.
We’re also using the IBM Data Science Experience and Watson Machine Learning to build our Carnak solution. Carnak stands for Cognitive Roadway Knowbot, which we invented to help cities better harness the wealth locked within their data. Carnak can help cities visualize the data and pinpoint specific areas on which to focus their resources, such as putting first responders near a high accident area. Since many cities have common data, it will work from San Francisco to New York and everywhere in between. In the future, we plan to introduce machine learning to Carnak so that city solutions can keep getting better over time.
Moving forward with blockchain and AI for the enterprise
My advice on blockchain and AI is to start small and not try to solve every problem at once. Often, consultants will recommend that a business reinvent all its systems to create a completely new one based on AI and blockchain technologies. I don’t share that philosophy. I believe it’s better to find touch points for AI and blockchain in what you already have, and build extensions from those touch points.
Today, businesses are ultimately built on networks of people and solutions. In the end, blockchain is all about building trust. Trust in the data, and trust in the blockchain. If you start small and show results quickly, I think you can build that trust within any business network.
See how Kilroy Blockchain is building synergies between AI and blockchain by watching the video featuring Karen Kilroy: