What do teams working with artificial intelligence need to succeed? At Seal Software, which makes contract discovery and analytics software, AI and blockchain experts thrive on trust and empowerment, says CTO Kevin Gidney. Gidney shares his thoughts on AI, innovation, and the value of failing and learning fast.
The Enterprisers Project (TEP): What new technologies are you working with that you think will have the biggest impact on your industry?
Gidney: Something that is having an impact on many industries is deep learning, or to be more specific, deep neural networks. The next developments are going to drive an increased velocity in the adoption and use within legal functions, mainly in-house legal and counsel. What we are seeing is the need for a secure, shareable knowledge system that has the combined teaching of all the parties. As in-house counsel and legal ops are given fewer people but more tasks, a sharing network of information understanding will emerge to cope with demand.
This new infrastructure to allow secure learning from combined knowledge without compromising data security is a challenge, and we have started to develop this based on the combination of AI and blockchain. This is where I see major developments within the next 12 to 36 months, which will turn the use of legal information and external events into a rich source of learning. An entirely new set of skills and resources will be created, where legal professionals create knowledge packs within set domains and share those for a subscription or within an open source model.
Much like the app stores of today, we will have a knowledge store for recognition of legal risk, compliance, improvements and standardization that would give customers an overall rating and suggestions for improvement on every contract produced.
TEP: What are the biggest barriers to having an innovative environment?
Gidney: The technological advancements around AI are seemingly constant. Development teams have to be extremely agile to adopt them rapidly. One area in which we are seeing massive investment is blockchain. The effect of blockchain on societies is being hailed by some evangelists as even more profound than the internet. Others are concerned with security and whether the immutability facet is 100 percent guaranteed. Having resources in an R&D team that can truly see how to utilize this technology for practical applications is a must.
With the interest in AI, blockchain, deep learning, and associated technologies comes one major problem and that is a potential skills shortage. These technologies are complex, and there are a limited number of skilled data scientists and machine learning experts. Being able to innovate is totally dependent on the skills and quality of the experts that you can attract to your organization. In the future, innovation will require not only a conducive environment (machine processing power for example), but also a trained and skilled staff.
TEP: Machine learning holds vast potential. What resources did you put in place to help your team experiment?
Gidney: It’s funny, as all data scientists are really children at heart. They are keen to learn, keen to try, are not afraid to fail and actually like to fail, but as fast as possible, so they can move on. With that in mind, doing things manually (like setting up systems) is just not an option. When you have many data scientists, they can compete for resources, or have systems used in the least efficient way. As a start-up, we don’t have a bottomless pit of money to allow us to run all systems all the time.
So the developers created a new test and deploy framework, that basically takes all the tasks from the scientists and lines them up, starts all the local or remote nodes, issues the jobs, and then publishes the results via a messenger channel such as Slack. This has increased the time and efficiency for them to fail or learn fast. In addition to the processes, it’s important that the developers feel trusted and empowered.
This is one of the core things we put into place early: trust and autonomy. It’s not always about what you do; it’s about how you feel about the people above, beside and with you. That means they have the confidence to know they can do close to anything to look at improving the product.
