Brian Kuhn: Blockchain, AI and solving the black-box problem

Technomancers caught up with Brian Kuhn (Partner, Co-Founder and Global Co-Leader of the Watson Legal Practice at IBM) to find out more about what the confluence of blockchain and artificial intelligence (AI) might mean for transparency in machine learning (ML) models.

Brian, what was your journey to IBM Watson Legal?

I was a practicing attorney on the financial services side and then I worked for Thomson Reuters, where I had an opportunity to pitch an idea to IBM around addressing the legal market with artificial intelligence.

The idea focussed on corporate law departments with business of law solutions. They took a chance on me and it was phenomenally successful, so I became the managing partner practice leader of a brand new practice within IBM focused on AI and law.

How do you see blockchain and AI working together?

AI is becoming ubiquitous across industries and there are many benefits – but there are problems. I see three problems with AI that blockchain can address:

Firstly, who trained the machine learning model and can you rely on them? How does their expertise influence the recommendations that the model produces?

Secondly, what data was used for training the model? Is it reliable, current and valid?

Thirdly, how do I defend myself if a machine learning algorithm produces a recommendation that’s simply too complex for me to parse?

Blockchain can address all of these problems by creating a forensic audit trail. Whilst blockchain might not be a perfect approach, it is absolutely the best approach we currently have to capture how a machine learning model produced a decision.

Do you see there being further development within IBM around solving the so called “black-box problem” in this way?

I think these problems will be addressed this way across businesses, and the expectation for transparency is going to increase as blockchain becomes more ubiquitous.

No one wants to rely on black box technology and if they have the freedom not to they won’t – they will demand insight.

What do you think will be the quick wins with blockchain and legal – which areas you can see where blockchain will become pervasive?

First of all I think it will be blockchain and AI! A subtle application for a new kind of security – protecting and guaranteeing the efficacy and provenance of the insights that AI produces. It’s not that complex to do that.

I also feel very strongly that in order for solutions to be developed, blockchain networks will need to focus on the current state of affairs when it comes to how people within a business network transact. For example, for microwaves and refrigerators that can be plugged into an electric grid, there needs to be more focus on that electric grid and on what the business value chain is that you’re trying to augment.

I don’t hear enough conversation about that from law firms, but I sure hear about it from clients because they’re doing it across their other lines of business and will certainly be turning to legal to do the same.

You’re a member of the GLBC (Global Legal Blockchain Consortium) what does that mean for you?

It comes back to the to the AI and blockchain confluence – an opportunity to explore something that’s going to be profoundly important.

It is really interesting because it’s not typical that this industry starts investigating a technological process first. It’s the opportunity to participate in owning the narrative and participate in being part of a defining new value that interests us.

Where should law firms begin if they want to start looking at blockchain and its use in legal?

I would suggest that they look outside the legal industry and they look to what the big four (consulting firms) are doing and they get a sense of what industries have had success.

They should try to form a point of view based upon what’s already worked rather than starting from scratch because there are already a lot of very good lessons to be learned.

 

 

Photo by Scott Rodgerson on Unsplash

Related Posts