Oct 30 2017

A few thoughts on artificial intelligence for both the sceptic and evangelist!


I find myself having more and more conversations about AI (artificial intelligence), whether it’s in Legal IT circles such as the recent panel I chaired at LawExpo 2017 in London or just a discussion with my brother on the ethics of the impact of this “next revolution” on the job market. My stand point has a mix of scepticism about how quickly this will happen and a dose of optimism that we will find a way to navigate this shift without mass unemployment.

I sometimes think though I am alone on both counts though. An example is this long but well written article on how we’re all going to lose our jobs to robots, for those in their twenties this article basically says you’ll all be out of a job before you retire! But then there are other more tempered articles that suggest not all jobs will expire but that we’ll adjust to use the tools in our jobs rather than the tools replace them. Or other more positive views of a future business world with AI.

My sceptical side can’t help but pull out articles like this one from 1992 which talked about the future of speech and pen recognition. “If I were a researcher, I’d feel I had a better return from studying speech recognition” [as oppose to handwriting recognition] and comment about how great a stylus would be for painting and drawing. Some 25 years later with Alexa and the Surface pen/Apple pencil we just about have the tech to achieve each to a reasonably successful degree!

But we will undoubtedly see a revolution that eclipses the industrial revolution and transforms the world of business, whether it’s in 5 or 50 years. And I look to the tech leaders like Bill Gates and Elon Musk who are sounding a warning that pretty much every politician is ignoring.

“I’m increasingly inclined to think that there should be some regulatory oversight, maybe at the national and international level, just to make sure that we don’t do something very foolish. I mean with artificial intelligence we’re summoning the demon.” – Elon Musk

Bill Gates has proposed possible solutions to managing the problem with taxes on robots and the long article linked to first in this post has other options for tackling the transition. We do need to start looking at what this revolution might mean for business and more importantly society, in a very barbell shaped economy more technology will only erode the middle even more and probably also erase the low end job market entirely!

However I think we have time, AI will form a huge part of technology solutions in the next 10 years I am sure, but in very specific applications and the majority of these I think will be in assisting humans rather than eradicating them. After all there will be plenty of companies that follow the Dilbert principle in this cartoon!


Jan 4 2017

Experience in Artificial Intelligence – lessons for Legal IT in 2017


To start 2017 I thought I’d have a quick look at 2016’s favourite Legal IT topic, Artificial Intelligence (AI). Reason for picking this topic is down to two pieces of technology that I brought into our household at the end of last year that utilise AI heavily. Those are Nest and Alexa.

The former has brought home to me what I think is going to be a key issue in adoption of AI in law firms, that is trust. My Nest thermostat “learns” over time and one of the key aspects of this is watching for when you go out of the house, the system then reacts by turning thee heating down. At the same time it triggers my Nest security camera to turn on. Over Christmas though I’ve noticed, on odd occasions, when we were out that the camera would turn on but the thermostat wouldn’t drop the temperature. Other times though it would, there seemed to be no consistency. After a lot of old style IT troubleshooting and a lot of googling, I eventually found that this wasn’t a bug, but that Nest had learnt patterns and our locations and kept the heating on when it thought our trip was local and brief, thus in its mind turning off the heating was less efficient (as otherwise it would need to heat the whole house from scratch on our return). The camera though it realised should be on immediately.

This is my trust point, until I fully understood what was going on I didn’t trust the technology. I thought it was just not working, so I had taken to manually overriding the settings when we went out. In reality it was working very well and actually better at predicting things that would save energy than I was! This issue though will be the same with Legal IT AI, getting the trust will take time and will probably need a full understanding of how the machine is learning before people will accept AI into the mainstream functions.

Alexa to me is voice recognition starting to become useful, moving from the smartphone (Siri) or the computer (Cortana) to being “in the room” makes so much sense and is much more useful. Whether it’s controlling the lights or simply putting on some music it genuinely is useful rather than a gimmick. It is impressive how Alexa is using all the data it is gathering to improve, however it also shows how far AI has to go in terms of human interaction. It is way beyond having to specifically phrase your commands or questions, but there is so far to go to get beyond a few “skills” it has now.

These two areas fuel my scepticism around AI. No that’s not fair,  it’s not scepticism it’s just wanting an injection of reality into AI within Legal. I am impressed with Alexa and Nest and the more I use them the more I get impressed by the learning, however my expectations for the technology (particularly for Alexa) were not overoptimistic. I think if we adopt a realistic approach for AI in Legal in 2017 then it can and will be a great enabling technology for firms. If we don’t temper the hype though, we’ll be disappointed or fail to trust it and it’ll go in the bin for years before we try it again!