Posted by Dan Bindman, Associate Editor of Legal Futures
The word ‘chatbot’ has been appearing with increased frequency in Legal Futures stories over the past year. As a popular human interface for computer software backed by artificial intelligence (AI), it is a phenomenon we are likely to see yet more of.
The combination of ‘chat’ – an essentially human activity – and ‘robot’ neatly encapsulates this interface. Other names for the technology are chatterbot and artificial conversational entity.
The application of chatbots in the law appears to be growing. As we reported last year, an Australian lawyer, Adrian Cartland, is using one to connect customers in a shopping centre with legal documents that can be readily automated, such as wills.
Josh Browder, the young founder of the DoNotPay chatbot, which has been used successfully almost half a million times to challenge parking tickets (I have used it myself), has extended it to multiple other areas.
He told November’s Legal Futures conference of his vision of chatbots linking directly with online courts and helping litigants in person. This is a digital native who was coding apps not long after he learned to walk, so anyone born in the analogue age should listen carefully.
It is tempting to dismiss this latest manifestation of legal technology as a fad which could never replace a good lawyer who is close to his or her clients, who faithfully serves the community’s legal needs, oils the wheels of local commerce, and so forth.
Or could it? Chatbots are already a normal part of online life – the digital assistant that you see at the bottom of many commercial web pages, asking if you need help, or want to talk.
They invariably have normal-sounding, region-appropriate human names, so as not to frighten the punters. In the UK it may be Emma, in Spain Javier, and in India Lakshmi.
If confronted by an enquiry their vast databases cannot deal with, it can be seamlessly passed to a human and you will never know it has taken place.
This could happen in legal practice. A lawyer sits at their PC, overseeing 100 chatbots, ready to step in if a client demands ‘proof of life’ or confuses the bot’s algorithms.
Chatbots are almost as old as the internet and far older than the worldwide web. The first one was known as ELIZA, created by Joseph Weizenbaum – later a stern critic of AI – in the mid-1960s. Using a technique known as pattern matching, it gave users the illusion that the program understood natural language.
One script it used mimicked a psychotherapist, repeating everything back at the user. What was extraordinary was how well humans took to it, with many conversing for long periods and insisting it had real intelligence.
This challenges the notion that people will only trust face-to-face interactions, although there is force behind this argument as well, particularly with older people.
Could this be the technology that finally persuades people – especially digital natives – to make a will, one of the hardest nuts to crack among unmet legal need?
With lawyers at the top of the pile validating the accuracy of the product and standing by to deal with complexity, with AI working away behind the scenes cheaply and tirelessly, and the chatbot providing an interface the public can engage with, this could be the legal services business model for the future.
Everyone captured by the attractions of this model insists it will expand opportunities for lawyers rather than kill jobs – just look at the activities of Billy.Bot, the chatbot barristers’ clerk launched last year by Clerksroom.
They cannot know this for sure, but the history of technological development is surely that as well as sweeping away whole industries, innovation can generate completely new markets and fortify old ones.
There is no reason to believe law should be any different.