Monday, August 26, 2019

Robot, AI, Pengacara: siapa yang akan diandalkan oleh klien di masa depan?

I’m a lawyer.

I sometimes mention the words “digital transformation” to my colleagues and get a blank stare.

Living in a connected IoT world, working in a digital, mobile manner using cloud computing services, for instance, is all still very new to many in the legal world. In some respects, we remain an old fashioned industry. However, the use of automation and in particular, deep learning and artificial intelligence are now starting to change the way lawyers are working. Over time, the continued advancements in AI and robotics will revolutionise the legal industry.

Should lawyers be worried? And, if yes, how so?

 A Starting Point: Automation vs Artificial Intelligence

Artificial intelligence is something we cannot ignore. But there are two areas where AI often gets incorrectly blurred: the areas of automation and actual AI itself.

Often, when one says the words “artificial intelligence”, the immediate reaction is to think of hardware in a factory which makes humans redundant because the humans are no longer required. Robots or machines will essentially steal jobs. To be fair, there is some truth to this, particularly in the middle class manufacturing sector, where automation and robotics may have an effect.

For instance, simple chatbots are replacing humans in call centers. Uber drivers will be replaced by autonomous cars. Industrial workers are being replaced by machines and robots. And everything in your house will be practically controlled by your cell phone or some other device. We are in the midst of a “Fourth Industrial Revolution”. The Internet of Things. Everything is connected, digitally.

But in spite of this job loss, automation should actually boost wealth, creating new jobs in the digital and service sectors. The replacement of manual, laborious activities in our lives using automation will result in greater efficiencies in the work place, leading us to work better.


Automation is very different from actual AI itself. AI goes one step further than pure automation, which simply is a function of — as the name suggests — doing things automatically. AI is about trying to mimic human intelligence.

AI requires a lot more work than what is required in order to automate. Automation is easy to predict, while AI, just like the human brain, can be programmed in many different ways. That is why the use of AI will, possibly in some sectors of many industries, still take many years to accomplish.

AI is, however, already being used in your everyday life and in many subtle ways. Going to Amazon and seeing what you might be interested in purchasing. Turning on Netflix and seeing it predict shows you might want to watch. YouTube’s recommendations. Spotify’s Friday Release Radar. Fraud alerts from banks, which then prevent your credit card from working.

The level of deep learning required by more intricate AI requires more than just new jobs in the service sector — it requires humans to learn, to analyze, manage and ultimately train the machines and robots which are attempting to perform human roles. That is now leading to a whole new level of job positions which requires training not just in universities, but starts in high schools. Coding is now becoming as relevant to your child’s education as English and Mathematics were years ago.

The Modern Legal World

So, what does this future hold for the legal industry? Will clients in, say, twenty years, be obtaining legal advice from robots? Chatbots? Or sooner?

Will clients even need human lawyers? Will judges be making decisions based solely on AI?

Much of the legaltech world boils down to two concepts: (a) matters which require data input and export and (b) matters which require legal knowledge and analysis.

Legal Data

Simple matters: forming companies; simple paralegal type functions, such as bundling documents; basic research and due diligence; issuing very simple legal opinions; drafting things like wills and basic court applications; and matters extending to the administrative functions within law firms, such as performing KYC on new clients and HR functions. Many of these processes are form filling exercises. Any process which is largely basic and can be conducted in volume based on precedent, will eventually be more or less automated, with the correct programming over time.

Will this lead to fewer lawyers? It will certainly lead to greater efficiencies in law firms — a lesser reliance on human administrative functions (HR, KYC, paralegals, even perhaps more junior associates/assistants). This is a combination of both automation and AI at work. There are some examples of firms already using AI technology, e.g. to weed out employee applications, as well as due diligence platforms such as Luminance, which is being used for M&A transactions. And there are great examples of contract review automation such as LawGeex which are becoming increasing popular platforms for clients.

Blockchain technology and in particular the use of self-executing smart contracts will also gradually create disruption within the industry as lawyers are needed less. The way data is generally kept using a decentralised secure and easily accessible digital ledger will certainly have an impact.

It will certainly affect how firms invoice/bill their clients. When contracts and legal opinions can be reviewed using a high level of AI with little-to-no human input, the traditional model of billing clients on a “time spent” basis is now practically obsolete. Lawyers should now be adapting to AI technology, because if they do not, they will lose clients to other firms which have adopted the technology faster.

At the very least, firms should look at adapting legal technology to help them review and assess their clients needs and to use that technology as a new means of working in a collaborative fashion with clients. Allen & Overy’s Fuse is already a good example of this kind of legaltech.

 Legal Analysis and Robotics

So, robots and the law. Can a judge be a robot? Can robots give more complicated legal advice based on legal analysis? This isn’t automated — it is something which requires very deep learning by the machine and AI programmers behind it.

AI will be very, very intelligent in the future, that is certain. And there are examples of risk-assessment AI tools already being used in some courts in the United States.
But technology can have its limitations.

For example, there are matters where human intelligence may be required, such as a complex court case where human creativity and judgment are needed in order to obtain the correct result. Or an extremely complicated legal opinion, where there are intricacies which really need a human touch to achieve the correct result.

That is not to say — in the very distant future — some of those limitations cannot be overcome. But it is not possible to predict the future. There is, however, one thing which AI will struggle to replicate… and that is empathy. Or at least, when we ever get to the point AI is able to replicate empathy, then I would expect that should be in the very, very distant future.

Digital transformation and in particular, artificial intelligence is now becoming something which is changing the way we live. Artificial intelligence has been described by Steven Hawking as humanity’s last invention and that it could even spell the end of mankind.

Even if traditional lawyers don’t like it, the legal world is transforming as a result of AI and deep learning. But bear in mind that AI is programmed technology. And it is only as effective and useful as those humans programming it. So it is absolutely vital to ensure those programmers have the sufficient skills and experience to do so.

And I’m pretty convinced that fairly soon, those blank stares I often get will disappear.

Author’s note: this article was initially published in July 2017 and updated in February 2018.

No comments:

Post a Comment