A Lucrative Micro-Niche: Deep Learning

Topics: Big Data, Data Analytics, Efficiency, Law Firm Profitability, Law Firms, Leadership, Micro-Niche, Midsize Law Firms Blog Posts, Practice Engineering, Process Management

deep learning

In our continuing series on lucrative micro-niche practice areas for lawyers, we examine deep learning and how artificial intelligence and big data can combine to give lawyers and law firms a way to mine for business in this rapidly expanding field.

The amount of data we generate every day is staggering — currently estimated at over 2.6 quintillion bytes. 2.6 quintillion! It’s a number so large that 2.6 quintillion pennies would, if laid flat, cover the entire Earth… five times. And that number is only growing with the increasing number of devices being made available and being used.

If data is the new oil, what new products can we make? That’s where deep learning comes in.

Deep learning is a micro-niche of artificial intelligence (AI). It uses mathematic procedures called algorithms to learn and recognize patterns. It’s nested below AI, machine learning, and “neural networks” because it’s a multi-layer neural network. What’s amazing is that deep learning enables computers to learn. Since we’re awash in data in the legal industry, we stand to benefit from learning how our clients are using it.

The basics are simple. The first challenge is to identify the types of problems that are most susceptible to being solved with this technique. In general, we mean data (numerical or text-based) that fit a category, classification, or label. The next step involves identifying a good supply of high-quality examples of that category, classification, or label.

If you can identify both the problem category and data-based examples of it, a data scientist can create a deep-learning model. And with that, deep learning becomes “just software” that can, from new data, identify the patterns you’d like to see in order to either increase revenues or decrease costs.

“Not surprisingly, early adoption of deep learning has happened in companies with access to massive amounts of data which can be analyzed to solve their problems,” says Pankaj Goyal, vice president of Hewlett Packard Enterprise’s artificial intelligence business. “This data might be in the form of text, image, voice, or video. These problems could be automation, new customer experience, or new product innovation.”

Your legal clients already are starting to use deep learning (and blockchain) and that mean you should start thinking about the problems that may arise from their use. For example, clients may have to show a court how their internal deep-learning system works; whether bias exists; and whether anyone’s privacy has been compromised. Moreover, you should be aware that courts may be sensitive to the revelation of patterns in their rulings. Earlier this month, the French government banned the publication of data related to rulings by French judges. A five-year prison sentence awaits anyone violating the law.

That said, we’ll switch to a few of the client industries you or your firm may serve that are working with deep learning, including:

Aerospace & Defense — Companies in this industry are employing deep learning for facial recognition to be used in security-screening purposes, remote sensing, object detection and localization, spectrogram analysis, network anomaly identification, and malware detection. Deep learning is also being applied to manual tasks and pilot operations in the cockpit as well as wearable computing for soldiers.

Automotive — The way an autonomous vehicle understands the realities of the road and how to respond to them — whether it’s a stop sign, a ball in the street, or another vehicle — is through deep-learning algorithms. For example, Audi and other auto makers use deep learning algorithms in their camera-based technology to recognize traffic signals by their characters and shapes.

Financial Services — Failing to properly identify and prevent fraud is an expensive proposition that costs the financial industry billions of dollars per year. As advancements in computing technologies and the expanding use of e-commerce platforms dramatically increase the risk of fraud for these financial companies, many are turning to deep learning to uncover transactional anomalies or suspicious patterns.

Manufacturing — Deep learning’s capacity to analyze very large amounts of high-dimensional data can take existing preventive maintenance systems to a new level.

Media & Journalism — With deep learning, machines can learn the punctuation, proper grammar, and style of a piece of text and can use the model it develops to automatically create entirely new text with the proper spelling, grammar, and style. Everything from Shakespeare to Wikipedia entries have been created. Deep learning is also being used to identify text that may incite terrorism or hate or text which constitutes deceit or misinformation.

Medical & Pharmaceutical — From disease and tumor diagnoses to personalized medicines created specifically for an individual’s genome, deep learning in the medical field currently has the attention of many of the largest pharmaceutical and medical companies.

Retail & Entertainment — Ever wonder how Netflix comes up with suggestions for what you should watch next? Or how Amazon comes up with ideas for what you might want to buy next, and those suggestions are exactly what you need but just never knew before? Those recommendation systems are, once again, deep-learning algorithms at work.

In this way, deep learning can be a valuable tool for customer service management and personalization challenges. For example, deep-learning analysis of audio allows systems to assess a customer’s emotional tone; and, in the event a customer is responding badly to the system, the call can be rerouted automatically to human operators and managers.

Robotics — Deep-learning applications for robots are plentiful. They can teach a robot to do something (like housekeeping, greeting guests, stocking shelves, and picking apples) by observing a human completing a specific task. Drones flying down rows of crops can learn to identify the weeds, and to use weed-killer only where necessary.

Conducting Litigation — More to the point, when you drag and drop in a complaint of a specific type, they’ll write an answer and spit out a first set of discovery requests, too.

Preventing Litigation — No one looks at yesterday’s emails to see if a “smoking gun” for a specific litigation risk is present. As we’ve said, there’s too much data. But some contend that deep learning may enable us to do what currently cannot be done.

With a focus on the business-relevant and high-frequency classifications of litigation, a system with a set of pre-trained deep learning models may ingest yesterday’s batch of emails and process the emails through each model to see which of the emails pattern-match to a litigation risk. Only the emails “related” to a risk category would be returned for review by members of the in-house team. By using deep learning as a filter, the in-house staff may spot a risky email in time to nip the risk in the bud.

To Conclude

What’s important to remember is that deep learning is in its emerging stages but is developing rapidly. General business managers and lawyers of the future will need to understand when problems are likely to be susceptible to a deep-learning approach, and how to manage the diverse teams with more technical skills that will be needed to be brought together to solve them.

The article was co-authored by Nelson E. (Nick) Brestoff, who holds engineering degrees from UCLA and Caltech and a law degree at USC. He was a California litigator for 38 years and is an inventor on eight patents that use deep-learning to identify and provide an early warning of either risks to avoid or financial advantages to obtain. He is also author of AI Concepts for Business Applications, a book to be published later this year.