Beyond the Hype: Putting AI to Work at Liberty Mutual #ILTASS18 #ILTACON

Session Description: Neota Logic and Liberty Mutual will share details of the design and implementation of Neota Logic’s AI-driven expert system platform to automate document creation and drive internal efficiencies at Liberty Mutual.

Speakers:

[These are my notes from the International Legal Technology Association’s 2017 Conference. I’m publishing them as soon as possible after the end of a session, so they may contain the occasional typographical or grammatical error. Please excuse those. To the extent I’ve made any editorial comments, I’ve shown those in brackets.]

NOTES:

  • It’s just software.  According to Kreutzer, “It’s only AI when you don’t know how it works; once it works, it’s just software.”
  • Why did Liberty Mutual even consider using AI?
    • As a life long geek, Jeff thought that AI was a cool thing to explore. The purpose of this exploration was to try to reduce some of the cognitive load, some of the lower-value tasks that slow the legal department’s lawyers down. Ultimately, it was intended to allow the lawyers to focus on higher-value tasks.
  • Forget the pixie dust. With all the hype about AI, it is easy to be dazzled by the “pixie dust” aspects of the technology. However, in Marple’s view, 90% of the work in an AI deployment project involves capturing, organizing and cleaning the relevant data. Without this essential work, you cannot get to the pixie dust.
  • Start with small, low-risk projects. To prove the value of the technology, they started with their internal non-disclosure agreement (NDA) process. (As with many companies, the NDA process was more onerous and time-intensive than the corporate legal department — and their internal clients — would like.)
  • Their timeline to deployment. The bulk of the time was spent exploring the technology options and identifying the right use case for their pilot. The actual process of creating and deploying their instance of Neota Logic took two to three weeks.
  • Their new toolbox. They are creating a toolbox that includes Neota Logic to be deployed throughout the organization. Because Neota is a business-facing tool, the business folks can use Neota to improve processes without involving a single developer.
  • What skills and competencies best support AI deployments?
    • Legal Engineers are perfectly placed to support AI deployments. They can be lawyers or business analysts. However, they need an understanding of the law and they must be willing to “lean in to technology.” Plus, they must have a healthy curiosity about the technology.
    • Some law schools are training their students in legal technology by using Neota Logic in their courses. Faster than we know it, most law schools will be training students in legal technology. And, when they enter law firms, they will simply practice law in a tech-enabled way.
Share

Keynote: Transforming the Business of Law with Cognitive Computing – Watson Legal #ILTAKEY2 #ILTACON

Session Description: “Artificial intelligence” (AI) is arguably an overused label for a vast array of technologies that vary in features, complexity and benefits. What is AI, and how do you know if you have a problem it can solve? Brian Kuhn (founder of IBM Watson Legal) and Shawnna Hoffman ( IBM Global Co-Leader, Global CoC) will address key misconceptions around artificial intelligence and cognitive computing and share insights from business-of-law use case workshops IBM conducted with corporate legal departments and law firms over the last two years. Brian and Shawnna will identify patterns of customer interest and success related to the application of artificial intelligence and cognitive computing to the legal domain, and he’ll discuss the top facts everyone needs to know about what AI can — and can’t — do.

Speakers:

  • Brian Kuhn, Esq., Global Leader and Co-Creator of IBM Watson Legal
  • Shawnna Hoffman, Global Cognitive Co-Leader, Global CoC

[These are my notes from the International Legal Technology Association’s 2017 Conference. I’m publishing them as soon as possible after the end of a session, so they may contain the occasional typographical or grammatical error. Please excuse those. To the extent I’ve made any editorial comments, I’ve shown those in brackets.]

NOTES:

  • What is AI.  A type of tool to extend our cognitive abilities. While computers have extended our cognitive abilities, 88% of the data we have created is “invisible” to us. It is rendered invisible because it is unstructured — narrative creative for consumption by humans rather than computers. The language is context heavy and may be rife with ambiguity. Historically, computers have been unable to parse unstructured data because they did not understand human language.
  • AI is an umbrella term. There is no single agreed definition. It is a constellation of capabilities (e.g., machine learning, neural networks, etc.)
  • Cognitive Legal. All cognitive systems understand human language, they can reason to extact ideas, they can learn from past results, and then they can interact in a natural way. They can read 800 million pages per second and, in the background, derive meaning and form hypotheses based on this work.
  • Cognitive systems democratize knowledge. They have the capacity to read and understand more content than humans could do in a lifetime. And the cognitive systems can do this in a matter of days or weeks. Better still, while human intelligence is focused on discrete bodies of knowledge that leads to deep expertise, cognitive systems can consume and comprehend multiple domains of knowledge and derive new insights that humans could never reach by themselves. This ability to synthesize information from across domains of knowledge will profoundly alter professions.
  • How do you know you have a problem AI can solve? IBM’s Cognitive Legal practice focuses on the business of law rather than the practice of law. This was the choice of their clients. They have been identifying use cases that:
    • provide tangible business value
    • identifiable key performance indicators
  • Identifying good use cases.
    • Wha are the painpoints?
    • What is hte business value of this use case?
    • Has the content that will fuel the solution been identified?
    • The best use cases provide
      • organizational beef
      • end-user benefit
      • strategic alignment
      • speed to implemention
  • Examples of Use Cases.
    • Outside Counsel Insight — designed to reduce spend on outside counsel by
      • automating manual invoice review processe to detect billing anomalies,
      • producing insights into trends and patterns that demonstrate the quality of outside attorney performance, resulting in outcome and price consistency,
      • producing brenchmarking insights at the case and line-item level that establish a reasonable level of effort for repeatable legal tasks, which then helps justify fixed-fee pricing arrangements versus hourly billing.
    • Early Case Insights
      • Consuming historical work product by outside counsel to provide appropriate advice for repeatable work and new insights to address current needs — without reinventing the wheel or pay again for the work to be done
    • Expertise Finder
      • using internal and external data
      • it could be used across a network of organizations or law firms. The underlying data would reside in public filings or even in a private blockchain.
  • Cognitive Legal “Cartridge”. We will be training “cartridges” in what and how the people of your firm think. Others could then purchase these cartridges and point them at their own data sources so that they can derive new insights by using what and how the people who trained the cartridges think. IBM Watson has done already by having three oncologists at Sloan-Kettering train a cartridge, and then applying that cartridge to the medical data collected by the largest hospital in Thailand.
  • Closing Insights
    • “Cognitive is a mirror you hold up to yourself.”
    • Use AI internally first. The power of AI lies in turning it onto your proprietary data. The richest data source is not located on the internet. In fact, 80% of data resides inside organizations and behind firewalls.
    • “We think about Watson in the context of AI — not artificial intelligence but augmented intelligence.”
    • Access to justice — 80% of people involved in domestic violence disputes represent themselves pro se in court. If there were a cartridge trained by domestic violence legal experts, that cartridge could help make sense of the data for each of those cases.
    • Legal Blockchain +AI:
      • blockchain contains enormous stores of data
      • AI can parse that data and make sense of it
      • blockchain and AI could be combined to create trusted and transparent self-executing contracts
Share

Have You Had Your Turing Moment?

2016 is the year I had my Turing moment.

As Wikipedia tells us, Alan Turing proposed in 1950 a test of “a machine’s ability to exhibit intelligent behavior equivalent to, or indistinguishable from, that of a human.” The proposal was that a human and machine would interact via a text-only channel. If a human evaluator observing the interaction could not distinguish the machine participant in the conversation from the human participant, then the machine would pass the “Turing Test.”

This summer I was introduced to a potential business collaborator. We had a brief exchange by email and then decided to schedule a meeting. He said that his assistant would contact me to find a mutually convenient time. She did and we did. Several times over the course of our project, my collaborator’s ever-helpful assistant took care of the scheduling hassles. In fact, I came to value her help so much that I spontaneously wished her a wonderful weekend one Friday. And she kindly returned the good wishes.

Because the readers of this blog are smarter than the average bear, you know where I am going with this. That fabulous assistant was not human. Rather, she was a scheduling bot created by Clara Labs. And in the course of our admittedly brief exchanges, I did not realize I was conversing with a machine. She passed the Turing Test with flying colors.

Why am I telling you this story? Because scheduling bots are only the beginning. Frog Design has just published its annual Tech Trends provocation that provides a glimpse of what is in store for us in 2017.  According to Frog Design, the report identifies “15 technology trends that will unlock opportunities for growth and enable organizations to provide more meaningful experiences to their customers, employees, and society.”

One of the trends on this list is the proliferation of business bots that can do far more than mere scheduling. Here is the scenario sketched by Frog Design’s Toshi Mogi:

Imagine an entrepreneur whose mentor has recommended they start a new venture, selling vintage electric skateboards to the aging hipster market. The entrepreneur will commission an assortment of business bots to bring their vision to reality. The R&D bot will crowdsource the selection of designs from on-demand freelance designers, the Operations bot will manage contract manufacturers and production schedules, and the Sales and Marketing bot will optimize e-commerce channels and product promotions. As business bots become more intelligent, their ability to perform complex operational tasks and harness digitally enabled platform services will help new entrepreneurs scale their ventures, faster and with precision.

Have you had your Turing moment yet? Don’t worry if you haven’t. If Frog Design is right, business bots are coming to you soon and will massively improve your productivity.

Alan Turing would be pleased.

[Photo Credit: Wikimedia]

Share

Law Firms — The Final Frontier

leonard_nimoy_william_shatner_star_trek_1968We interrupt our ILTACON blogging to mark a significant occasion in the history of humankind: The 50th anniversary of Star Trek.

This iconic television show premiered on September 8, 1966, and since then has changed the way several generations think about technology and space. In fact, it has changed our lives. According to Elizabeth Howell, several Star Trek technologies that seemed far-fetched when first seen on TV are now an indispensable part of modern life:

  • communicators >> cellphones
  • tricorders >> MRIs
  • tablets
  • global positioning systems

While I am an enthusiastic fan of the franchise, I cannot claim to match Marshall Honorof in his devotion. He has watched every single Star Trek movie and television show. In his account of his personal space odyssey, he makes the following observation about Star Trek:

Something about the show’s optimistic message stuck with me. Technology can improve our lives. We can conquer our deeply held prejudices. There is other life out there, and it is willing to cooperate with us. And, importantly, no matter how far we come as a society, there will always be room for adventurers.

[…]

While it’s not a novel observation, the reason ‘Star Trek’ feels unique, even in a world of more ambitious sci-fi properties like ‘Battlestar Galactica’ and ‘Black Mirror,’ is because it alone asserts that technology will make our lives better, not worse.

So it seems that others have figured out something that eludes too many inhabitants of law firms: technology can make our lives better, not worse. The challenge is whether we can “conquer our deeply held prejudices” and use our Vulcan, Bajoran, and Betazoid abilities to leverage technology to make something better of a woefully underperforming industry. We have underserved clients, unhappy lawyers, and unappreciated allied professionals. Isn’t it time for a change?

Go boldly!

[To learn more, see the great infographic on The Evolution of Star Trek.]

[Photo Credit: Wikipedia]

Share

No Free Tech Ride

I_VotedFor the last few decades, we have been like children let loose in a candy store, never stopping to consider the negative consequences of stuffing our faces with sweets.

What is this candy store? The Internet.

For the last few decades, we have been on a deliriously fun tech ride. We have come to take for granted the ease, the convenience, and the sheer joy of immediate gratification the Internet provides. All of this and more can be yours with just a little connectivity.

Of course, we know that if you consume too much sugar you destroy your teeth and set yourself up for diabetes and a host of other bad health outcomes. And, as we have come to learn, all that consumer tech convenience sometimes come with a very large price tag: hacker convenience.

That is the reality. There is no such thing as a consequence-free candy binge. And, there is no such thing as a free ride on technology.

I was reminded of these realities when I saw Bruce Schneier’s recent article in The Washington Post with the scary title “By November, Russian hackers could target voting machines.” According to Schneier,

Over the years, more and more states have moved to electronic voting machines and have flirted with Internet voting. These systems are insecure and vulnerable to attack.

But while computer security experts like me have sounded the alarm for many years, states have largely ignored the threat, and the machine manufacturers have thrown up enough obfuscating babble that election officials are largely mollified.

For years cyber security experts have warned about “walking talking security threats.” By this they mean you and me — the people who use ridiculously easy-to-guess passwords or do not lock our devices or refuse two-factor authentication or surf the web merrily over insecure public wifi. These reckless behaviors can lead to a host of terrible outcomes, including identity theft or breaches of an employer’s firewall.

If this were not bad enough, Schneier suggests that people with the ability to harm society have indulged in even more reckless behavior.  Schneier is talking about a whole other level of tech-ignorant behavior that has resulted in an insecure electronic voting system could imperil our democracy. If he is right, we should be angry and terrified.

At every level of society, starting with you and me, we need to take a serious look at how our negligent use of technology can have disastrous consequences. The party has been fun, but now it is long past time we grew up. After all, we now know there is no such thing as a free tech ride.

[Photo Credit: Denise Cross]

Share

Working Backwards to the Technology #ArkKM

Session Title and Description:  Working Backwards to the Technology:

Focusing on User Experience to Enhance the Practice of Law Technology is an important—indeed critical—enabler for knowledge management, but allowing the technology tail to wag the KM dog can lead to serious problems. Focusing first on user experience is difficult, time-consuming, and expensive; but doing so will help ensure success.

Speaker:

Patrick V. DiDomenico, Director of Knowledge Management, Ogletree, Deakins, Nash, Smoak & Stewart, P.C.

[These are my notes from the 2015 Ark Group Conference: Knowledge Management in the Legal Profession.  Since I’m publishing them as soon as possible after the end of a session, they may contain the occasional typographical or grammatical error.  Please excuse those. To the extent I’ve made any editorial comments, I’ve shown those in brackets.]

NOTES:

  • Steve Jobs: “…you’ve got to start with the customer experience and work backwards to the technology. You can’t start with the technology and try to figure out where you’re going to try to sell it.”
  • Lessons:
    • Technology is an important enabler for KM. So don’t be a Luddite.
    • Good user experience is time-consuming and expensive to create, but worth it.
    • If the user experience is too demanding, the user will become frustrated and depleted. It’s a lot like decision fatigue. (See the Isreali Parole Board study.)
    • If the technology provides bad or incorrect information, the customer experience will be suboptimal.
    • The best technology disappears and the best user interface disappears. They leave a great user experience.
    • Don’t settle for bad user experience. Question things.
Share

Your New Robot Colleague

robot-1241845-1600x1600Inside the legal industry we often complain about the glacial pace of change. However, sometimes change can sneak up on you unaware — even in the legal industry.

The topic of artificial intelligence (“AI”) in the practice of law is one many of us had considered long before the advent of IBM’s Watson. However, while we could imagine the usefulness of AI in a legal practice, we were not sure where, how or when we might see it in reality.

Well we now have an answer to the question of when. A September 22 press release made this announcement:

RAVN Systems, experts in Enterprise Search, Unstructured Big Data Analytics, Artificial Intelligence (AI) and Knowledge Management solutions, announced today that international law firm, Berwin Leighton Paisner (BLP), is implementing the company’s Artificial Intelligence platform, known as the RAVN Applied Cognitive Engine (RAVN ACE).

But that’s not all. Not only do we have AI in action, but it is now in the form of a new member of the team: a contract robot. This contract robot works in BLP’s real estate practice. Its job is “to extract data from standard legal documents, cross-check the data internally, check it against external sources, and write the output into a spreadsheet, ready for the next stage of the process.” Sounds like a terrific junior associate.

Here’s how Matthew Whalley, Head of Legal Risk Consultancy at BLP, describes that new team member:

The robot has fast become a key member of the team. It delivers perfect results every time we use it. Team morale and productivity has benefited hugely, and I expect us to create a cadre of contract robots throughout the firm. If the reaction to our first application is any indication, we will be leading the implementation of AI in the Law for some time to come.

So here we are. The robot is real. The next question is: how will your life change with the arrival of your new robot colleague? If you believe there will be no change at all, watch out. Your new colleague is about to prove you wrong.

[Photo Credit: Cecile Graat]

 

Share

Revolutionary Integrations — #ILTACON #ILTA116

ILTACON 2015 LogoSession Summary: Attorneys need information about their matters from a variety of sources, and the days of having to jump from one tool or system to the next are over! See how firms are enabling collaboration, matter management and project management by strategically fitting together technologies to create a single platform where attorneys can create, collaborate, share and retrieve knowledge. They are simplifying the way attorneys access and interact with dozens of different technologies and creating next-generation systems designed to support and streamline attorney workflows. See firsthand how they are making it happen!

Speakers:

  • Meredith Williams, Baker Donelson
  • Jeffrey Rovner, O’Melveny & Myers
  • Ginevra Saylor, Dentons (moderator)

[These are my notes from the International Legal Technology Association’s 2015 Conference. Since I’m publishing them as soon as possible after the end of a session, they may contain the occasional typographical or grammatical error. Please excuse those. To the extent I’ve made any editorial comments, I’ve shown those in brackets.]

NOTES:

  • Audience Overview: There were about 90 attendees. When asked by the presenters, only about 5 attendees indicated that their firms had active matter pages.
  • O’Melveny’s Matter Pages. The firm introduced the concept of matter pages five years ago. They remembered the ease of having all matter materials within a single redweld. With digitization, however, the various materials related to a matter were scattered as far as the attorney was concerned: documents were in the document management system, correspondence in email inboxes or archive folders, financial information was in the time/billing system, etc.
    • The initial concept:
      • matter updates: news posted here for the benefit of the entire team, and also emailed to members of the team
      • financial information: amounts accrued/billed/realized, leverage, etc.
      • list of timekeepers
      • links to matter documents and practice support materials
      • ethical screen information
      • real-time information
      • interactive elements
    • The current approach: In addition to the original materials they have added
      • budgeting tools, including tools for alternative fee arrangements
      • key financial indicators (KPIs)
      • modules to support legal project management
    • The matter pages are a front-end to a wide range of data sitting in the data warehouse (in SQL tables in the original systems of record). They use stored procedures to avoid doing complex things on the fly.
    • They use Recommind to retrieve content from the document management system.
    • The visibility of matter pages is controlled by ethical screens and, in the absence of a mandatory screen, access can be limited to a defined group.
      • The matter pages are composed of modules. These modules have granular security so that the firm can restrict access to specific modules or to specific content within modules.
  • Baker Donelson’s Electronic Matter File.
    • “If you force them they will come.” They achieved this by consolidating all the relevant data into a single interface
    • Client/Matter Dashboards. These dashboards are created automatically in SharePoint 2010 as soon as a new matter is opened. The dashboards are designed for information consumption rather than collaboration.
      • They have almost 4000 dashboards.
      • The dashboards include basic information on how the client wants to be contacted.
      • They use Recommind to push the information into the dashboards.
    • Client Dashboards:
      • client profile details
      • documents
      • Interaction contact & event details
    • Matter dashboards:
      • critical content: financial data on the matter
      • matter budget
      • documents
      • correspondence
    • Extranets
      • Extranets enable collaboration by providing the ability to
        • see Information about the File
        • Manage the Client or File
        • Work the File more efficiently
      • Designed with mobility in mind
      • Client-facing extranets:
        • SharePoint team calendars — organized by matter
        • case assignment information — which Baker Donelson personnel are managing specific client matters
        • quarterly reports generated by Contract Express
        • wherever possible, they generate documents for each matter via Contract Express (document assembly)
        • discovery banks of related content
    • Next phase = BAKERPRACTICE
      • the KM team observed several lawyers as they worked — this revealed all the hassles of “dancing among the systems” in order to “work the file.”
        • behind this new effort is two years of due diligence plus four years spent clarifying their universe of matter types for the firm
      • they will have to create a new interface that allows lawyers to work a matter from a single place
        • a lawyer will see a list of files
        • then the lawyer the lawyer can drill down to the task that lawyer needs to accomplish
        • when the lawyer closes a document, the system will show the lawyer how time that lawyer spent drafting, show the likely client-matter number, and then ask the lawyer if she would like to report that time now.
        • when the lawyer closes an email, they will receive a similar billing prompt
      • they have retained an external UI/UX firm to make sure they get the user-facing elements right
      • they will be choosing participating vendors shortly
      • they estimate that BAKERPRACTICE will result in significantly more accurate time reporting (and billing)
  • Start with Why
    • Bring meaning to information
    • Matter management – matter centricity alone is not enough
    • Enhance collaboration
    • Simon Sinek:  “People don’t buy what you do, they buy why you do it.”
  • Lessons Learned.
    • Do not take the lawyer outside their process.  Learn their process and then build to that.
Share

No More Pretzels!

Let me begin with a health warning: Be careful as you watch the video below. It will give you a sympathy backache. That said, it’s worth watching it to see what a naturally gifted human pretzel can do.

Now think about how you might perform as a human pretzel.  No matter how much physical flexibility you may have, chances are that you cannot come close to the standards that Victoria Jacoby attains in the video.

Actually, let me rephrase that. Chances are that you cannot come close to her ability to contort her body, but I’m willing to bet that you far exceed her accomplishments when it comes to contorting yourself and your technology to accomplish everything you need to do everyday.

A classic case in point is email. Its ubiquity is a testament to its perceived usefulness. However, I’d suggest that we have been pushing its usefulness beyond the boundaries of safety and sanity.

So what are smart and safe uses of email? Craig Jarrow of Time Management Ninja suggests the following:

  1. Non-urgent communication
  2. Follow-up
  3. Praise
  4. Timeshifting
  5. Filtering
  6. One-t0-many communications
  7. Sending documents/pictures
  8. Mobility

If those are the good uses, what are the bad uses? In 2007 Dave Pollard outlined the bad use cases in When NOT to Use Email:

  1. To communicate bad news, complaints or criticism
  2.  When you are seeking information that is not simple and straight-forward
  3. When you are seeking approval on something that is involved or controversial
  4. When you are sending a few people complicated instructions
  5. When you are asking for comments on a long document
  6. To request information from a group on a recurring basis
  7. To convey instructions to a large number of people
  8. To achieve consensus
  9. To explore a subject or idea
  10. To send news, interesting documents, links, policies, directory updates and other “FYI” stuff.

For each of these cases, Dave Pollard provides what he considers to be the better way of communicating. (You can find a concise summary of the alternatives in his post, Getting Rid of Email.) In addition, he has created a detailed decision tree you can use to determine what mode of communication is best in each circumstance.

People are fond of saying that “Lawyers live in email.” A more accurate way of describing this is as follows: lawyers spend their days as human pretzels when it comes to email. They contort themselves and their technology, pushing it to do things it was never meant to do.

And then we get mad when things go wrong?

Perhaps it’s time we shifted from the bad use cases to the better use cases for email. Perhaps it’s time we finally outlawed pretzels — of the human and technological kind.

Share

Working with Special Snowflakes

snowflake-310071_1280We were taught as children that no two snowflakes are alike.  Some of our teachers went so far as to suggest that we were like snowflakes, each possessing unique characteristics, each to be valued in her own right.

Isn’t that lovely.

This kind of affirmation is helpful in the right time and place, but at some point in our education most of us learn that there are a lot of characteristics that humans share. Moving beyond crude stereotypes we discover, for example, a shared fight-or-flight response and nesting impulse. And that’s just the beginning.

There are, however, some people who are rarely pushed to look beyond their particular circumstances to understand how much they share with the rest of humanity. Who am I talking about? Lawyers.  Yes, the work they do is different from the run of the mill. Yes, they do need a special education to undertake this work. Yes, their work can have enormous consequences for others. But the same could be said for doctors, engineers, architects, etc. Yet lawyers persist in believing that they are a breed apart, a group of special snowflakes.

Unfortunately, too many technologists enable this point of view by telling lawyers that tools can and should be adapted to accommodate lawyer preferences. Thus you have technologists larding up standard software such as MS Office with customizations and embellishments meant to placate the special snowflakes in our firms. And then we act surprised when we calculate the cost of implementing new technology or upgrading existing technology. At what point do we say that the system performs reasonably for 80% of the work lawyers do and we should think twice (or thrice) about customizing for the remaining 20%?

As you consider the decisions you make about your law firm technology or knowledge management systems, consider the extent to which you are enabling special snowflake syndrome. Codependency is unhealthy for all involved.

Share