It’s still early days in the GenAI revolution. And there is as much trepidation and confusion as enthusiasm about what tomorrow brings.
Will generative AI wipe out half the legal profession? Could it kill the billable hour? Might it make lawyers happier?
These sentiments were on display at Everlaw Summit ‘23 as speakers and guests dove into how the newest LLMs work, ways to introduce AI (responsibly) in the workplace, and the potential risks and rewards for the legal industry. Audiences also heard the latest on AI-influenced training for young lawyers, ways to use the same billable time to make clients happier, and staying relevant in careers against increasingly competent robots.
Over the course of the three-day conference, plenty of positive themes shined through. Like this one, summarized by Ironclad Chief Community Officer Mary O’Carroll:
“GenAI isn’t about eliminating people,” O'Carroll said. “It’s about empowering people.”
GenAI Will Take a Diverse Village to Deploy
Given its potential to transform legal practice and free attorneys to focus on higher-level investigation and litigation strategies, GenAI is not just another piece of technology. Powerful, broadly accessible, but also layered with potential risk, it calls for a multidimensional approach.
Tech-forward law firm Seyfarth Shaw has had plenty of change management experience with technology automation, from machine learning to robotic processes. “But nothing quite prepared us for the tsunami that was Chat GPT,” said Stephen Poor, Seyfarth chairman emeritus and labor and employment partner, during a panel on the responsible ways to deploy Generative AI.
To evaluate the potential impacts and uses of GenAI, Seyfarth panned out to bring together diverse perspectives from across the firm. “Rather than looking at it as just the adoption of a tool, we wanted to understand the capabilities of the technology and what it meant for the organization, both short term as well as the medium term,” Poor said.
The single largest piece, he said, is the people – how are they going to be affected by the technology and what are their concerns? How can they get excited about it? That type of consideration informed an approach that was neither top-down nor completely decentralized, but a hybrid of the two. Seyfarth recruited a diverse pool of those who were excited about the new tech, those who wanted nothing to do with it, and “nontraditional thinkers” from the ranks of associates, partners, and other firm professionals.
“This is a new world,” Poor said. “We want people involved in the thinking who had not typically raised their hands before.”
LLMs Are Transforming Law Practice and Training Next-Gen Lawyers
In a panel about large language models in action, IBM Vice President and Associate General Counsel Donna Haddad echoed the idea that GenAI touches the entire organization and must be dealt with accordingly.
Haddad, a founding board member of IBM's AI Ethics Board, stressed that with employees in all corners of a company experimenting with the technology, it’s important to understand how it’s being used, what data goes into the system, and how it’s being secured to abide by client confidentiality and other professional ethics obligations. One of the keys to success, Haddad said, is to build a culture of autonomy and responsibility throughout the organization, so that everyone is issue-spotting, evaluating risks, and escalating concerns.
Speaking on the same panel, Dr. Megan Ma of Stanford’s Center for Legal Informatics (CodeX) described what she’s most excited about. She sees potential for large language models to enhance training for the next generation of lawyers. She’s exploring ways that LLMs can help map out litigation strategy or contract negotiation and building simulation tools.
“If you're an early-stage lawyer or a junior associate or if you're fresh out of law school – or even in law school – can you give yourself the opportunity to live these experiences rather than learning from the books only?” Dr. Ma said.
Taking the Mystery Out of GenAI
In his Summit keynote, Everlaw Founder and CEO AJ Shankar helped break down how GenAI works and spelled out some of the most advantageous applications for the legal profession. He assured audiences that ChatGPT is not (yet) intelligent like a human – it doesn’t actually understand the information it presents, and it still makes mistakes. But, as he explained, its computing power and extensive training make GenAI tools very competent at certain tasks that can really transform legal work.
Customers in the EverlawAI Assistant beta program are seeing that play out firsthand. Lewis Brisbois was able to reduce the time to complete core tasks – such as summarizing key documents and preparing for depositions and mediations – by at least half.
“Instead of an attorney taking 15 minutes to read and summarize the contents of a 50-page Word document in Notes or scour a 100-slide PowerPoint to identify the content needed to build your case,” said Gordon Calhoun, a partner at Lewis Brisbois and a leader in the ediscovery space. “EverlawAI Assistant produces competent summaries that one would expect from an experienced attorney in less than 30 seconds.”
That means more time for lawyers to devote to higher-value goals, such as stronger preparation for depositions and dispositive motions and crafting winning case strategies.
Billable Hour Not Going Away, But Clients May Be Happier
Fears about GenAI doing away with the billable hour are likely overblown. That’s what Jay Carle, co-chair of ediscovery and information governance at Seyfarth, told audiences during a panel discussing highlights from the Everlaw 2023 Ediscovery Innovation Report. Carle is enthusiastic about the possibilities to deliver greater value to clients with the same billable hours. In fact, he worries about lawyer over-reliance on AI.
He said he believes these new tools can be used responsibly and defensibly to help legal professionals work at a higher level, by for example, finding key information during ediscovery more quickly and kickstarting projects like a statement of facts or a pleading.
“I work much better when I've got a starting point rather than having to start from a blank page, it's much easier to edit,” he said. “You end up with a better work product and sometimes it'll come back with a new way of saying something or presenting information in a way that you wouldn't have done that maybe is a little bit better."
At the same time, he said relying on AI-generated document summaries alone could be problematic. Are you at risk if you rely solely on GenAI outputs to make a legal judgment?
“It's still going to be important for the lawyer to go and read those documents or those cases,” Carle said. “Because, what if it hallucinated? What if it missed the point?”
Future Proofing Your Career in the Age of AI
If you’re getting pangs of fear that Chat GPT will make you obsolete, you’re not alone. This one is hitting knowledge workers across industries. At Summit, The New York Times tech columnist Kevin Roose offered survival strategies drawing on his best-selling book, Futureproof: 9 Rules for Humans in the Age of Automation.
The advice for years had been to develop hard skills, like STEM and coding, to stay employable. Roose said that’s no longer enough.
“You cannot outhustle an AI. They don’t need vacation. They don’t get sick. They don’t sleep,” Roose said. “You are not going to compete in this economy based purely on the amount of cognitive labor that you are willing to put out.”
Instead, workers need to double down where the machines are weakest. What will keep lawyers in the game is the human part of their job – the face-to-face interactions, the client work, the situations that can’t be solved through a boilerplate approach.
Take, for example, Roose’s accountant, who asked himself how he could turn tax prep, a chore, into a fun experience. As a former stand-up comedian, he decided to combine taxes with humor. Today, Roose says his accountant successfully handles taxes for a niche industry of actors, writers, and artists and hires other former comics as accountants.
“I have more fun doing my taxes than going to actual comedy shows that I paid money to see,” Roose said.
Succeeding by Making the Power of AI Your Own
As Everlaw CEO AJ Shankar said, large language models may have developed an intuition about human knowledge and expression to the point that machines can produce meaningful outputs, but they do so without any deeper understanding. Human creativity and judgment are still very much needed. AI, he says, is like the intern who can help legal professionals do faster, higher quality work – and hopefully increase professional satisfaction and quality of life in the process.
“So keep that in mind,” AJ said. “You have the world's most comprehensive intuition at your fingertips.”
What will you do with it?