Will Gen AI Dumb Us Down or Smarten Us Up?
January 04, 2024DownloadsDownload Article
Most people have ambivalent feelings about generative artificial intelligence (“Gen AI”). A recent survey from Pew Research Center indicated that 52% of respondents were “more concerned than excited” about the use of Gen AI in daily life; just 10% were “more excited than concerned” and 36% were “equally excited and concerned.”1 This marks a notable uptick in apprehension from Pew surveys in 2021 and 2022, when only 38% were “more concerned than excited.”2 Such misgivings are understandable as applications and potential uses of Gen AI come into clearer view.
Despite the media’s many Terminator allusions to “the machines taking over,” most respondents primary concerns about the use of Gen AI are more mundane: They are likely to see Gen AI as a potential threat to their jobs and livelihoods. While IT-driven advances in recent decades mostly have impacted lower-skilled jobs, Gen AI is widely seen to be more disruptive to white collar jobs held by more educated workers. A previous Pew Survey from July 2023 indicated that workers with at least a bachelor’s degree were more than twice as likely to have jobs with the most exposure to AI than those with just a high school diploma (27% vs. 12%), while 19% of all workers were in jobs most exposed to the impact of AI — meaning that their important work activities will be either assisted by or replaced by AI.3 Understandably, the distinction between “assisted by” and “replaced by” is an enormous one, and as there is little evidence to date about exactly how this will play out by profession, there is some reasonable basis for respondents’ concerns.
The Pew report concluded that jobs where analytical skills are more important had more exposure to AI impacts. Specifically, Pew’s analysis concluded that just over one-half (52%) of workers in professional, scientific and technical services jobs and more than one-third (37%) of workers in finance, insurance and real estate jobs were most exposed to AI.4 Interestingly, slightly more survey respondents in the banking & finance sector said the use of AI will “help more than hurt” in their jobs than those who felt otherwise.
Gen AI and Restructuring: Just Another Brick in the Wall?
For the restructuring profession, the prospect of Gen AI infiltrating our realm is another step on a long road that has changed the way we do our jobs for more than two decades — all for the better, so far. Let’s retrace some of that terrain.
Changes in the way we work since the turn of the century broadly can be placed into one of these categories:
- Efficiency Gains: Changes that reduce the time needed for professionals to complete a task.
- Productivity Gains: Changes that allow professionals to better perform a task, doing it more comprehensively or on a larger scale than their predecessors could.
- Outsourcing Gains: Changes that essentially relieve professionals of performing a task due to easily accessible third-party resources or tools that can do it.
Those in our profession long enough can recall the frequent drudgery of office life for junior professionals before PACER’s online access to filings and documents on U.S. bankruptcy court dockets became available in 2001. Prior to PACER, requests for filing documents had to be made through third-party services, such as Washington Document Services, that retrieved them from courts, photocopied them (for $0.08 per page!) and shipped them overnight or second-day. As the workday started, it was a common sight to see large brown UPS boxes of documents sitting on consultants’ desks, which would dutifully be photocopied and circulated to the case team. It was the Stone Age for document retrieval and management — a critical part of any restructuring case. That dreaded routine ended once documents on PACER court dockets became retrievable online. But even then, soft copies usually were scanned documents and not yet word-searchable, so zeroing-in on relevant sections or specific word mentions buried within documents hundreds of pages long, such as a DIP financing proposal, was still an onerous and time-consuming process. It was a few more years before the PDF file format was made openly available and became an industry standard for electronic document sharing.
These tasks can now be done in minutes, marked up and summarized. Such a huge change in this workflow process has made us more efficient workers, enabling fast and precise searches of court dockets and a vast ocean of filed documents — an indisputably positive work development for younger professionals typically charged with such assignments. PACER didn’t change the essential nature of the task to be done (e.g., analyzing a proposed disclosure statement/POR), but it greatly reduced wasted non-learning time for professionals, allowing the job at hand to be done much more efficiently. Today, other advances in search can return relevant documents from myriad other sources, such as SEC filings, and prominently highlight all mentions of key words or phrases. We can search through years of SEC filings on EDGAR in a matter of seconds and find all documents that mention the term “asbestos litigation,” for instance. That is power at our fingertips.
There have also been technology-driven changes that have made us more productive workers, mainly from continuous improvements in standard-issue business software that make it possible to process and analyze huge amounts of data on our laptops and present these findings in a compelling manner. The functional capabilities of the Microsoft Office suite are more advanced today than they were one or two decades ago. These changes represent improved business tools that have ramped up the productivity of restructuring professionals in an increasingly knowledge-based workplace. Restructuring professionals with deep knowledge of powerful business software tools are in high demand and are more productive workers than their predecessors were years earlier.
Another workplace change over the last dozen years or so is the proliferation of publications dedicated to restructuring activity and leveraged credit markets, mainly Debtwire and Reorg Research (Reorg) but several others as well. These services began with dedicated news coverage of restructuring events covered by beat reporters but have since expanded to include analysis of critical events and key documents by their teams of lawyers and analysts, such as evaluating the impact of a proposed restructuring transaction on creditor groups or summarizing and evaluating key negative covenants contained in a credit agreement.
These services used by restructuring practices have greatly reduced data and intelligence asymmetry in the marketplace, since most practitioners subscribe to them and consume their content. Fewer events and developments fly under the radar for long. For larger restructuring practices, it is harder for data-driven prospecting efforts to identify potential business opportunities that aren’t already known by competitors.
Recent offerings by these third-party services effectively represent an outsourcing of certain work by restructuring shops that utilize them in lieu of doing it in-house. Should junior analysts be charged with reviewing and summarizing key provisions of a credit agreement or performing a liquidity analysis of a distressed company when Reorg or CreditSights often have these reports, prepared by qualified professionals, available on-the-shelf or upon request? It’s debatable.
More recently, there are new service offerings from third-party vendors that harness the power of algorithms and AI to provide restructuring practices with virtual law clerks, automated PowerPoint decks for work pitches or other presentations, AI-generated analyses of earnings releases and investor calls, valuation analyses, targeted buyer lists, and AI-driven prospecting screens to identify restructuring candidates. Unlike previous workplace changes, such as PACER, that reduced inefficiency in a workflow process, these third-party services have mostly reduced or eliminated some workflows traditionally done by restructuring professionals. Gen AI is the latest innovation that can do the heavy lifting attached to certain work routines, either by improving productivity in the workplace or by eliminating the need to perform certain workflows in-house.
Are these recent changes unambiguously favorable? Let’s answer that question with an analogy. Consider the activity of parking a car. Most cars today are equipped with a rear-view camera display on the dashboard that is activated when the car is in reverse gear. This camera helps with parallel parking by ensuring the driver doesn’t hit the car behind him and letting the driver know when to recover the steering wheel. It assists the driver with the parking task. But there are also smart cars that automatically parallel park without the driver’s involvement. This feature doesn’t assist the driver, it replaces the driver in accomplishing this task. The driver relinquishes the job of parallel parking to the software and sensors of the vehicle. Drivers who become dependent on this novel feature will see their parallel parking skills degrade over time from lack of practice. New drivers today may never need to learn how to parallel park. The job gets done, but the driver arguably is “dumber” compared to others who do their own parallel parking. Does it matter at all?
Similarly, restructuring professionals, especially junior ones, may never have to perform many tasks that their predecessors routinely did a decade or two ago, instead relying on new or improved tools that accomplish the tasks with much less personal involvement, if any. Many of these tasks were once considered “rites of passage” along the trajectory of a restructuring career path. Today it is conceivable that young professionals may never have to slash their way through a credit agreement or POR and summarize it for managers, as an example. Again, does it matter?
What is often overlooked in these discussions is the residual learning that occurs when junior professionals get their hands dirty performing routine tasks that may not appear to be learning opportunities. Working with large amounts of data or complex legal documents in depth offers not just topical knowledge but a keener sense of the subtleties encountered, knowledge that might not be imparted when algorithms or Gen AI do most or all of the heavy lifting. If all they need to do is prompt an AI tool to perform a task, professionals who are spared the grind of these assignments in their early career development might encounter knowledge gaps and/or lack some acquired sensibilities as their careers progress — much like drivers who cannot parallel park or are unable to find a destination without GPS. Clients should no more want a financial advisor to blindly accept AI-driven results without scrutinizing those results, especially in the early adoption period of the Gen AI revolution, than we would want a radiologist to subordinate her medical judgment of our CAT scan to AI. As a rule, bad mistakes can happen when we surrender our common sense or good judgement to any sleek black box — even one powered by AI.
The inevitability of change is a cliché by now. Some embrace it, others reluctantly accept it. But it is coming, as it always has, and the restructuring profession is no exception to the continuum of change in the workplace. Some Gen AI tools relevant to the restructuring profession that we’ve seen to date seem underwhelming — more sizzle than steak — though admittedly it is early in the game.
There is a popular narrative that Gen AI will relieve working professionals of routine tasks that they dread doing anyway, freeing all of us to do more challenging and value-additive work. That’s simplistic. We can’t all be doing high-minded or mission-critical work all of the time, so the ranks of the profession could be trimmed over time if Gen AI lives up to its hype. However, such an impact would likely happen gradually, mostly through natural attrition. Remember, too, that at its core, the restructuring profession is a people-driven business that is highly dependent on fact-gathering and analysis, argumentation, negotiation and compromise. Gen AI will play some role in this process, but given the many ad hoc work streams and highly interpersonal aspects of the restructuring process, it is doubtful that the machines are coming for our jobs.
1: Michelle Faverio and Alec Tyson, “What the Data Says About Americans’ View of Artificial Intelligence,” Pew Research Center (November 21, 2023).
2: Michelle Faverio and Alec Tyson, “What the Data Says About Americans’ View of Artificial Intelligence,” Pew Research Center (November 21, 2023).
3: Rakesh Kochhar, “Which U.S. Workers Are More Exposed to AI on Their Jobs,” Pew Research Center (July 26, 2023).
4: Rakesh Kochhar, “Workers Views on the Risk of AI to Their Jobs,” Pew Research Center (July 26, 2023).
Most Popular Insights
- 10 Global Cybersecurity Predictions for 2024
- Global CFO Survey 2024
- Bridging the Gap Between Artificial Intelligence Implementation, Governance, and Democracy: An Operational and Regulatory Perspective
- The Power of Positive Paranoia: A Key Trait for Every CEO and General Counsel in 2024
- A Targeted Approach is Key to Implementing AI