Predicting the Future of AI for Lawyers: Finding, Minding, and Grinding

We’ve been inundated with talk about artificial intelligence (AI). I’ve been told all the ways it can change my life and my practice. I’ve been impressed with some such solutions and wasted time and money on others.

Despite all the hype, one comment keeps rattling in my head. I think back to when OpenAI released a new version of ChatGPT that could create art at the user’s request, and the following comment gained some Internet traction: “I want AI to do my laundry and dishes so that I can do art and writing, not for AI to do my art and writing so that I can do laundry and dishes.”

Who wouldn’t feel this way? Of course, that leaves the question of what’s “art” when practicing law, and what’s “laundry.” The answer to these questions will vary by the practitioner, by the role they play in their organization, and by practice area.

For the purposes of this article, I look at the future of AI through the lens of what’s art versus laundry according to the storied categorization of lawyers into finders, minders, and grinders. The reason I focus on this is because, in the solo or small firm context, attorneys are usually charged with fulfilling more than one of these roles.

This is why it can be exhausting to be a solo. Most people tend to be extremely proficient in one of these roles but might be lacking in others. This skills deficit creates tension, both internally and organizationally. If your knack is to make rain, it can gall you to draft briefs. If your knack is to draft briefs, it can pain you to talk to clients or attempt to bring in business. What is “art” or “laundry,” and thus how you want to approach incorporating AI into your practice in the future, will consequently vary, as will the tools that make your practice more fulfilling.

As a preliminary matter for those not familiar with this nomenclature, the terms “finders,” “minders,” and “grinders” were coined by David H. Maister in his 1997 book Managing the Professional Service Firm, which discusses the various roles in a professional setting necessary for it to succeed. In the legal context, these terms are typically understood to mean the following:

• Finders. These are the rainmakers who bring in the work.
• Minders. These are the attorneys who “mind” the file, meaning they run a matter day-to-day. Depending on the firm, they could manage the associates or interact with the client on matter status, but typically, they serve in some managerial role.
• Grinders. These are the producers of the work product. They draft the briefs, respond to the discovery, handle document review, etc.

What’s Art and What’s Laundry in a Law Firm?

For me, here’s a rough overview of how I’d distinguish art from laundry in each of these roles.

When I’m Grinding

Art is creating quality work product. It’s focusing on the critical provisions in a contract, as opposed to looking for the right template or sifting through boilerplate. Laundry is staring at a blank page. Laundry is proofreading. It’s confirming that internal references are accurate after rounds of revisions. It’s fixing party names and ensuring all terms are defined.

When I’m Minding

Art is a high-functioning firm machine. It’s reviewing and running with high-quality work product. It’s maintaining harmonious client relationships. It’s bills that go out on time and are promptly paid. The laundry is human resource management. It’s slogging through poor-quality drafts. It’s bills that need to be discussed because there were inaccurate time entries. It’s typos.

When I’m Finding

Art is connecting with the right people and closing more engagements. Laundry is the paperwork. It’s missing opportunities to follow up. The laundry is time spent on looky-loos who have no interest in paying for legal services.

How Far Away Is the Future?

So, how far are we from being artists as opposed to mere laundromats? AI has made some remarkable inroads toward eliminating what grinds me about grinding, but there’s still some room for improvement when it comes to minding and finding.

Grinders

In the transactional space, there are drafting tools that can help draft documents from scratch. Spellbook, for example, has been at the forefront of the AI revolution for lawyers. With Spellbook, you can take written comments from a client or another attorney and ask Spellbook to make revisions throughout the entire agreement. You can ask it to review a document, and it will suggest missing provisions and find undefined definitions, inconsistent provisions, and incorrect internal citations. It’s so accurate that I can even ask it to revise party name references and subject-verb agreement with great success (say goodbye to CTRL+f).

On the litigation side, there’s been less activity. Briefpoint, however, has been an exception. It can take a complaint or propounded discovery and generate detailed, case-specific requests for admission, requests for production, interrogatories, and other discovery documents in less than a minute. You can also add your own library of clauses or requests so it can be even more tailored to your style of practice. Briefpoint can create quality discovery in less time than it would take to write an email to an attorney or paralegal to instruct them on what you want.

Tools such as Spellbook and Briefpoint have made huge strides toward making drafting more pleasant, but what’s on the horizon? Legal-specific generative AI (GenAI) solutions have typically been focused on a single query or document at any given time. For example, you ask ChatGPT a question, and it provides an answer. Drafting tools generate one document or edit one thing. But what if you made a change in one document, and your tools anticipated the edits that would need to be made in other documents? According to Spellbook cofounder Matt Mayers, that’s the next step:

Traditionally, AI tools have had a singular output, such as an answer to a question in a chat. Now, AI agents are capable of completing more complex projects. We recently launched Spellbook Associate, the first legal AI agent that understands, plans, and executes multi-step tasks from start to finish. For lawyers, this means more than just automation—it’s about unlocking efficiency and precision. For example, it can draft complete financing documents from a single term sheet, prepare employment packages in minutes, create disclosure schedules, and review entire document sets for risks and inconsistencies.

Wow. What would normally take detailed instructions to an associate could potentially be done in a quick prompt or document upload. This next phase of AI means lawyers will spend less time manually updating related documents, which is a huge source of errors and typos in more complex transactions. If a key term changes in one agreement, agentic AI can automatically adjust corresponding provisions across the relevant documents.

Minders

To the extent that minders are meant to add to or review a grinder’s work product, Spellbook, in particular, shines. I can ask Spellbook specific questions that someone grinding the document might’ve missed. Even just asking Spellbook to simply review the entire document for typos, undefined definitions, wild party names, etc., can start me off on a productive path toward improving the document.

When it comes to the management role of minding, however, I haven’t seen too much help yet come from legal-specific AI. As a minder, I need to staff, track hours, track realized time, analyze matter budgets, ensure payment of bills, etc. Sure, nearly every AI tool promises in some form or another to produce “data” to help inform decisions, but I haven’t seen anything yet to truly make minding from the management perspective feel more like art. The minder is still left to monitor reports and data, such as time billed versus realized, and act on it independently. The data insights thus seem much more akin to using more powerful detergent.

That is, however, unless Spellbook Associate can augment the output in work quality or reduce the volume of grinders. The higher number of juniors you’re managing likely contributes to a higher number of headaches you’re having to deal with, both from work product quality control and human resources management. Merely having a more effective class of grinders will certainly reduce human resource management concerns. To the extent a minder is able to reduce head count while maximizing the reach of your ace grinders, AI is a valuable tool.

Finally, one of the biggest complaints against lawyers is that they take too long to respond. This is where tools such as Microsoft Copilot and Google’s Gemini come in. They can help draft and summarize emails, and they try to improve their feedback based on the edits you make so your responses can read like they’re coming from you.

Gemini has already released development road maps for future use cases, including hiring and customer sentiment insights. Although its calendaring functionality is limited at the moment, I don’t think it will be too far off for it to serve in a scheduler role, and I sincerely hope it will be able to create deadlines and calendar according to court rules.

Finders

Perhaps due to the nature of finding, finders haven’t had quite as many AI tools built for them that are specific to the legal industry.

Nonetheless, finders would likely benefit from incorporating AI communication tools such as Copilot or Gemini into their workflows. As these communication tools evolve, they mirror your communication style and make drafting emails much easier.

Automated email messages will also improve. Everyone knows when they receive an automated message. My inbox is littered with examples with typos in my name or more general salutations for marketers wary of that faux pas. But what if email communications were tailored by AI? That would mean more quality messages, more quality connections, and, by extension, more engagements for the firm.

Email automation is cool, but let’s take it a step further. What about duplicating the firm’s rainmaker to be in multiple places at once? That possibility isn’t too far off.

One of the major sticking points of the Hollywood actors’ strike was limiting the use of GenAI by studios to create actors’ likenesses on the big screen. The actors’ concern was that studios could use AI to completely replace them with AI-generated content. The next step for this type of technology is generating dynamic video content in real time and coupling it with the conversational power of tools such as ChatGPT.

Rumor has it that this technology is already in use. One client mentioned that he had an hour-long video call with two individuals at a prospective investment firm. He later learned that the people he spoke with for almost an hour didn’t actually exist—they were AI generated.

With technology this powerful, the firm’s rainmaker could be in multiple places at once, closing multiple clients at once, at least if the meetings are on Zoom. The same technology the actors’ guild was concerned about could be used to create an increasing number of video ads without the need for the firm’s video actors to even appear for filming. It could also be used to streamline and create more consistent client intake teams.

Using an AI-generated rainmaker to talk with prospective clients comes with some obvious ethical concerns that should be the subject of much more discussion, but this possibility is likely to become a reality much sooner than we might anticipate.

Less Detergent, More Paint?

When I’m analyzing the next tech I’m looking to adopt, my first question is, “Will it help me do less laundry?” As you analyze AI for the future, keep in mind your strengths and weaknesses so you can identify tools to augment the areas that give you that laundry-like feeling.

On the whole, I see slightly less laundry in the future, but perhaps not at the pace we might like. I’m excited to see what developments agentic AI might bring, but one limiting factor I see before attorneys feel like Picasso is the lack of cross-platform functionality. I’d like the insights and commands from my email to speak to my other systems. The fatigue caused by context switching is real; our need to switch from one program to another saps our energy and attention. Imagine if a painter needed to leave the room every time the artwork required a different color.

Until then, there’s still a lot of clothes in the hamper.