As Rain’s lead writer and someone who makes a living putting words together, I wish I would have had the foresight to purchase a policy from Old Glory Insurance. As faux spokesman Sam Waterson explained in a 1995 Saturday Night Live faux ad, the company offered policies that insured against “the unfortunate event of a robot attack.” While I was not physically attacked by a robot, my ego certainly was wounded by one when a non-Rain BDM client of mine, whose marketing company had been recently acquired by a larger competitor, replaced me and my content writing services with generative AI.
While my hunch is that this loss of a decade-long client relationship to AI would have been excluded under Old Glory’s policy, the existential (and financial) hit was nevertheless jarring, even though I worried about just such an occurrence since the first time I gave an experimental prompt to ChatGPT.
My client, apologetic and clearly not happy with the directive from his new corporate overlords, told me it was simply a matter of dollars and cents; the dollars I charged for a blog post or other piece of content would cost only cents when created by generative AI. It was undoubtedly faster, too, cranking out full articles in the time I took to draft the first couple of sentences (not to mention any research I may have needed) to write the piece.
But the more I thought about it, and the more conversations we had here at Rain BDM about the ifs, hows, and whens of leveraging generative AI in our content writing, the less concerned I became about a forced early retirement. In fact, I have a feeling my former client will soon realize the penny-wise, pound-foolishness of swapping out a human for a robot, especially when it comes to crafting sophisticated, nuanced, creative, and, most critically, accurate client-facing legal content.
There are many convincing reasons to think twice before handing over your firm’s content to a faceless, fallible, and forgettable algorithm.
The Rain View of AI: We Prefer Humans.
Speaking on Rain’s behalf, we’ve been fielding an increasing number of inquiries about the use of generative AI, which isn’t surprising. Current and prospective clients ask if we utilize ChatGPT, Gemini (formerly Bard), Copilot, and other tools to produce blog posts, web pages, press releases, and other written content.
The short answer is no, we don’t do that. Rain has a stated policy against using generative AI in any content drafting, whether for clients or our own communications.
If you want the long answer, keep reading.
Algorithms Lack Nuance.
Proudly and unabashedly, Rain is really good at creating content. Thanks to our crazy-talented team of lawyers-turned-writers and our meticulous proofreader, we translate even the most complex legal concepts and intricacies into digestible, memorable material that resonates with each client’s target audiences. We work closely with attorneys to divine and channel their voices and styles. That penchant for personality and personalization is a compelling part of the value we deliver.
AI evangelists are quick to proclaim that what you get from ChatGPT et al. is only as good as what you put into the prompt and that subpar content is the fault of the user. If you want superior output, you have to tell it exactly what it needs for the alchemy to happen.
Okay, we’ll play that game.
Here’s a query we posed to Gemini: “Explain how U.S. federal courts have interpreted the Sherman Antitrust Act since 2020 relative to Google and Apple in the style of Dr. Seuss.”
In the land of Apps and Bytes, where tech giants take flight,
Two big companies, Google and Apple, hold immense might.
But Uncle Sam, with a keen eye, on fair play and such,
Thinks these tech titans might have grown just a smidge too much!
The Sherman Act, a wise old law, from eighteen-ninety,
Says competition’s the key, for a market quite shiny.
But courts since 2020, have peered at these two,
Wondering if their app stores might be a bit askew?
They say that Google, with its grip on Android’s domain,
Might be making it tough, for other app stores to gain.
Like a Grinch with a heart three sizes too small,
They might be squishing competition, standing oh so tall!
Apple’s App Store’s another case, with a judge’s raised eyebrow,
Forcing them to open the gates, to let some payments flow out.
Not just their own in-app deals, like a tollbooth so tight,
But letting others in on the game, to make things just right!
Now, these are just rumblings, the trials still unfold,
But the message is clear, for these tech giants of old:
Don’t be a Snuffleupagus, blocking the way,
Let competition flourish, for a market that’s fair, hooray!
Without question, the speed at which Gemini produced the text is astounding, and its ability to incorporate relevant factoids is impressive. But just because the text rhymes doesn’t make it Seussian, and while the nod to the Grinch is clever, referencing Snuffleupagus is an affront to both Whoville and Sesame Street.
Next, we asked ChatGPT: “Write a 300-word blog post about part-time FMLA leave targeted at business owners and human resources managers.”
Title: Understanding Part-Time FMLA Leave: A Guide for Business Owners and HR Managers
As a business owner or human resources manager, navigating the complexities of the Family and Medical Leave Act (FMLA) can be challenging, especially when it comes to part-time employees. Part-time FMLA leave adds another layer of intricacy to an already intricate process. Understanding the ins and outs of part-time FMLA leave is essential for ensuring compliance and supporting your employees during their times of need.
Part-time FMLA leave allows eligible employees to take intermittent or reduced-schedule leave for qualifying reasons outlined in the FMLA. This means that part-time employees, just like their full-time counterparts, may be entitled to job-protected leave for medical and family reasons, such as caring for a newborn, a sick family member, or dealing with their own serious health condition.
For business owners and HR managers, it’s crucial to familiarize yourself with the specific requirements and regulations surrounding part-time FMLA leave. This includes understanding eligibility criteria, notification procedures, and documentation requirements. Ensuring clear communication with employees about their FMLA rights and responsibilities is also essential to avoid misunderstandings and potential compliance issues.
Implementing effective tracking and documentation systems can help streamline the administration of part-time FMLA leave. This includes accurately recording leave taken, tracking available leave balances, and maintaining thorough documentation of any correspondence or medical certifications related to FMLA requests.
Moreover, fostering a supportive and inclusive workplace culture can make a significant difference in how employees perceive and utilize FMLA leave benefits. Encouraging open communication, providing flexible work arrangements when possible, and offering resources for managing work-life balance can contribute to a positive employee experience and promote retention.
In conclusion, navigating part-time FMLA leave requires diligence, clear communication, and a commitment to compliance. By understanding the requirements, implementing effective processes, and fostering a supportive workplace culture, business owners and HR managers can successfully manage part-time FMLA leave while supporting the well-being of their employees.
Again, A+ on speed. The substance and restatement of the law are solid, and for SEO purposes, there are search-friendly keywords aplenty (including “part-time FMLA leave” repeated six times after the title). Yet, one could easily mistake this for a section from a payroll company brochure. Where’s the insight and the application? What’s at risk for the company for non-compliance? When and how can the employer lawfully terminate the employee?
Efficient and Sufficient? Yes, But What About Proficient?
Sure, we could have expanded the prompt to elicit a more in-depth response, and by requesting 300 words (actually, ChatGPT spit out 321), perhaps we over-restricted the algorithm. But if the time it takes to become a better query drafter, who then has to double-check for accuracy and substance, is greater than writing an original piece that meets and exceeds the client’s expectations, what’s the business case for that?
Our clients pay us for content services under a monthly outsourced marketing department structure, a fixed fee by content type and length, or less frequently, a straight hourly rate. Relative to the first two arrangements, efficiency is critical. It would be in Rain’s best economic interest to take advantage of generative AI, right?
And Then There’s Accuracy.
Generative AI generates information, and the cleaner and more streamlined the information, the better the output. Case law is messy. Procedural histories can be labyrinthian. Precedent is controlling until it isn’t. And when the data sets from which AI services pull currently reach back less than a decade, the basis for the information is definitely suspect and presumptively flawed.
Categorically, Rain does not publish legal content without a lawyer’s review and approval. As good as we are and notwithstanding our law practice experience, our writers will never know the law as well as our clients. As much as they rely on us for our written communication strengths, we rely on them to ensure the content is correct. It’s a non-negotiable partnership.
Sadly, some lawyers have placed too much reliance on generative AI to prepare court filings, putting them in ethical jeopardy. Search Google for “lawyers misuse of AI,” and you’ll see various examples. By avoiding AI for content drafting and partnering with our clients to confirm accuracy, we steer clear of that risk entirely.
An Exception for AI as a Content Assistant.
In its present state and how most law firms think about using it for content development, AI is more of a novelty than a disruptive innovation. We expect that to change over time, but not today, tomorrow, or in the foreseeable future.
However, AI has aspects that are useful to our team. For example, we sometimes consult ChatGPT or Gemini about suggestions for blog titles, keyword variations, and subheadings. If a client provides an industry-speak heavy medical record, engineering report, or scientific journal article, we might ask AI for a plain English explanation.
We depend on Grammarly to embolden our writing by streamlining sentences, detecting passive voice usage, and ensuring the drafts we deliver are 100% original (and where others’ work product may not be). We use Descript to transcribe video and audio files so we can then edit and caption the final clips. On the off chance we cannot find a suitable photo or graphic to accompany a blog or social media post, we might enlist Canva’s AI Image Generator for options.
People Using AI Rather Than AI in Place of People.
Lest you think we are AI alarmists (although we recall “The Terminator” on the big screen when it was first released), we’re not. Generative AI is inescapable. As the algorithms, applications, data sets, and guardrails evolve, it will almost certainly become a fixture in business and society.
But we are not there yet. Not by a long shot. Until then, our money is on Rain’s team of humans.