Scott Stevenson doesn’t want every employee’s idea to get an AI glow-up.
The co-founder and CEO of Spellbook said employee proposals containing “AI nonsense” routinely land on his desk. It could be a project plan or a marketing pitch. Whatever it is, he said, it can look much more polished than the thinking behind it.
Stevenson said he would prefer to see the rough cut first.
“I want the D+ version of your plan,” he said.
Stevenson is now asking employees at the Toronto company, which is developing an AI tool for drafting and analyzing contracts, to resist the temptation to turn every idea into polished business prose.
His message is remarkable, coming from the CEO of an AI company: Use the technology, but don’t allow it cloud your thinking.
AI slop on social media and networking sites like LinkedIn has attracted a lot of attention. Stevenson worries about a workplace version of the problem.
Other companies are also grappling with this challenge: the head of AI video company Synthesia warned staff earlier this year about “AI sloppiness,” writing overly long-winded documents could mean that instead of one employee taking ten minutes to tighten up a document, “we have ten people spend ten extra minutes reading sloppy material.” Business Insider previously reported on AI software startup Clay’s new company-wide AI writing policy. It focuses on four principles, including “writing is thinking” and “longer is not better.”
Stevenson is trying to address the problem with his own guidelines. He recently sent a memo to the company’s 200-plus employees updating the company’s AI policies and outlining how he wants them to use the technology.
Here’s an excerpt:
While we are an AI company and believe in the usefulness of AI, we think it works best today for very narrow work such as coding or contract review. We don’t yet believe that models are good at casual long-form writing. Documents are too long and say too little, use repetitive language that reduces confidence in your thinking, and sometimes contain baseless nonsense. Although it may seem powerful and professional to be able to create a polished long document with a prompt, this impression is often only superficial. Good writing is concise and authentic. I’d rather you send a messy five-bullet proposal than five pages of polished fluff. Moreover, AI has become writing increasingly recognizable. We want our customers to always feel our care when we communicate with them, and we don’t think AI-generated writing makes this possible. Our new AI policy is as follows:
- Indicate whether a document or message you send internally is created with AI: this helps the reader process it properly.
- AI should not be used for writing the body of project proposals: we want to see your authentic reasoning. You can use AI for additional material such as spreadsheets and tables, but not for the core of your reasoning.
A. If in doubt, send a “D+ Version”: we often don’t want a long or polished document. Concise, authentic bullet points are often enough to get a conversation started. Thinking about how to make your argument concise is helpful to the reader.
- The letter you send will be treated as your own: A lawyer is liable if he presents an argument containing hallucinatory case law. We will also hold ourselves accountable for our own writing. Please check everything you send to the team carefully as it will be treated as your property
- Customer communications should always be handwritten: We want our 5,000 customers to experience authentic human care in every interaction.
Is your company reconsidering its policies on AI in the workplace? What do you think about receiving texts produced with AI? Let us know in the comments below.
