Statement on AI Use
We don't.
We often hear that "AI" (we prefer the term "LLMs", since there's nothing intelligent about AI) is already here, and so we may as well adapt to the times. This boosterism presents a false dichotomy: embrace AI technologies or be left behind.
We see how AI technologies super-charge environmental degradation at a time we need to reverse it. We see how the technologies are built on the stolen creative and intellectual property of thousands of unpaid and unacknowledged creatives and writers. We see how AI infrastructures are forced on communities with little recourse to reject them. We see trillions of dollars recklessly flowing into a bubble that is on the verge of bursting, threatening tremendous global economic impacts.
At the same time, many of our clients come from sectors like education, government, NFP and NGO, and the creative sector, where discussions about the potential benefits – as well as the harms – are active. In the education sector, for example, many universities are embracing AI so long as it is acknowledged and limited to specific conditions. Yet even within universities there are differing opinions, with some disciplines fully embracing AI to ensure students are job-ready, while others critique it with deep skepticism.
Along with their many concerning aspects, AI models promise to make writing easy. What they actually do is flatten writing, rather than enhance it. AI writing is predictable, turgid, bland. For the purposes of writing, especially where authenticity, original thought, or creativity are concerned, we cannot recommend it.
In limited use-cases, we acknowledge that many clients will find AI models helpful. It seems quite good at proofreading, or offering structural advice. When someone chooses to use AI tools to edit their work, what is missing is the nuanced work a human editor does to enhance a piece of writing while maintaining the author's own voice. It can be tempting to use such tools to edit work as it is more affordable than paying a human to do so, and it is inevitable that many will prefer that option. But this also means that we are facing a period where individual expression and craft is at serious risk.
Many of our clients may choose to use AI for all sorts of reasons throughout the process of producing their work and we make no claims or judgements about this decision. We welcome all texts, regardless of AI use.
In our own editing process we do not and will not use any AI tools. All our edits are done by a human. At no stage do we feed clients' work into any AI tool or model for comment or feedback, to make any suggestions, or to do secondary work like develop a quote or draft an email. Our files are stored on a cloud system that is end-to-end encrypted and not used to train AI models, and so we are confident that our clients' work is not used for AI training.
We do, at times, make use of an editing software called PerfectIt. This tool has existed long before AI tools and uses macros to run through a series of basic editing checks. It does not use AI, and we will continue to use this tool unless they change their AI policy.
In short, our AI statement is that Ellipsis Editing does not use AI. Every email is read and responded to by a human. Every comma is added or removed by a human. And this statement was fully drafted, edited, and published by a human. We are proud of the work we do, and we will continue to work in this way until the industry can no longer support small, independent editing services like ours.