My inbox is full of press releases from furnishings companies claiming their latest collections exhibit exceptional craftsmanship, are deeply rooted in cultural heritage and are redefining modern design. These descriptions are AI-generated.
Reading redundant drivel is going to make us all go mad, but for thought experiment’s sake: Is there an instance in which AI-generated product descriptions are fine? Or tell us anything useful?
I consult with interior design and home furnishings executives on how to grow their businesses with media and technology. For much of the past year, the focus has been on developing their internal artificial intelligence protocols.
It should come as no surprise that the resistance to adopting AI generally comes from the employees (and rightfully so). Whereas eagerness to infuse AI into every facet of a business generally comes from the C-Suite. Problems arise when both are cagey about discussing AI-usage with one another. For any company, deciding whether to use AI — or to what degree — begins with an open, good faith conversation about:
- Company values
- Detaching productivity from time
- Shifting the focus onto consistency and quality of the work product
- Valuing taste
I could write 10,000+ words on each of those points, but in order to limit the scope of this article, I am going to use the example of the AI-generated press release or website product descriptions as a springboard to discuss some common concerns that arise when building a company’s internal AI protocols, namely: ethics, creativity and sustainability.
AI and creativity
Naturally, there is going to be some repetition when writing about consumer goods as opposed to, say, art. When originality isn’t the point, perhaps AI-generated descriptions can suffice.
Although online, that repetition hurts SEO and that will only intensify as AI search capabilities are universally adopted. (Read: Google recently laid off their search team)
Otherwise, at least treat it like MadLibs: Recognize what is overused, and come up with unique, accurate replacements.
At least tweak it, please
The cynical (and former tech executive) side of me says AI is not going anywhere, if you can’t beat them, join them.
To paraphrase what the late, great David Lynch said back in 2007, ‘AI is a pencil, everyone has access to a pencil, it’s how you use it.’ I say this with trepidation, but I am inclined to agree. A common take within the tech industry right now is: AI is not going to take your job, a person who knows how to use it will. With a million caveats, I agree.
The more you experiment with generative AI large language models, the more you familiarize yourself with its capabilities and limitations. It quickly evolves and, sometimes, it regresses. It can summarize and it can paraphrase, it can outline and it can edit. It gets the gist. But it cannot make your point for you. This is where taste comes in.
The art of prompting
As you begin to dance with LLMs, prompting becomes an art and skill in and of itself. Human discernment (taste) is also what enables someone to decide when Chat GPT’s hallucinations are too frustrating, when it is being repetitive, inaccurate or simply is not performing to your standards. Then, it’s better to do things the good old fashioned way.
With discerning taste, a good use of AI might be to use it as a brainstorming tool – working backwards to determine the things you don’t want to say.
There are other positives too, such as the mounting evidence to suggest that LLM’s are making life easier and more inclusive for people with intellectual disabilities ranging from dyslexia to Autism.
As with the introduction of any disruptive technology, many people will sound the alarm bells and resist change. With AI in particular, many will use it to coast and cling to the status quo for as long as possible. Very few will take the opportunity to reclaim their time and be honest about what makes them most useful, which, in this case, is synonymous with most human.
Much of our time is filled with fluff and filler, why not outsource it to the robots or eliminate it entirely? And as the bots take up a greater share of the busy work, how can we be cognizant not to replace it with more busywork or bulldoze scared employees in the process?
Mistakes will be made as we figure this out collectively. Self-awareness and honesty about what does and does not require a human touch is going to be imperative.
AI and sustainability concerns
If you’ve made it this far, maybe you’re someone who cares about the environment. Let’s talk about it.
AI takes a concerning amount of energy to produce results. As reported by MIT Technology Review, “OpenAI and President Donald Trump announced the Stargate initiative, which aims to spend $500 billion—more than the Apollo space program—to build as many as 10 data centers [for AI] each of which could require five gigawatts, more than the total power demand from the state of New Hampshire. Apple announced plans to spend $500 billion on manufacturing and data centers in the US over the next four years. Google expects to spend $75 billion on AI infrastructure alone in 2025.” This amount of energy expenditure is, as they point out, not a norm in technology. It is unique to artificial intelligence.
If you are someone who writes off AI entirely for environmental purposes: That’s great. Much like those who refuse to buy from Amazon in favor of shopping locally: You’re amazing. But I’m not here to admonish. Same-day delivery is pretty awesome.
As the co-author of Green Interior Design, I try to walk the walk as much as possible. My big sustainability flex is that although I live in what many consider to be a driving city, I haven’t been behind the wheel of a car in years. I also realize that isn’t possible for most people.
That is one of the main points the book drives home: There isn’t just one shade of green. How sustainable something is lands on a spectrum. Modern life is not conducive to living a zero-emissions lifestyle, but we can be more conscious of the things we do everyday to help mitigate our human and corporate contributions to climate catastrophe, including what we utilize artificial intelligence for.
So what do we do — Use AI, not use AI? That, too, is a question better answered on a spectrum: How and to what extent. Before using it to tell us your product or service is exemplary of fine craftsmanship, deeply rooted in cultural heritage, boundary-pushing and redefining modern design, and any other tells, consider how even if they are accurate descriptors, these buzzwords have lost their meaning thanks to LLMs.
If there is a lesson to be learned from their overuse, it’s this: AI is trained on us. It is a mirror reflecting back what we want. When prompted to “write a press release to best sell this furniture collection,” its answers reflect back a cultural craving for both comfort and novelty.
There are many aspects of artificial intelligence we didn’t unpack – intellectual property concerns, AI product rendering, non-creative uses and additional benefits of AI, future of work and employment, etc. Let me know if you’d like more content like this and we can delve deeper.
