✏️ Content Design & AI Fright: Why generative AI is great, but not revolutionary for content & content design
A lot of the information and opinions I have are based on the book Life 3.0: Being Human in the Age of Artificial Intelligence by Swedish-American cosmologist Max Tegmark (1). If you haven’t read it, I highly recommend it, especially if you fancy AI scenarios, what consciousness means, or are curious about AI ethics. It’s a personal favourite! I want to point out I’m also not a quantum computing expert or cosmologist, but someone who enjoys reading about AI ethics and developments, meaning this is a think piece specifically related to generative AI in the world of technical writing and content design; you don’t have to agree with it, and perhaps you shouldn’t (what a concept in a world that’s increasingly polarised (2)).
Generative AI is great; from what I can tell, it’s here to stay. I use it almost daily as a technical writer. You can plug it into your help centre and it can help users find articles, make on-brand, short text for UI projects, and explain complex grammatical structures when something sounds or reads overly wordy, something that’s useful when creating and justifying a style guide. However, I can’t help but hear of the anxieties and doomsday scenarios off to the side. It also makes me wonder why people fear what we don’t understand yet, what we do understand, and what we can’t change (I guess that in itself is worrying!)
The early uses of generative AI, since being available en-masse in 2022 (3), caused media hype and, let’s be real, an overkill of articles on Linkedin and X telling people how to use it effectively (weird, given it’s so new), along with people expressing anxieties, especially in my field of content creation and design (again, weird given it’s so new). Sure it can do all kinds of things, but it’s a technology that is, arguably, sorely misunderstood and limited. This is in contrast to every major company including generative AI in some way, shape, or form in its products, software, and services (something that’s, arguably, not revolutionary or new, although public discourse seems to say otherwise, be it for financial shareholder glory).
I argue that, as of 2024, generative AI is doing what Google and its search engine capabilities have been doing since 1998, just in a more convenient user interface where we pop in a question and get an answer. To me, it’s not that impressive, especially given the school of thought on how problematic and dangerous AI can be. From an image generation perspective, it’s also problematic, given copyright infringement can cost companies more money than what it’s worth (4). It’s a new technology that isn’t government-regulated, is a prime candidate for instability, and is, again, severely misunderstood. Today, you might have access to generative AI, but as we move forward, it’s arguable that some companies in jurisdictions might not be able to use generative AI in the future, due to the risks alone and government regulations that don’t exist yet.
To highlight what generative AI can and can’t do, I went right to the source and asked ChatGPT:
What it can do 👍🏻
Information & Research
Answer questions on a variety of topics, including science, history, technology, culture, and more. Help with real-time updates, like news, weather, sports, etc. Assist with understanding complex topics or breaking down difficult concepts.Creative Assistance
Write or edit creative content such as stories, essays, poems, and blog posts. Generate ideas for writing, art, or other creative projects. Help with brainstorming and outlining content for presentations or speeches.Text-Based Tasks
Proofreading and editing your writing for grammar, style, and clarity. Generate summaries or paraphrases of articles or texts. Help with writing formal letters, resumes, or cover letters.Learning & Study Support
Assist with solving math problems, including algebra, geometry, calculus, etc. Help with language learning by practicing vocabulary, grammar, and conversation. Offer explanations and guides for various academic subjects.Code and Programming
Help write and debug code in various programming languages (Python, JavaScript, etc.). Provide explanations for coding concepts or algorithms. Assist with data analysis and simple automation tasks.Image and Art Generation
Create custom images based on descriptions, such as illustrations or design ideas. Help brainstorm visual concepts for projects (e.g., logos, scenes, or character designs).Entertainment and Fun
Generate jokes, riddles, or fun facts. Recommend movies, books, or music based on your preferences. Play word games, quizzes, or help with trivia.Personal Assistance
Organize your thoughts and help prioritize tasks. Create checklists, schedules, or reminders (in a text format). Suggest wellness routines, mindfulness exercises, and self-improvement tips.
What it can’t do 👎🏻
Accessing Real-Time Information (Beyond Certain Limits)
I don’t have access to live databases or private accounts, and I can’t retrieve personal information from other websites. I can’t provide real-time updates or make live purchases (e.g., stock prices, live event details, or transactions).Physical Tasks
I can’t interact with the physical world. For example, I can't make phone calls, send emails, or physically assist with tasks.Personal Interactions
I can't recognize or interpret images of people or emotions beyond what’s provided to me in context. I can't create or interact with other specific individuals beyond general conversations.Medical, Legal, or Financial Advice
I can provide general information or explain concepts, but I’m not a substitute for professional advice. For instance: Medical diagnoses or treatments Legal counsel Personalized financial planning or investment adviceAccess to External Systems or Software
I can't directly interact with your computer systems, software, or external platforms, such as running or installing programs, accessing your files, or controlling devices.Conducting Real Transactions or Actions
I can’t process payments, place orders, or perform online transactions (like purchasing goods or booking services).Generating Inappropriate or Harmful Content
I avoid creating or promoting content that is harmful, illegal, or inappropriate, including offensive language, hate speech, or violence.Predicting the Future or Giving Absolute Certainty
I cannot predict future events or give definitive answers about uncertain or speculative situations (like lottery numbers or outcomes of future events).Personalizing for Specific, Real-World Individuals
I can't know personal details about people unless explicitly provided in the conversation. I cannot track or remember your personal preferences or histories between interactions.Content That Violates Copyright
I can’t provide or recreate copyrighted material like books, movies, or music, except in cases where it is used in compliance with fair use (e.g., summaries or explanations).
With the points above, it sounds more like a helper than an ‘all-in-one’ solution, especially for creating technical content. Unless ChatGPT or other generative AI systems come tomorrow with a concept of self, which to me means it can talk to developers, publish, audit, and do the ever-growing list of responsibilities, I’m not worried, and I encourage anyone in the world of content design and technical writing to look at the bigger picture and question why they’re worried. Couple this with the list of current constraints, I feel like I will be pushing up daisies well before ChatGPT takes over the world of technical writing, and even then, will it be as good as a human? It’s not for certain, but if we, as humans (a terribly flawed and quite near-sighted species) are creating this technology and it needs to be so self-aware and so general it can do everything a technical writer can do, will it ever reach that point? Humans have spent thousands of years trying to define what ‘self-aware’ and existence are, and I feel like this won’t happen tomorrow, and even if it did, who will be there to fact-check and manage the outputs? Who is the expert? Who gets to decide who and what an expert is?
At the end of the day, do we all become prompt engineers (5)? More efficient at what we already do? Obsolete? Are companies mentioning generative AI in their products because it's a buzzword, and although helpful, is it possible it can replace all of us? I don’t think anyone knows for sure, given AI itself can’t predict its future, and even cosmologists and AI experts can’t predict what will happen (cue Teagmark’s book). It will depend on our adaptability to whatever the “AI overlords” throw at us and how governments treat it, manage it, and allow countries and businesses to use it, and if government cooperation and investment in people are any indicator, especially in most societies in 2024, it’s going to be a very long time. It’s arguably too early and unstable a piece of technology for all of this to be remotely considered, and until then, it’s a gamble for everyone who uses it, especially technical writers and content designers, who in a lot of cases sign non-disclosure agreements to keep business secrets safe. Even if your company is using internal AI developments, someone still needs to manage the information and resources. Again, a time saver, but nothing new or impressive here.
I won’t be impressed until AI can solve some of the world’s biggest issues, like climate change, war, poverty, or terrorism, let alone do all the work of a team of technical writers or content designers. When the constraints I mentioned above are black and white and solved, which, again, doesn’t seem near (6), it does make a nice search engine and content design companion.
