SINGAPORE: I was conducting a generative artificial intelligence (Gen AI) workshop at a secondary school last year when I noticed one teacher sighing heavily. She seemed less than happy as I demonstrated how Gen AI could create digital paintings in the style of Renaissance art.
“I don’t see how Gen AI can replace my 30 years of teaching art,” she said. “How should we teach art to students now?”
“Exactly the way it has always been done,” I replied. “We need to teach them how to draw well, otherwise, they will not be able to make use of AI to create great art.”
This advice puzzled her. Then I showed her how Gen AI is best used to add colour and texture to a line drawing or sketch, instead of generating a whole new image that the artist cannot control. “We must remain in control of AI!” I said firmly.
For those who create content for a living, Gen AI feels like an existential crisis. With today’s Gen AI apps, a few keystrokes are all you need to generate entire stories, music, images and even videos.
As a writer and artist trained in the traditional way - thousands of hours of practice on paper, getting my writing torn apart (figuratively and literally) by editors - I understand how my brethren feel.
I can also relate to the recent petition by 88 Singapore writers denouncing Singapore's National Library Board (NLB)’s approach towards Gen AI.
Last year, NLB rolled out programmes such as StoryGen, which uses Gen AI to create new variations of fairy tales such as Little Red Riding Hood, as well as workshops like Children Write: Publish A Book With Gen-AI.
The writers’ petition raised four major concerns: The ethical issues of intellectual property misuse, as many AI models are trained on copyrighted materials without permission; the psychological impact, as reliance on AI undermines the development of writing skills and creativity; the risk to literary quality, since AI-generated works tend toward mediocrity and lack originality; and the significant environmental cost of AI, including its energy and resource-intensive nature, which contradicts NLB’s sustainability values.
Now, I must tell you that I did not write the paragraph above.
I fed the petition letter into ChatGPT and prompted it to “do a one-paragraph summary of the four concerns”. The chatbot spat out the information in two seconds.
Am I less of a professional writer for doing this? Can I put my name to this article if parts of it were not actually by me? And could you even tell?
This is what the local writers want NLB to warn the public about - that Gen AI is too easily used to substitute human effort. I agree with the writers’ concerns, but Pandora’s box has been open for too long.
The two-year-old ChatGPT is now the 9th most popular website in the world, and had 3.7 billion visits in October 2024, according to online tracker Similarweb. Gen AI is integrated into your Android phone or iPhone. A recent Google-Ipsos survey across 21 countries found that 80 per cent of Gen AI users use it for communications and writing.
Using Gen AI is now the norm, not the exception. The petition is well-intentioned, but it is hard to convince people to pull back on using Gen AI (especially when it is free).
If there is any public message to send now, it is that we humans must choose whether we want to be augmented or automated by Gen AI.
Two years ago, an acquaintance asked if I could provide English coaching for her. It was a difficult task because she didn’t like to read books and she never received any feedback on her broken English.
I showed her how ChatGPT could clean up her written work with a simple prompt like “check grammar”. She was surprised to see how many grammar errors ChatGPT caught and fixed. I also use the same prompt to help me catch typos in my own writing; it’s like having a newsroom editor by my side.
Gen AI is a godsend to millions of people who struggle with the written word.
On LinkedIn, I’ve seen how people were too shy to post content, but started doing so when they realised ChatGPT could help them articulate their thoughts more clearly. In this way, Gen AI augments our capabilities and helps us communicate in better ways.
But Gen AI will not make you a good novelist.
If you do not have a strong command of language and its nuances, you will not be able to evaluate Gen AI’s output with a critical eye. You will also be unable to edit the generated text to fit your own writing style, because you don’t have a style to begin with.
This is the irony of using Gen AI - it takes a subject matter expert to wield it well.
Of course, there will be people who fully automate their writing using Gen AI.
LinkedIn has become flooded with posts and comments that are obviously written by AI. Google and Pinterest are choked with AI-generated photos. YouTube will be next, once AI-generated videos get better and cheaper to use.
These users do not care about creating authentic content or engaging other people. They just want AI to do the thinking and writing for them. If they apply the same passive attitude to their work, they will find themselves being replaced by AI agents in the future.
Perhaps that day has already arrived. Meta is rolling out tools for people to create AI characters on Facebook and Instagram.
Social media will soon be filled with AI bots talking to each other.
In this new age of “AI Slop”, people are starting to flee the mass-produced drivel and seek the warm voices of fellow humans. 35 million people subscribe to Substack newsletters and there are about 4.4 million podcasts today.
To my fellow writers, let’s seize the day too.
If you haven’t already started doing so, choose your favourite social media platform and publish excerpts of your work, share daily thoughts, comment on the world, review the latest shows, and so on. For instance, I have started writing a book on marketing and I’m posting draft chapters weekly to get crowdsourced feedback from LinkedIn and Facebook friends.
I am confident that humans still want to engage with each other. And as the world sinks under this AI-generated deluge, I will still be here typing one word at a time for my readers.
But don’t mind if I ask ChatGPT to check my spelling.
Ian Yong Hoe Tan is a strategic communications lecturer at the Wee Kim Wee School of Communication and Information, Nanyang Technological University. He has more than two decades of experience working in the media and technology industries.
Continue reading...
“I don’t see how Gen AI can replace my 30 years of teaching art,” she said. “How should we teach art to students now?”
“Exactly the way it has always been done,” I replied. “We need to teach them how to draw well, otherwise, they will not be able to make use of AI to create great art.”
This advice puzzled her. Then I showed her how Gen AI is best used to add colour and texture to a line drawing or sketch, instead of generating a whole new image that the artist cannot control. “We must remain in control of AI!” I said firmly.
For those who create content for a living, Gen AI feels like an existential crisis. With today’s Gen AI apps, a few keystrokes are all you need to generate entire stories, music, images and even videos.
As a writer and artist trained in the traditional way - thousands of hours of practice on paper, getting my writing torn apart (figuratively and literally) by editors - I understand how my brethren feel.
I can also relate to the recent petition by 88 Singapore writers denouncing Singapore's National Library Board (NLB)’s approach towards Gen AI.
Related:
PANDORA’S BOX HAS BEEN OPEN FOR TOO LONG
Last year, NLB rolled out programmes such as StoryGen, which uses Gen AI to create new variations of fairy tales such as Little Red Riding Hood, as well as workshops like Children Write: Publish A Book With Gen-AI.
The writers’ petition raised four major concerns: The ethical issues of intellectual property misuse, as many AI models are trained on copyrighted materials without permission; the psychological impact, as reliance on AI undermines the development of writing skills and creativity; the risk to literary quality, since AI-generated works tend toward mediocrity and lack originality; and the significant environmental cost of AI, including its energy and resource-intensive nature, which contradicts NLB’s sustainability values.
Now, I must tell you that I did not write the paragraph above.
I fed the petition letter into ChatGPT and prompted it to “do a one-paragraph summary of the four concerns”. The chatbot spat out the information in two seconds.
Am I less of a professional writer for doing this? Can I put my name to this article if parts of it were not actually by me? And could you even tell?
This is what the local writers want NLB to warn the public about - that Gen AI is too easily used to substitute human effort. I agree with the writers’ concerns, but Pandora’s box has been open for too long.
The two-year-old ChatGPT is now the 9th most popular website in the world, and had 3.7 billion visits in October 2024, according to online tracker Similarweb. Gen AI is integrated into your Android phone or iPhone. A recent Google-Ipsos survey across 21 countries found that 80 per cent of Gen AI users use it for communications and writing.
Using Gen AI is now the norm, not the exception. The petition is well-intentioned, but it is hard to convince people to pull back on using Gen AI (especially when it is free).
If there is any public message to send now, it is that we humans must choose whether we want to be augmented or automated by Gen AI.
Related:
AUGMENTED HUMANS: “AI, CHECK MY GRAMMAR”
Two years ago, an acquaintance asked if I could provide English coaching for her. It was a difficult task because she didn’t like to read books and she never received any feedback on her broken English.
I showed her how ChatGPT could clean up her written work with a simple prompt like “check grammar”. She was surprised to see how many grammar errors ChatGPT caught and fixed. I also use the same prompt to help me catch typos in my own writing; it’s like having a newsroom editor by my side.
Gen AI is a godsend to millions of people who struggle with the written word.
On LinkedIn, I’ve seen how people were too shy to post content, but started doing so when they realised ChatGPT could help them articulate their thoughts more clearly. In this way, Gen AI augments our capabilities and helps us communicate in better ways.
But Gen AI will not make you a good novelist.
If you do not have a strong command of language and its nuances, you will not be able to evaluate Gen AI’s output with a critical eye. You will also be unable to edit the generated text to fit your own writing style, because you don’t have a style to begin with.
This is the irony of using Gen AI - it takes a subject matter expert to wield it well.
Related:
AUTOMATED HUMANS: “AI, DO MY HOMEWORK”
Of course, there will be people who fully automate their writing using Gen AI.
LinkedIn has become flooded with posts and comments that are obviously written by AI. Google and Pinterest are choked with AI-generated photos. YouTube will be next, once AI-generated videos get better and cheaper to use.
These users do not care about creating authentic content or engaging other people. They just want AI to do the thinking and writing for them. If they apply the same passive attitude to their work, they will find themselves being replaced by AI agents in the future.
Perhaps that day has already arrived. Meta is rolling out tools for people to create AI characters on Facebook and Instagram.
Social media will soon be filled with AI bots talking to each other.
WRITERS, SEIZE THE DAY
In this new age of “AI Slop”, people are starting to flee the mass-produced drivel and seek the warm voices of fellow humans. 35 million people subscribe to Substack newsletters and there are about 4.4 million podcasts today.
To my fellow writers, let’s seize the day too.
If you haven’t already started doing so, choose your favourite social media platform and publish excerpts of your work, share daily thoughts, comment on the world, review the latest shows, and so on. For instance, I have started writing a book on marketing and I’m posting draft chapters weekly to get crowdsourced feedback from LinkedIn and Facebook friends.
I am confident that humans still want to engage with each other. And as the world sinks under this AI-generated deluge, I will still be here typing one word at a time for my readers.
But don’t mind if I ask ChatGPT to check my spelling.
Ian Yong Hoe Tan is a strategic communications lecturer at the Wee Kim Wee School of Communication and Information, Nanyang Technological University. He has more than two decades of experience working in the media and technology industries.
Continue reading...