Search the Community

My first MCF blog post of 2026 started with me looking through what had been posted in the two months I hadn’t had time to write anything - a way to stimulate my imagination and avoid writing similar topics in a short period of time and risk monotony (aren’t I considerate of you dear readers hahahaha *pukes*). My Editorial colleagues often tell me not to worry about it since it’s not part of my KPIs, but nevertheless I feel bad about not keeping myself on track with my aim of a post a month. Anyway, going through the posts, I came across Matt’s article about his experience learning that some new music discoveries were AI-generated, and thus I found a new topic to write about. OH, THE HORROR My experience with AI-generated music can be summed up in two incidents. Screenshot of Shiori Shinomiya's artist page on Spotify The first was when I came across an artist by the name of Shiori Shinomiya from a song recommended in Spotify’s Discover Weekly playlist. On first listen, I thought I had discovered yet another anonymous singer with the vocal chops of J-Pop stars like Ado, Minami and 9Lana. Given the way I listen to music on YouTube, I must have scrolled through the comments without looking at the thumbnail art. The second instance was when I was recommended by YouTube (I’m very sure this time) a jazz playlist. I set it as background music while I was working on something, and when I went to check the tracklist when I was done, lo and behold, there wasn’t one. Skeptical, I checked out other videos on the channel’s page, and saw that just about every video I opened had comments on the music being AI-generated. I then blocked the channel from my feed, feeling somewhat disturbed at the possibility that I had unwittingly enjoyed AI-generated music (even if the saxophone solos in a couple of tracks were distractingly good). It felt as if a Buddhist as devout as my mother had been tricked into eating beef, or perhaps a Muslim unknowingly ate pork and enjoyed it. Returning to the aforementioned Shiori Shinomiya, while I was listening to ASHITAMO, I noticed something off about the vocals. There’s this part right at the beginning of the first chorus where Shiori starts screaming, but unlike her other songs, the way her scream distorts feels less like clipping as a result of screaming too loud but more the product of digital manipulation, which I surmised could be the auto-tune not being able to handle RAW HUMAN EMOTION. At the bottom of the Spotify page, I saw that the publisher was a certain Cocoa Music Japan. Wanting to know who else this label (I presumed at the time) had, I searched them up, and another unfortunate revelation unfolded before my eyes. Across this Reddit post, everything pointed to the works of Shiori Shinomiya and her supposed labelmates being the product of AI prompts. Users pointed out the repetition of melodies and lyrics used across songs by the different “artists”, the suspicious lack of presence for these “artists” (when even underground Japanese idols have small but dedicated fanbases, and virtual idols who never show their faces have massive in-person performances), and probably the most obvious red flag I didn’t notice: AI-generated cover art and thumbnails. At this moment, I felt a tinge of despair. I’d experienced a mixture of these red flags individually, but had chalked them up to plausible reasons. This was Ado's first performance on The First Take, a YouTube channel that has brought in all sorts of musicians for "one-take" performances (a topic for a different time). It’s common for Japanese artists to remain faceless for at least the early part of their careers, with reasons ranging from age-related privacy concerns to artistic immersion and even simply separating their music from their daily lives, long before Ado, probably the biggest “faceless” Japanese artist of the 2020s, came to prominence with Usseewa. The use of AI-generated art for covers was something I would not do personally (I could take photos just fine), but could understand why a small musician would do it (lack of money to pay an actual illustrator or graphic designer), and thus far the music I’d heard from such people was good enough for me to ignore it. Not the most ethical position to take, so I guess the despair I felt upon realising the possibility of having listened to and enjoyed AI-generated music was the price to pay. RUMINATING IN DESPAIR How had I not noticed? I must have presumed only humans could make music that stirred the soul, yet here was evidence that AI could do the same. What did that mean for me as both a listener and a creator? I was now contemplating the same thoughts as many creatives who feel existentially threatened by generative AI being used for the arts. The biggest question I had for myself was: at what point in using AI is a creator no longer fit to call themselves one? In the context of music production, there have been AI-assisted tools before the AI we know today came into the fold, from mastering tools like LANDR to stem splitters like the built-in function in Apple’s Logic Pro digital audio workstation (DAW). A display of the latest Yumenokessho voice banks (exclusive to SynthV) at Bushiroad Expo 2026 in Taipei. Source: https://x.com/yumenokessho/status/2019228417787588932/photo/1 Meanwhile, vocal synths like Vocaloid and SynthV have promoted their latest products as having AI that tunes the voice banks to sound more realistic (to decent effect as I found out myself back during a hands-on demo of the Yumenokessho voice banks at Anime Festival Asia 2023). Unlike generators like Suno and Udio, which are trained on a database of other people’s music without explicit permission and require only a written prompt to generate a whole song, these vocal synths not only require users to come up with their own melodies and lyrics, but also pay the voice providers (the people whose voices were recorded to make the various voice banks). This difference in effort is one way creatives are distinguished from ordinary people typing prompts into Suno or Google’s Lyria to make up for their supposed lack of skill, but I feel that alone is insufficient in defining the lines by which we do so. Then there’s the question of the non-technical aspects like songwriting. There are already tools to generate chord progressions (though to be fair there are only so many in the Western major/minor scales), but what about higher-level elements like themes and motifs? If asking AI to generate entire lyrics is frowned upon because it’s not drawing from a real person’s experience, is it cheating if a songwriter asked their LLM of choice, “How can I write the concept of survivor’s guilt into a metaphor?” and then come up with ideas from there? The consensus on what constitutes “real” art then becomes impossible to agree on. THIS CREATIVE’S POV Current discourse on generative AI in the context of the creation of the arts is divided between “AI will make artists obsolete” and “AI will never contain the soul of human-made art”. As a consumer, while generative AI may have elicited an emotional response from me, I realise that these AI-generated or heavily AI-assisted songs are never going to be all-time favourites precisely because they haven’t got that personal touch. Be it a personal struggle, funny lyrics, or simply blowing minds with out-of-this-world sound design, everyone has their own way of connecting with others that may not resonate unanimously. In an attempt to cater to the lowest common denominator and maximise revenue (let’s be real - easy money is exactly why the AI hype train is a thing), the products of AI end up as milquetoast pieces of content (how I hate that word) to be consumed and forgotten as the masses move on to the next shiny toy. As a creative, my view of art is that it is a means of self-expression that can outlast the creator. I may not be able to make work as polished as AI, but only I could make something that’s wholly me. Every success, every mistake I make, is embedded into my work as a record of myself at that point in time. Letting someone else, let alone AI, dictate what I am in my work - to misappropriate the Hayao Miyazaki quote again, is an insult to life itself. In the Reddit post regarding Cocoa Music Japan, there was a user claiming to represent the company, and their justification for the use of AI across a few comments was that these real artists were not confident about their current abilities and thus enhanced their performances (though I still wonder if something was lost in translation from Japanese to English - if they’re even a real Japanese company at this point). Even now, I understand the sentiment of wanting to start things off right - rushing head-first into things is not common advice precisely because so few people can make it work, but at the same time I’ve come to understand that it is only through failure that we can move forward with whatever it is we’re doing. Here, I’d like to raise up some quotes from a speech that gave me some comfort amid my recovery from what I perceived to be my biggest failure a few years ago: “It is our failure to meet our perceived ideal that ultimately defines us and makes us unique.” “There is nothing more liberating than having your worst fear realised.” “Your perceived failure can become a catalyst for profound reinvention.” “But today I tell you that whether you fear it or not, disappointment will come. The beauty is that through disappointment you can gain clarity, and with clarity comes conviction and true originality.” DOES FAILURE BIRTH SUCCESS? Perhaps the fact that I still struggle with viewing failure in such a positive light might make Conan’s speech moot, but worse is getting stuck in overthinking loops because you’re afraid of making a single mistake. I will plead guilty to consulting LLMs for advice to offload my mental load, from purchases like SD cards (ironically inflated in price due to AI) to career tips and tricks, but not only has my overthinking habit not decreased in frequency, the advice I’ve taken often has led to regret. The day after I purchased a 256GB Sony TOUGH SF-M SD card for SGD$190 (the price has since increased amid Sony’s announcement that they are stopping SD card production), I found a SanDisk 128GB SD card for SGD$185 that would serve my needs better long-term because unlike the Sony card (which lets me record 5 hours of 4K 25fps footage by the way) I would be able to properly shoot 4K 120fps video with it on a future camera. This was the latest of a number of purchases where I was let down by what the LLMs had reassured me would suit my needs despite my gut telling me otherwise. Photo by Solen Feyissa on Unsplash AI gradually encroaching on my life has led me to find ways to wean myself off it for guidance, but thus far I haven’t been too successful. Between the fact that making bad purchases (or mistakes in general) is expensive, and just about every company (and governments *sigh*) is pushing for the use of AI in every facet of life, my catastrophising mind has enabled AI’s death grip over me. Perhaps that’s why I draw the line at letting AI do the actual work. Be it the videos I shoot and edit for Sgcarmart, this very long blog post (is it an essay now?) or my personal projects, I’m trying to exert some form of control over my life and make something out of it that I can call my own. Maybe what I need to do is to learn to trust my judgement and make the best decisions I can with the cards I’m dealt at any given moment, and most importantly be willing to take failure as it comes. But how do I do that? I shan’t ask Claude for once - it’s time to play Russian Roulette (capitalisation intended) and take a leap of faith. ~ Wei Feng Cover image: Photo by the blowup on Unsplash

Anyone tried it out? I saw a YouTube it can even solve excel formulation, essays...etc. https://www.bbc.com/news/technology-64538604 Quoted "It has been two months since the public launch of AI chatbot ChatGPT by the firm OpenAI - and it did not take long for people to start noticing what a game-changer this really is. Whether you have asked it to write you a song in the style of your favourite musician, sneaked in a homework question (500 words on the end of World War Two? no problem), tasked it to write copy for your company website, write a speech or even churn out specific program code, ChatGPT has proved that it can deliver - and in a convincing way. There has been acres of reporting about its potential threat to a wide range of jobs, and indeed to our entire model of education if students can get their coursework done and university applications written instantly via ChatGPT or its rivals. " Unquoted

https://sg.style.yahoo.com/quit-teaching-because-chatgpt-173713528.html I Quit Teaching Because of ChatGPT This fall is the first in nearly 20 years that I am not returning to the classroom. For most of my career, I taught writing, literature, and language, primarily to university students. I quit, in large part, because of large language models (LLMs) like ChatGPT. Virtually all experienced scholars know that writing, as historian Lynn Hunt has argued, is “not the transcription of thoughts already consciously present in [the writer’s] mind.” Rather, writing is a process closely tied to thinking. In graduate school, I spent months trying to fit pieces of my dissertation together in my mind and eventually found I could solve the puzzle only through writing. Writing is hard work. It is sometimes frightening. With the easy temptation of AI, many—possibly most—of my students were no longer willing to push through discomfort. In my most recent job, I taught academic writing to doctoral students at a technical college. My graduate students, many of whom were computer scientists, understood the mechanisms of generative AI better than I do. They recognized LLMs as unreliable research tools that hallucinate and invent citations. They acknowledged the environmental impact and ethical problems of the technology. They knew that models are trained on existing data and therefore cannot produce novel research. However, that knowledge did not stop my students from relying heavily on generative AI. Several students admitted to drafting their research in note form and asking ChatGPT to write their articles. As an experienced teacher, I am familiar with pedagogical best practices. I scaffolded assignments. I researched ways to incorporate generative AI in my lesson plans, and I designed activities to draw attention to its limitations. I reminded students that ChatGPT may alter the meaning of a text when prompted to revise, that it can yield biased and inaccurate information, that it does not generate stylistically strong writing and, for those grade-oriented students, that it does not result in A-level work. It did not matter. The students still used it. In one activity, my students drafted a paragraph in class, fed their work to ChatGPT with a revision prompt, and then compared the output with their original writing. However, these types of comparative analyses failed because most of my students were not developed enough as writers to analyze the subtleties of meaning or evaluate style. “It makes my writing look fancy,” one PhD student protested when I pointed to weaknesses in AI-revised text. My students also relied heavily on AI-powered paraphrasing tools such as Quillbot. Paraphrasing well, like drafting original research, is a process of deepening understanding. Recent high-profile examples of “duplicative language” are a reminder that paraphrasing is hard work. It is not surprising, then, that many students are tempted by AI-powered paraphrasing tools. These technologies, however, often result in inconsistent writing style, do not always help students avoid plagiarism, and allow the writer to gloss over understanding. Online paraphrasing tools are useful only when students have already developed a deep knowledge of the craft of writing. Students who outsource their writing to AI lose an opportunity to think more deeply about their research. In a recent article on art and generative AI, author Ted Chiang put it this way: “Using ChatGPT to complete assignments is like bringing a forklift into the weight room; you will never improve your cognitive fitness that way.” Chiang also notes that the hundreds of small choices we make as writers are just as important as the initial conception. Chiang is a writer of fiction, but the logic applies equally to scholarly writing. Decisions regarding syntax, vocabulary, and other elements of style imbue a text with meaning nearly as much as the underlying research. Generative AI is, in some ways, a democratizing tool. Many of my students were non-native speakers of English. Their writing frequently contained grammatical errors. Generative AI is effective at correcting grammar. However, the technology often changes vocabulary and alters meaning even when the only prompt is “fix the grammar.” My students lacked the skills to identify and correct subtle shifts in meaning. I could not convince them of the need for stylistic consistency or the need to develop voices as research writers. The problem was not recognizing AI-generated or AI-revised text. At the start of every semester, I had students write in class. With that baseline sample as a point of comparison, it was easy for me to distinguish between my students’ writing and text generated by ChatGPT. I am also familiar with AI detectors, which purport to indicate whether something has been generated by AI. These detectors, however, are faulty. AI-assisted writing is easy to identify but hard to prove. As a result, I found myself spending many hours grading writing that I knew was generated by AI. I noted where arguments were unsound. I pointed to weaknesses such as stylistic quirks that I knew to be common to ChatGPT (I noticed a sudden surge of phrases such as “delves into”). That is, I found myself spending more time giving feedback to AI than to my students. So I quit. The best educators will adapt to AI. In some ways, the changes will be positive. Teachers must move away from mechanical activities or assigning simple summaries. They will find ways to encourage students to think critically and learn that writing is a way of generating ideas, revealing contradictions, and clarifying methodologies. However, those lessons require that students be willing to sit with the temporary discomfort of not knowing. Students must learn to move forward with faith in their own cognitive abilities as they write and revise their way into clarity. With few exceptions, my students were not willing to enter those uncomfortable spaces or remain there long enough to discover the revelatory power of writing.

https://vulcanpost.com/843379/team-of-ai-bots-develops-software-in-7-minutes-instead-of-4-weeks/ Back in July, a team of researchers proved that ChatGPT is able to design a simple, producible microchip from scratch in under 100 minutes, following human instructions provided in plain English. Last month, another group — working at universities in China and the US — decided to take a step further and cut the humans out of the creative process almost completely. Instead of relying on a single chatbot providing answers to questions asked by a human, they created a team of ChatGPT 3.5-powered bots, each assuming a different role in a software agency: CEO, CTO, CPO, programmer, code reviewer, code tester, and graphics designer. Each one was briefed about its role and provided details about their behaviour and requirements for communication with other participants, e.g. “designated task and roles, communication protocols, termination criteria, and constraints.” Other than that, however, ChatDev’s — as the company was named — artificial intelligence (AI) team would have to come up with its own solutions, decide which languages to use, design the interface, test the output, and provide corrections if needed. Once ready, the researchers then fed their virtual team with specific software development tasks and measured how it would perform both on accuracy and time required to complete each of them. The dream CEO The bots were to follow an established waterfall development model, with tasks broken up between designing, coding, testing, and documenting of work done, with each of them assigned their roles throughout the process. What I found particularly interesting is the exclusion of CEO from the technical aspects of the process. His role is to provide the initial input and return for the summary, while leaving techies and designers to do their jobs in peace — quite unlike in the real world! I think many people would welcome our new overlords, who are instructed not to interfere with the job until it’s really time for them to. Just think how many conflicts could be avoided! Once the entire team was ready to go, the researchers then fed their virtual team with specific software development tasks and measured how it would perform both on accuracy and time required to complete each of them. Here’s an example of fully artificial conversation between all of the “members”: Later, followed by i.a. this exchange between the CTO and the programmer: These conversations continued at each stage before its completion and information being passed for interface design, testing, and documentation (like creating a user manual). Time is money After running 70 different tasks through this virtual AI software dev company, over 86 per cent of the produced code was executed flawlessly. The remaining about 14 per cent faced hiccups due to broken external dependencies and limitations of ChatGPT’s API — so, it was not a flaw of the methodology itself. The longest time it took to complete a single task was measured at 1030 seconds, so a little over 17 minutes — with an average of just six minutes and 49 seconds across all tasks. This, perhaps, is not all that telling yet. After all, there are many tasks, big and small, in software development, so the researchers put their findings in context: “On average, the development of small-sized software and interfaces using CHATDEV took 409.84 seconds, less than seven minutes. In comparison, traditional custom software development cycles, even within agile software development methods, typically require two to four weeks, or even several months per cycle.” At the very least, then, this approach could shave off weeks of typical development time — and we are only at the very beginning of the revolution, with still not very sophisticated AI bots (and this wasn’t even the latest version of ChatGPT). And if time wasn’t enough of a saving, the basic costs of running each cycle with AI is just… $1. A dollar. Even if we factor in the necessary setup and input information provided by humans, this approach still provides an opportunity for massive savings. Goodbye programmers? Perhaps soon, but not yet. Even the authors of the paper admit that even though the output produced by the bots was most often functional, it wasn’t always exactly what was expected (though it happens to humans too — just think of all the times you did exactly what the client asked and they were still furious). They also recognised that AI itself may exhibit certain biases, and different settings it was deployed with were able to dramatically change output, in extreme cases rendering it unusable. In other words, setting the bots up correctly is a prerequisite to success. At least today. So, for the time being, I think we’re going to see a rapid rise in human-AI cooperation rather than outright replacement. However, it’s also difficult to escape the impression that through it we will be raising our successors and, in not so distant future, humans will be limited to only setting goals for AI to accomplish, while mastering programming languages will be akin to learning Latin.