After nearly 20 years of living in Taiwan, I’ve reached the point where being so useless in Chinese has become a personal failure. I reached lower intermediate level, but then plateaued. 3 years in, my Chinese was pretty great. 10 years in, it was unbelievably bad for the amount of time living in a foreign country.
At a dinner recently when a colleague asked how my Chinese was, my wife said “embarrassing” and she’s not wrong. For some reason – probably because I’m procrastinating against other big projects and mastering Chinese feels like a productive distraction – I’m newly committed to improving.
PS – this wrote this post over several months, and it was mostly an assessment of the challenges, which I feel I’ve now conquered… scroll down to see how I plan to become fluent in Mandarin this year.
The challenges of language learning (study plan)
I’d already been consuming some YouTube videos – I’m in that awkward stage where I can understand a lot but can hardly say anything; and I can’t read, although the characters I learned a long time ago have stuck with me, so I know I could learn the characters if I wanted; Taiwan uses traditional Chinese characters which have a lot more detail.
I actually got kicked out of university because the teacher’s lesson planning involved students writing characters on the board with a marker. I could memorize, read and even type out characters on my phone or a computer… but writing traditional characters from memory by hand would be a lifelong and mostly useless pursuit (I don’t even write in English anymore, I can’t read my own handwriting and pens and paper feel weird).
Anyway, I got a refund and haven’t been back to that university since… but I’m finally ready to try again. This time however, I’ll focus on comprehension before characters.
300 hours of video games = listening practice?
This summer I got into a fantasy videogame; it’s a rare game in that I can keep playing over and over again, in 40 minute sessions, and it never really gets boring. There have been some studies that keeping your mind semi-occupied with one task my let your brain learn in the background, so I found a bunch of videos with longish phrases to just listen to while I play.
I also went through videos for HSK 3 and 4, which is about my level.
Did it work? Not yet.
The hypothesis was, if I can easily sit down and play 2 or 3 hours of video games a day, and if I could also force myself to learn Chinese at the same time, that would be time well spent. And I am, actually putting in the time. I think the main issue is, I need to pick ONE video on repeat instead of trying to learn new things.
Could I just watch a video 3 times with full focus, instead of listening on repeat 10 times while playing video games? Yes… it would take less time but more brainpower; the main thing is that, even if objectively, obviously, it’s less time and effort to do it that way… if I never do it that way it doesn’t matter.
Doing it the dumb, slow way is better, if I actually do it.
Ps. I actually think an article like “I learned Chinese in 3 months with video games and magic mushrooms” would be an amazing piece of content.
Mandarin Fluency requires speaking
The next next big thing; after my comprehension is pretty decent, is to put myself in live situations where I’m forced to respond. I see some titokers just show up and rizz locals in Chinese but I don’t have that kind of confidence. When I was little, my parents thought I was slow because I wouldn’t speak… until I finally did and spoke in full sentences.
I don’t talk much anyway, and only when I can be reasonably clear and express valuable meaning.
But fluency is obviously the goal, and that’s not something I can learn just by consuming content or studying, I’m going to have to find ways to put myself into activities or groups of native speakers, and that’s a challenge for later (solved it: I built an AI Mandarin tutor to practice with – more on that later.)
It’s also very very frustrating to feel like I’m at level 6 comprehension but level two speaking. Even very, very simple and basic sentences, I have trouble saying or speaking out loud, my tongue and mouth just don’t want to move in the ways that Chinese requires. So I can’t JUST train my brain, I have to actively exercise my tongue and mouth muscles in unfamiliar ways; I need to speak and voice out loud all the very simple stuff and repeat it until it becomes comfortable.
I don’t like doing it that way because I was a “talented and gifted” student so I feel like I can cram last minute and ace the test, with less work than everybody else – but like most aging TAG students who have failed to live up to their potential, I’m finding my special advantages are also somewhat limitations (when I first arrived in Taiwan, I was in class with a friend and I thought he was so slow and bad and pronunciation – but he actively used it worked hard and became fluent years ago, leaving me behind).
Fluency isn’t just knowledge. It’s a reflex – a muscle memory.
Chinese Language Learning Tools (best apps)
My sudden obsession with learning Chinese is mostly an extremely cultivated form executive dysfunction; it is the activity I practice most (not doing the thing I need to be doing) so naturally I’m the best at it.
I had a lot of big goals and the end of 2025, was feeling pressure to complete them, and instead used Chinese as a form of revenge procrastination that still felt productive.
While getting organized, I discovered this post I wrote a decade ago on how to learn Chinese in 3 months – which is an indictment and a reproach. Friends are skeptical of my video game method, but I am feeling signs of improvement. I’ve had a couple of dreams in Chinese, which is a good sign.
I learned Spanish by living in Argentina for a year – absolutely full immersion, before smart phones or the internet. I like to believe I’m clever; learning a new language should be possible.
Last year a YouTuber held a gameshow-contest full of foreigners in Taiwan speaking Chinese and that was super motivational, because I’m competitive. However… I haven’t progressed with all the books I bought at ALL. I have been listening to grammar and vocab videos, from HSK4~HSK6. Some of it feels too hard, but I’m absorbing a lot of vocab.
I was using the flashcard apps for a few days but it became overwhelming, with several decks it was hundreds of cards a day. There are some listening ones that I still think are very USEFUL, but I just don’t have the PATIENCE or willpower (or time-awareness?) to be consistent or form new habits.
I’ve been watching more advanced YouTube videos – the upper intermediate ones mostly don’t use English subtitles which is annoying, but some of the good blogs at least define and explain the main vocab words.
And I’ve also been watching a lot of native content on Netflix.
Previously, when I tried this, I couldn’t follow.
I’d excuse myself by saying, I watch TV to relax. Or, there’s no Chinese-language content I enjoy (I really like supernatural/fantasy stuff, and it’s more drama or comedy.) There is a some pretty great Korean or Japanese shows though and some have Chinese audio.
While previously I couldn’t get into any shows… this time I’m recognizing all the new grammar and vocab I’ve been learning which feels important. I can recognize new words, and that confirms and solidifies my expanding knowledge. I still need more WORDS, but I can already tell my understanding has gone up, I can follow a lot of sentences I wouldn’t have been able to before.
And that’s honestly a big step, because now I can slowly switch to full native Mandarin immersive content – instead of “studying” I can just binge watch 8 hours of some show. I am definitely going to run out of content; there’s a lot of classic Chinese battle or historical shows, and I’m slowly going over things that are good enough but not really amazing… but it’s such a dopamine boost to hear a sentence and know the words and understand the meaning, and to realize that about yourself, that you have unlocked this new channel to comprehension, the path to knowledge is an eternal lust.
I Got Frustrated With Language Learning Apps — So I Finally Built a System I’ll Actually Use
Despite my earnest ambition, which I’m happy to report has not waned, I mostly just kept playing video games and listening to Chinese podcasts on YouTube. The thing is, there’s some really really good content. I didn’t stick with any of the apps, so honestly I’m not even going to mention them here (I’ll make a review video and post that instead).
I took notes when I heard new words, and I have definitely improved my comprehension and vocabulary, but it feels too slow. I have a stack of books but they’re intermediate, which means characters only… no pinyin. So I’d need a language tutor, or to take pics and translate them with my phone – I even bought a new pair of Rokid glasses to scan and translate, hoping they’d also read out loud to me (they won’t).
The books, the paper, the effort… it’s exhausting just to have them in the room. I’m a fully digital citizen and a mouse and keyboard feel more natural than a pen and notepad. But I needed a specific, actionable plan of attack, something I can commit to and track progress, something easy and organized.
So I built my own. I don’t know if it’s the best thing ever for learning Chinese, but most other methods haven’t clicked for me and so I’m going to focus on this one platform hard and see what happens.
Mostly, it’s a Chinese flashcard system like Anki, with an emphasis on daily practice. Each character has a breakdown of the radicals and a mnemonic to boost recall, along with 3 examples sentences. You can play anything out loud, you can hide the pinyin or English, and it works for simplified or traditional characters.
You can also immediately practice and review with small quizzes and tests, or chatting with an AI tutor who is always up to speed on your progress – even voice mode so you can have a real conversation!
I put it up online just so I could access it easier – it’s free, and I like the idea of people studying together, so I hope it’s useful. You can play around with it here – you don’t even have to login: Chinesetruffle.com
PS. I’ll have to find a way to set limits for the expensive stuff like voice chats.
Learning Mandarin Channels on YouTube
I’ve been watching a lot of “comprehensible input” content like Chinese podcasts or vloggers – I’ve found a satisfying supply of great content and eventually I’ll make a big list to support those channels. But I suddenly became obsessed with the idea of my own channel, based around the idea of an “illiterate doctor” because I can’t read Chinese (yet).
This is uncomfortable, because I go pretty much under the radar here in Taiwan. But that also means I don’t get to practice… so hypothetically, getting a little bit local famous could help me make new friends to speak with.
It’s basically a reverse of the good content I enjoy on YouTube – me speaking English with Chinese subtitles. More importantly, putting it out there makes it inevitable.
I’ve built a system to help me study, but I needed the added pressure of public opinion.
So I’ll make videos chronicling the whole process, and the world (more likely my 3 subscribers) will hold me accountable. Eventually I’ll speak some Chinese, and when I’m fully fluent I can just be a famous Taiwan YouTuber or something.
But because I can’t do anything normal…
I’m adding a twist.
My digital AI clone is the perfect language tutor…
A few months ago I played around with Sora – it could make me rap in Mandarin walking down the streets of Taiwan. It was sickening… I could probably flood TikTok with AI slop and people would think my Mandarin was amazing – then if I did get famous it would only be the fake AI version of me.
But I do want to play with this tech in a responsible way, so I’m going to take the English transcripts of my videos, translate those into Chinese, and then have a virtual me speaking my ideas, fluently in my voice. He’ll be way more handsome and much smarter.
Probably has all his shit together.
I’ll have very big WARNINGS on those videos to make sure people understand it’s an AI clone and not actually me, and that the whole thing is an experiment in language acquisition (I suppose, as a man of letters with a PhD in Literature, I’ll be granted this indulgence).
I’ll record my Chinese learning journey, and then watch my virtual clone unleash the same discourse in fluent Mandarin – so I can see exactly how it would look and sound coming out of my own mouth.
I’ll say all the things I have to say about all the things I like to talk about.
I’ll translate those for my clone…
and then I’ll listen to myself read them to me.
Who is to say that this isn’t the fastest way to learn a language?
Has anybody else tested it? I’m probably the first.
I love how it’s so dystopian and scifi – I could literally write a novel about all this.
Are we in the Matrix yet? Should I follow the white rabbit?
PS I WISH I could get my videos to look as good as my digital clone, but I haven’t figured out the studio lighting. There’s a danger of chasing an impossible ideal; feeling discouraged because AI has surpassed you. But I’m sending him down trails I’d dare not tread on my own.
Did AI Just Take my Job?
All the things I’ve been doing online over the past few years, can now be done with AI. But it can only be done well if the person in charge has the experience and knowledge to recognize great writing from slop. The thing is, I’m not very good at my job.
That is to say, I’m excellent at what I do… I just refuse to do it.
I am Melville’s Bartleby.
AI is letting me do a lot of things all at once, which means I can finally harness the rush of enthusiasm that comes with shiny new ideas, and actually build some shit before I collapse on my fainting couch for three days.
PS it’s “illiterate professor” because I can’t read Chinese… yet.
When do the dreams become real?
What if this better-looking personal avatar, who kind of looks and sounds like me in a recognizable way, could produce better content that people actively preferred. I could just ramble and make him lip-synch; or I could turn my eternaldumpsterfire™ of brain farts into clear, informative, intriguing videos that are more useful and more satisfying to actual humans?
What if instead of posting a few times a year, it was a more regular occurrence. A hermit, sending out a robotic simulacrum to greet the public. Not to pander, but to provide – to participate and prop up? He would be like casting a magic spell, assisting thousands… how could I not?
We are Bruce Willis in a future where self-isolated humans interact with the world through remote-controlled robotic surrogates. If my AI clone is my ideal self, a magic genie with infinite abilities who will follow my command, I can give up the struggle of production and allow myself to be served.
Is the struggle for skill acquisition the thing that gives art value? Surely not, since modernist painters refused “skill” as a lecherous and intrusive cage. Is it feeling? The last two years have easily proven that most humans like and share AI content (without knowing what it is, the content engages users on its own merits.)
One insight I’ve had about AI, is that things manifested into reality have value, even if they’re not perfect yet. The cost of art is too great, so great art is few.
While the quality of the thing itself may not rival the thing I would have made, you can’t compare qualities, only ontologies (I just had a discussion with chatGPT about this word, and I was right – I doubted myself enough to check, but it took a circular conversation before he confirmed my definition).
What I mean is, a thing made with AI is better than a thing unmade by you.
Not ALL things, for sure. You’re encouraged to claim as much joy as you’re entitled.
But for all those other things you’ve been avoiding, that wouldn’t ever get finished at all, that you’d maybe even already given up on… there is fresh hope.
The raw creative potential of the void.
That’s why if you gaze into the abyss, the abyss gazes back. You want to look into “not yet but could still be.” The dazzling potential of all possibilities. One will catch your eye and consume your soul.
AI can very quickly help you realize imperfect things – and sometimes imperfect is all you need, because not all projects require the full extent of your brilliance.
Language Mastery vs. The Future of Creative Work
I’ve basically given up on most of my online channels – I’m simply not making any content and don’t currently feel inclined to. I feel like I’ve said most of what I had to say. Of course I didn’t do it well. I could do it better. But to what end?
Over the last few months I’ve actually been winding down – I spend more than I make with online fees and tools, so I’m letting go of some unprofitable projects and shifting my attention (theoretically) to the things people actually want to pay me for.
Rather than trying to build the perfect home YouTube studio and manually make all those videos… I could have a virtual me do it instead. The question is of course, whether or not viewers would like it more or less than my organic self. I could see it going either way because honestly, I’m hard to listen to sometimes.
But I like the idea of an open partnership, where every body is in on the joke, and I let the virtual me do the deep organized dives so I can do the crazy rantings.
AI removes a lot of the burden or productivity, which means I can stay focused longer – there are tasks I’ve hired out before that have never been accomplished, wrestling with human service providers. Skills I didn’t have. So I have graveyard of Big Ideas in piles of folders containing folders, spiraling back in time.
Chinese has been one of those elusive abilities, like playing the piano, that my ideal self would someday acquire; and yet as the sand falls, the dream fades.
Replacing Procrastination with Productivity
I’ve become very good at doing something worthwhile to avoid the thing I promised myself I would do. I spiral out in a tantrum of manic productivity when I’m avoiding projects that matter.
And while I haven’t set New Year’s Goals for a while, it’s mostly because my goals are always the same. Finish my fiction.
I have countless series and story ideas, and I am broken under the weight of unattainable goals. I used to be young and eager but now I’ve seen what a decade can do. I can see the math – I’ll never finish my best work in this lifetime.
So I’m certainly optimistic about having new tools to help me write or edit drafts, but skeptical about how much work it’ll take to fix and whether it’s worth the trouble (someday soon, it will take less work, and may be worth the trouble).
I started January with somewhat of a clean plate and could hypothetically spend the next few months focusing solely on writing more novels – and that’s certainty what I want to do and plan to do – but instead I built both a Chinese learning app and a brand new YouTube channel… things that weren’t really on my immediate radar.
I listen to the muse and go where she calls me. I channel an energy beyond me.
On the other hand, my Chinese sucks and I’m no longer willing to live with the indignity.
I’m a hamster on a wheel, running for his life in the wrong direction.
But without the essential exercise, the vessel will grow stiff.
I imagine through this detour, I’m gaining some very valuable new skills, which may soon be wielded in my favor. It’s the accidental mishap by which your protagonist snares the crucial key.
I feel shifting winds and am ready to chase them into the unknown.
I’ve already lost confidence in pursuing this adventure, but the app is all working.
It’ll help me, and maybe it’ll help others.
The emotional burden of creative toil
I’m well aware of my procrastiviticlivities™ (the tendency to procrastinate aggressively yet productively in the wrong direction) but I also feel inspired. I’m using AI to fix or finish a decade’s worth of creative baggage – the imperfect, unfinished things that shared so much of our time, heart pieces stuffed in abandoned folders.
Often, I was thwarted by simple technical issues, and once AI taught me exactly what to do, I learned so much that it became easier and easier. Things that seemed insurmountable, which would have drained my battery, could be faced with impunity.
I have courage, more than I had before, because I know the limits of what I am capable of just got wider. In a strange way, I’m vision-boarding my ideal self, with an AI clone that has already achieved my desired path; he’s been through it, he knows what to do, he can teach me – or at least, he can remove trivially stubborn obstacles that would halt my progress.
While I get that highly polished digital AI avatars are weird if you meet someone in real life, the good news is to say I almost never meet people. Does my online presence and platform have to be me exactly as I am, rather than what I aspire to become?
The wild jungle adventure is being formalized. I’m downloading Plato’s Realm of Forms.
I’m ticking off a list that has been gathering dust. I’m clearing the path, reclaiming my own energy and attention and focus, cutting ties not by giving up but by taking control. Reducing the mental bandwidth of cancerous side-projects, so I can pour all my passion into my favorite things, is a Faustian bargain I’m ready to make. Otherwise, I am my art’s existential threat.
I feel like I’m drawing an arrow, we’ll see where it lands.
The challenges of learning a new language
We were talking about Learning Chinese, remember? It just kind of spawned into a treatise but it’s been awhile since I’ve written one of those.
I started this post in October and immediately stalled out. Down below is a list of my early efforts; the books and apps didn’t get used at all, while the video game method was faithfully executed. But the rankling discomfort of unsatisfied ambitions – paired with the brand new abilities AI allowed me to access – prompted a frantic week of creative abandon (yes, it does still feel creative, when you’re coding. It’s a lengthy and frustrating process, but the thrill of making something new never gets old).
And now, hopefully, I’ve resolved my obstacle by devising a solution. Even tonight I had to convert my *first ever* video for that new channel from CapCut’s simplified Chinese to the traditional characters that Taiwan uses. It felt impossible, and I could have easily lost the day – and possibly moved onto something else – but instead chatGPT worked me through options and Cursor actually built me a tiny desktop application that will automatically translate my captions.
An idea made manifest.
I can’t take the credit for it, I don’t even know how it works – but I don’t need to know. That specific bit of coding trivia is not a useful thing to study; it’s a single nail in a thousand miles of train tracks.
I had to invent the tools and support I needed, because the plethora of Chinese language learning resources failed to provide me with the energy I needed to deploy them. But my months (years?) of frustration, also helped me figure out exactly what I needed.
Can AI help you learn languages faster? (Let’s find out!)
Mandarin Learning Books (these didn’t work for me)
I went out and bought a stack of Chinese textbooks, because that’s what you do when you’re trying to be a serious person. Then I remembered: textbooks are where motivation goes to die.
Practical Audio Visual Chinese (PAVC) is the main offender. It’s old. The videos are ancient—full-on 80s energy, low quality, mumbling audio—and the most irritating part is my book doesn’t match the YouTube videos I found, so I can’t even “listen along” without getting derailed. Going through it feels like dragging a sofa up stairs.
My original plan was simple: translate the characters into pinyin, scribble it in, say it under my breath, grind forward. Except you can’t really “sound out” Chinese characters. If you don’t know them, you don’t know them. So a book that already includes pinyin helps a lot—like Let’s Learn Mandarin, which at least doesn’t make me feel like I’m doing homework in a broken museum exhibit.
I also picked up A Course in Contemporary Chinese, which seems to be the newer replacement for PAVC (I think). I’m not totally sure—that’s just what the internet claims. I’ve got Far East Everyday Chinese too. Some of these come with CDs, which is funny because I don’t own a CD player. I might literally go buy one, like it’s 2004 and I’m rebuilding my personality.
The problem is: forcing myself through textbooks one lesson at a time is tedious. I like the ritual—walk to a coffee shop, sit down with a book, pretend I’m disciplined—but it’s slow and frustrating. Also, I forgot my reading glasses one day and realized part of my “burnout” is just me squinting at paper like an idiot.
I’ve been using ChatGPT to bridge the worst part: I take pictures of dialogues and grammar pages, it converts them to pinyin and translates them, and then I can annotate the book. That feels like progress… but it’s exhausting, and it’s the kind of exhausting that makes you quit. Especially because textbooks love vocab I will never use in real life, which makes the whole thing feel pointless.
The real problem isn’t grammar—it’s vocabulary. My grammar is fine. What I need is a ridiculous amount of words. That’s the bottleneck. If I knew more words, everything else would be easier: textbooks, listening, speaking, even just staying motivated. Right now I’m stuck in that classic intermediate swamp where everyone assumes you can read characters fluently and magically transition into “all Chinese, no help.”
So I’m shifting toward vocab-first tools that I’ll actually stick to.
- There are a couple YouTube videos with ~1000 words each. If I just looped those consistently, I’d pick up 2000+ words (and maybe some characters, depending on whether I’m watching or only listening).
- I keep returning to a flashcard website based on A Course in Contemporary Chinese. It works great on desktop, but on phone/tablet it’s a “sit down and commit” thing—you can’t half-ass it in micro-moments. One lesson done properly takes like 20 minutes, and that’s probably the point.
- I never fully “got” Anki, but I’m close now. I imported a bunch of decks—deleted the ones that were messy, simplified-only, or missing audio. But I found at least one that’s solid. The best part is the default behavior: show up each day, swipe through ~20 new items, and it tracks everything. That feels smooth and satisfying, which is basically the only real secret to building a habit.
- I also found an app called Immersive Chinese, and honestly I’m impressed. About 160 lessons from beginner to intermediate, mostly one longer sentence at a time, which I prefer. It’s not a chore. It’s the kind of thing I can actually do daily in a coffee shop without wanting to throw my phone into the harbor. My only complaint is I want way more content. Part of me wants to clone the whole app so I can feed it my own sentences.
That’s basically the theme here: the fast, slow way works—but only if you do it. I’ve been failing at this for a long time, so I need a system that I’ll actually stick with, not the one that looks impressive on a shelf.
3-month goal: 2500 new words/characters.
I’m focusing on vocab because it’s the lever that moves everything else. Most study material assumes you understand about 80% of what you’re reading or hearing. I’m not there yet. Once I am, textbooks and listening practice won’t feel like punishment.
Textbooks are great in theory because they’re structured and progressive, but I get furious learning phrases I’ll never say. So I want structure without wasting my attention.
Using ChatGPT as a language partner is a weird one. It’s useful, but I’m not fully comfortable with it yet—especially voice mode. Still, it can build vocab lists, structured lesson plans, and custom dialogues. The thing I actually want to do is more personal: make a list of topics I love talking about, record myself rambling in English, translate it into Chinese, then practice repeating the exact same content. That way I’m learning language I’ll actually use: my own thoughts, my own interests, the fun stuff.
I’ve also thought about learning through songs, because it’s smart in theory—but most lyric videos don’t include pinyin, and right now characters are my weak point. There are also good YouTube channels doing conversational Chinese, but they often skip pinyin and English translations because they assume “Book 3 = you’re ready for full immersion.” I’m not. Not yet.
Also: I hate podcasts. The jingles and intros make me want to die.
Daily practice plan (about 2 hours/day, flexible):
- One textbook lesson (vocab + listen + understand) — 1 hour
- Anki reviews + new cards — 20 minutes
- One Immersive Chinese lesson — 20 minutes
- HSK 4/5/6 YouTube study — 30 minutes
- Optional: games + looping vocab/sentence videos (repeat the same batch until it sticks)
My textbooks have ~75 chapters. One a day (weekends off) is about 3 months—except it feels painfully slow, and a lot of it is repetition. It would be faster if I cram vocab first.
At 20 Anki words/day, that’s about 600/month, ~1800 in 3 months. Add in the 2000-word YouTube loops and I’m close to the 2500 goal. The bottleneck is attention: I need something I can do while doing something else. Smart glasses might actually help here—especially if I can run vocab videos while walking or biking, and even get real-time translation overlays for daily life.
And then there’s the next step: speaking. That’s its own problem. I need to talk more, consistently, ideally with a native speaker around specific topics. But I’m an introvert and I barely speak to people in English, so yeah… we’ll cross that bridge after I stop feeling like every conversation is a memory test.
Bottom line: I’m choosing the thing I’ll actually do. A finished system beats a perfect system I never use.
The Actual (New) Chinese Learning Plan
It’s pretty similar, only now I have a system and a process that’s AI-enabled, at least in the sense that AI built me the structure I was missing. I want to learn all the vocab words up to HSK6. So I added all the HSK levels into my Chinese Flashcard app, and added goals to track progress and completion.
I’ll do one book a month for six months, using the built-in tools to focus on listening, shadowing and speaking for reinforced learning. The insightful former me up above knowingly predicted “The bottleneck is attention.”
Building a Learn Chinese Youtube channel turns it into an activity, rather than a chore. I can treat it as a fun crazy experiment and play with some new AI tools – exploring the fastest and easiest way for me to finally get fluent and connecting with people through the process; investigating the synergies between human creativity and machine learning.
I feel like the bionic man, sustaining long bouts of exertion through limb-supporting safety guardrails. I have built myself a wheel to run in. But building the wheel was the hardest part and the momentum itself will generate continued motion.
AI writing and communication: better than humans?
You may have noticed a tone shift up above. I asked chatGPT to clean up the messy notes I’d been making when I started to explore Chinese learning materials. I *could* have trained him in my exact style, the problem is, my exact style is misspelling every word and then never finishing sentences. I’m not entirely convinced that AI can’t already write much better than I can, if we define better as ‘external quality validation’.
The writing is a bit too punchy, but it’s trained on “good” writing, or maybe “effective” writing. It’s probably more engaging, smoother and easier to read. It even created some absolutely bangers, I think “serious person” is mine but “textbooks are where motivation goes to die” is a sharp punchline and a powerful hook. Also later when I said “I hate podcasts,” chatGPT added, “the jingles and intros make me want to die.”
If you’re familiar with AI you can probably see the signs. But again this is raw, based on my original but without a trained writing style – the low effort article slop you’re seeing everywhere is just the easiest, laziest shit from the worst free models… and even that is now consistently better than the easiest laziest shit from the worst humans.
I don’t believe my imperfect mess is necessarily the thing that gives my writing value, especially for blog posts like this one where it is an absolute goblin. I’ve been using AI to write articles on some secret blogs and those articles are honestly stunning, they sound like what I could be at my very best, if I tried very hard, and had unlimited time and resources – but why invest that much in some random article on an invisible blog?
I have a very strong suspicion – which worrying, because it’s also so obvious – that well written, informative, SEO’d blog content would not only bring more traffic but also help more people; and I’m also convinced AI could do it better than I could (or, as I mentioned earlier, that AI could do it at all, because I absolutely will not.)
The sanitized version of my notes has a very distinct style. It’s not bad. Maybe it’s even good. Maybe it’s better. It obviously doesn’t sound like me, but holy shit that’s the point, and I hope everyone scrolled past the meandering life update to get what they needed.
UncannyValley SideNote: I just had the unnerving realization, while editing, that I’d gotten stuck into a pattern loop, ending all of these sections in a pithy, vaguely poetic structure – this is one of the things that AI constantly does, that I have to vigilantly fight against (repetition and getting stuck in loops of purple prose without actually saying anything.) Which sucks becauseI felt those sentences. They resonated with me, I created them deliberately to my own satisfaction. Am I becoming the AI? By analyzing bad AI writing have a become a better editor, or is my AI companion already softening my brain?
Derek Murphy has a PhD in Literature, is a USA Today and Wall Street Journal bestselling author, and has spent over a decade helping authors write, publish, and sell books through Creativindie. When he’s not critiquing manuscripts or building writing tools, he’s probably in a castle somewhere with cats.
Add Comment