So Substack just dropped an unwanted feature on us: “Going forward, you’ll be able to scan text to see how much of it was likely written by hand or with AI assistance,” says Chris Best in Against Claudefishing.

Well thanks, Chris. Didn’t realise I needed to be harassed even moar by anti-AI trolls. Trolls, whom, I question their critical thinking skills day by day because you totally hate people using AI but you’re using AI to hate on people using AI?

Hilarious.

Anyway, if you haven’t gotten the memo, despite assertions that they’re much better now, AI “detection” tools are notoriously unreliable.

I had a conversation with my lecturer friend yesterday and I asked her if she ever had false positives/negatives with her students. Without hesitation, she said: “Yes. Definitely!”

Fortunately, she belongs to a college where students are not automatically penalised just because an AI tool says they wrote with AI. Her students are interviewed about their essay and asked to defend their thesis. I assume many can produce past versions of their essays as proof of their work as well.

AI tools also flag English as Second Language writers more. Neurodivergent writers, those with dyslexia, those with brain fog, will be swept up too as many use LLMs to order their chaotic thoughts and communicate better.

On a personal note, I tested an essay with Pangram. It came back 100% human, but ironically, the shorter, summarised version of the essay was flagged as 60% AI.

100% of the shorter text was derived from the original essay. I used AI to edit it to 50% of its original size, giving the LLM (Deepseek) firm instructions NOT to add any new copy but just cut down the text and rearrange paragraphs. After it shortened the prose, I further edited it manually.

So, that piece of copy was dinged because somehow and AI left markers on it?

And how is that shorter version not 100% my own work when it’s made up of entirely my words?

I don’t care what Substack does with the platform

Add Pangram. Add baby dancing emojis with every “human verified, Pangram-approved” copy if you like. What I care about is my mental peace on the platform.

As a minority voice on Substack, I’m already getting harassed and bullied just because I don’t have the same opinions as the majority on the platform. I’m getting tired of being dragged into exhausting political and cultural battles when I express my often-contrary opinions on AI and China – hot topics for these people.

Before Notes, Substack’s annoying social media organ, my newsletter was a self-contained home where only invited ones or those who deign to follow me, could chatter in peace. After Notes, my essay can be picked up, shared and exposed to people I don’t want: those steeped in hatred, bigots, racists etc etc.

Case in point, yesterday I wrote a Note thanking two Substack writers for elightening me On the matter of Greek erasure in Christopher Nolan’s The Odyssey. Someone reshared my post and called me a racist for writing that essay. (The reshare or classically, the “retweet”, is an oft-used device to bully other people online.)

Since Substack introduced social media dynamics to the platform, my mental peace has been eroding day by day.

Instead of siccing unwanted features like Pangram on us, Substack should be creating features that ensure that my Substack is a safe place for discussions, especially from people who are minorities on this platform.

On Mastodon, I can choose to “downplay” my Note so it won’t go viral, or only share my note to specific people or followers. I can also choose to block certain keywords so that I don’t see hateful posts. These guardrails don’t exist on Substack, and the bullying is getting worse each day.

With Substack’s introduction of Pangram, I expect it to get even worse.

However, I doubt Substack would care because outrage is engagement, engagement is money, and there’s no better engagement than outrage.

I’m glad I made a shift in my content strategy with Substack

A few essays ago, I implored writers to have a personal website and not make Substack their digital home. I’m implementing this advice personally by writing my essays first on my website and then sharing a bite-sized version of it on Substack (while inviting them to read the full essay on my website at the same time).

PS: Interestingly, as if to proof my point right, wordpress.com actually shared my post on their Reader portal, and I attracted over 50+ comments overnight. If I had written it on Substack, it probably wouldn’t have caught their attention. All that viral juice? It comes back to my website, baby!

Will I stop using Substack? Like I said in the past, the puritanical pursuit of platform purity is a futile and silly exercise.

I’ll continue my current strategy of writing my essays on my website and promoting them on Substack. Substack will be my “community hub” and publicity engine to get new readers for the website.

I’ll also probably hang out more in Seablings, the Southeast-Asian community I’m building on Substack. I’m enjoying the little forum-like environment created by the weekly Seablings Soiree posts.

As much as I desire to disengage from the toxicity of Notes, it is still a valuable resource to find new voices and connect with like-minded writers. Unfortunately, blocking is the only weapon we have against the toxic elements on that platform. The best thing I can do is limit the time I spend on that increasingly toxic place.