Thoughts on AI
October 20, 2026
I took a stroll around Philly Chinatown yesterday and I noticed something: many of the displays, menu images, and visuals are AI-generated. In theory, coming from a graphic design background, I should have had a visceral aversion to this phenomenon—but what I actually felt was akin to an indifference.
From a business standpoint, it makes sense: It’s cheaper to use AI to generate printable assets than to hire a graphic designer, which would easily cost 10-20x more than a one-month ChatGPT/Claude/Gemini subscription. Once anything becomes commoditized, people would start to believe everything (skills included) can be replaceable with a lower cost substitute. What I see are sloppy color choices, nonsensical images, and sloppy typesetting; what the restaurant owner sees is something cheap and practical.
It’s like how, with my plebian palate, I can’t bring myself appreciate fine dining as much as I do with dishes from an average noodle house.
And that’s when it suddenly dawned on me: what I’m actually worried about with AI’s disruption of white-collar work isn’t just the impact on employment, but that people without the relevant expertise or training fundamentally can’t tell what’s good from what’s bad.
I see this a lot in the academic field of HCI. There are lots of people who are technically competent, but never properly trained in English writing. So, they assume that with enough prompting, they can simply will the AI to generate a semblance of a manuscript… only for the submitted paper to land in the hands of a reviewer like me (read: an English pedant), and be rejected momentarily after. All this does is wasting everyone time and energy. And as much as I want to empathize with non-native speakers who have to write in English, I can only say that this is just a part of the process. It’s not like I grew up an English-speaking environment either—I have had to put in my fair share of practicing and learning the language. (Though, I do admit that being a linguaphile does the process make easier.)
This is also why I think statements like “as long as it works, isn’t that good enough?” is a terrible notion. Anyone with a basic understanding of statistics knows that accepting mediocrity as the default inevitably lowers the average.
Much like just because I can’t appreciate fine dining, it doesn’t mean I think fine dining establishments should all be turned into family restaurants.
It’s very interesting to hear from people outside my circle, because while my friends and I are concerned, the specific things we’re concerned about are more granular—like how it’s been more and more challenging to grade papers, teach students, do “slow” AI research, design/make things responsibly, etc. Like, what we see are the different views of the cogs in the machine… What people see outside of our circle is the hulking machine that’s towering over them.
It makes sense that AI leaders’ sentiments are striking fear into the public’s mind, since they are all over the news due to how influential AI as a technology has become, and to some extent I can see where these AI people are coming from. It’s not my main area of research, but I’ve seen enough of how the sausage is made, and I share a similar sense of dread.
But it’s undeniable that the AI/tech leaders have strong incentives to amplify and keep this narrative alive, because there’s so, so much for them to gain—whether it’s from an economic or political standpoint, if they play the game right, it’s probably the most leverage they can use to concentrate power and wealth.
Personally, these are the conclusions I came to. Take X as where we are currently. AI leaders’ techno-determinist vision paints the future as:
X → some years later → end of human civilization or whatever catastrophic event
And I just don’t see how they connect, because that’s not how reality works. The whole world (8.3 billion people with very different values and priorities) doesn’t just set itself down a path of destruction and accelerate towards it.
What’s more likely to happen, if I have to guess based on looking at major events like the Great Depression of the 1930s, the 2008 Great Recession, COVID, etc., is that things will suck for a number of years, and that’s going to be tough to live through (we’re going through it right now), but things will correct themselves in due time. So my expectation of the chain of events is more like:
X → things will suck for some years but won’t be world-ending → global society gradually figures out how to cope with AI → AI eventually becomes stabilized as a natural part of everyday life → back to waiting for the “next big thing”
There’s a tendency for the collective consciousness to gravitate towards a catastrophic, world-ending event, because for some reason, when we think of something big that goes south, we imagine the worst-case scenario—just like when we get a diagnosis that there’s an unidentified mass growing inside our body, our first instinct is to jump to “oh god no it must be cancer”, when in reality about 80% of diagnosed lumps are some sort of benign tumor. That is not to say a catastrophic event will never happen, but if you’re really worried about a civilization-ending dystopia, you’ll have much better odds betting on a slow, boring decline that will surely disappoint the cult members who are earnestly waiting for judgment day.
With tech changing like crazy these days, we all have to roll with a lot of sudden shifts, and even make trade-offs we never thought we’d have to make. On subreddits where professors and teachers gather, I often read people talk about how students these days depend too much on AI, can’t focus, and so on…
I’m not saying that that’s not the case, but I think we could all try to be a little more understanding and forgiving.
Because no young person today chose to be born into a world run by phones and social media. And nobody would pick to start their career in their 20s right when AI is flipping everything upside down. Feeling disappointed, lost, angry, or helpless is just totally understandable.
This makes me realize more and more how important it is to hold on to “hope”. Not the cheesy slogans people shout in anime or movies, not the kind of blind optimism about the future, but something closer to a discipline that you can keep consciously reflecting on… An inner backbone that can take whatever the unknown-unknown throws at you—the good and the bad.
I’m reminded of what Andy said to Red in “The Shawshank Redemption”:
Remember, Red, hope is a good thing, maybe the best of things, and no good thing ever dies. I will be hoping that this letter finds you, and finds you well.