[2026-08-24] my ai manifesto, i guess
LOCK IN is my 2026 mantra. It really just means "let's focus on what's important and stop being distracted by true crime podcasts running in the background 24/7.” And locked the fuck in I was this week… squatting on a kitchen stool, shrimped over the counter with only my laptop screen to light up the night, racking my brain for ways to automate away the undefinable "life admin” tasks that OpenAI propaganda keeps harping on (a feeling more than a tangible list). I impulse order 200 NFC tags hoping that each small tap - to start a nap timer, to open my daily checkin with my roommate, to {third example I can't even come up with right now} - will file down my anxiety and put me in a position to make a bajillion dollars while giving me the space to meet the future love of my life in the Whole Foods granola aisle. I'm interviewing for an AI engineering startup and I ask my brother (a 24-year old Berkeley EECS software engineer living in the Bay) how he feels about autonomous dev tools and he says "idk what those are.” I dream of drowning often, but that night I dream he dies instead and I wake up sobbing.
When I first started working, locking in meant keeping a tight schedule and running errands between boozy brunches and birthday parties and bakery heart-to-hearts, working at "100% efficiency" (a confident commitment I made to my first manager, which was met with a quiet look of concern and the advice to maybe not do this because I would quickly burn out -- waddya know, he was right!), moving with an aggressive swiftness around Midtown, and walking the routes-of-least-repeat-so-I-can-get-my-laundry-and-fill-my-water-and-put-my-keys-in-my-work-bag-in-one-trip in my apartment. A bit unnecessary, yes, a little neurotic, perhaps, but all symptoms of a borderline crippling obsession with efficiency. In hindsight, any friend or coworker or stranger looking through my window could probably sense the optimization coursing through my veins and spilling into how I move around the world.
As someone who can't trust my own brain to conjure an objective reality, it's tempting to delegate away decision-making to a more "objective” source. The truth is, I find myself instinctually reaching for Claude on more occasions than I'm proud of because I simply can't guarantee the validity of my own anxiety and OCD-riddled inner world. Instead, I could hypothetically press a button on my phone to record my thoughts, automatically transcribe those thoughts to words in an iPhone note (dated and timestamped), feed those words into an artificial brain I build on my laptop, and automatically schedule focus blocks to "work on my anxiety app” or "sand my chair” or "send out wine & craft night partiful” based on my open projects and my calendar. (I did indeed build this. Ask me about the efficacy. Or don't.) But the robots won't know that I've been putting off sanding for weeks because it's simply too hot and humid for me to bear being on the roof or that I've been too anxious to work on my anxiety-tracking app. The robots won't consider that I dreamed that my brother died and I woke up with my whole day ruined or that I got too high at a rave with coworkers and spent the next half day paralyzed, ruminating over my embarrassment. The robots can't tell me what's important to me. And the robots are as reliable and trustworthy as humans are, which is to say they're not very reliable and trustworthy, but the difference is that something that doesn't have a heart can't be held accountable. And so, at the end of the day, I still consciously try to put agency back into my own hands.
These days, my brain is whelmed with pithy phrases I hear in the office stairwell, on LinkedIn and Instagram, or from the casual suited passerby in my favorite cafe, and I feel my own voice slowly getting quieter and less familiar. I voice-to-text my friends, but as someone who has always claimed to be better expressed and understood in written words than spoken, the speed and lack of second thought feels reductive, like I didn't think through what I really meant, didn't give my recipient the courtesy of doing the hard work of distilling meaning out of nothing, and so it's now up to them to decide what the point is, shrug!
I Wispr nonsense straight into Claude Code and instruct it to follow the Amazon (yes, the Jeff Bezos one) Writing Culture standards and my friend texts me that Harvard closed their Writing Center and let go of its director and back when we were Writing Center Fellows at Princeton it was all about motive and what's underlying the thesis of a paper but now we shift the burden of clarity to agents who can only interpret as much as we give, motive or not. We think so much about output and so little about what it feels like to receive it.
As for my brother… is it willful ignorance or an intentional refusal to engage? Do I live in a bubble (hint: the answer is yes) or can there actually be a future that isn't as doomed as I increasingly feel is guaranteed? All around the Internet and even in my own circles, I see people dreaming of opening cafes, crafting, and obsessing over the very concept of whimsy - things that can't be AI-ified. But even this movement feels artificial, like we're somehow converging on the same patterns and interests, guided by the invisible hand of the Algorithm. And to be fair, some of this is less related to AI and more a reflection of the shallow interconnectedness bred by copying what performs best online… the Westernification of cafes, the phenomena of culture-adjacent moments like having a mahjong night and saree-inspired fashion or experiencing a very Chinese time in your life - all of which are individually quite fun trends in my opinion, just as much as they are illusions for real understanding.
My mom, a network engineer, used to ask me and my brother to get on a Microsoft Teams link to write code for her, but now she goes to Claude Code and doesn't call us for help anymore. My dad, whose LinkedIn bio says "A Curious Mind,” used ChatGPT to digitize the reverse dictionary my grandpa invented and hand-wrote over the course of 6 years. Decades after my grandpa wrote the first character, the dictionary is finally set to be published in the year he turns 97. In the meantime, I make my little apps, ask AI to write emails at work, and feel guilty for contributing to an objectively worse future for the next generation. I become more productive at the cost of feeding my own perfectionism and compulsion to optimize, telling my brain "you can be better, faster, stronger...you're not doing enough just yet...just a little more…” and I continue to reach for that little more until I lose all sense of reality, preserved simply as a shrimp at the keyboard in the dead of night.
I originally wrote this by hand, frantically. There were several moments where I considered switching to dictation. I have so many thoughts that I'm scared to lose lest I immediately record them into existence. But maybe that's the point - that not all fleeting thoughts are important or deserve to be heard. Some thoughts are better left unsaid and forgotten, so as not to dilute the ones that actually matter. Words are cheap; meaning and connection and understanding and intention and thought are not.