Flâner in Komachi | 迷歩小町 (xiǎo tīng)

My relationship with AI is complicated

To summarize why I wrote this post: 人在江湖，身不由己。(= "One has to compromise in this world.")

I drafted this up on 3rd of September. Then I was feeling low, either due to anxiety from job applications, or PMS, or SAD, or all of the above. Today I still don't feel quite myself, but I want to accomplish something to cheer myself up, while also 给自己一个交代 (="to be at peace with myself"). Plus, I found songs to sustain my energy throughout the writing process.

So, here we go.

It starts with...

My first encounter with AI dates back to when DALL·E first came out. I was impressed with the technology but viscerally hated the images it generated. I have been following a lot of artists on various social media, as a result, I learned about the copyright infringement issue. That was a big red flag. It's morally unacceptable, so I stay away from using any image generation AI tools.

I used to get MIT Technology Review newsletters sent to my inbox, in order to get exposure about what's happening in the industry. I was trying to "stay relevant" as I was a newcomer to the tech world. I forgot when I stopped catching up with the news, probably after feeling too overwhelmed and desperately wanting to focus on my life.

Then ChatGPT came out, making everyone excited. I used it very sparingly, mainly to polish my resumes. Because I think of resumes as merely collections of garbage words with a formula, it just makes sense to use AI for it.

And then time flies.

Here and there, I picked up random stuff from different conversations and news feeds. I wasn't actively keeping up with the latest trends. The tech landscape changed rapidly, but my work (note: it's a traditional industry) likes to stick with the old workflows. FOMO creeps up.

Recently, upon losing my work, I finally decided to step out of my comfort zone and stop overthinking (I constantly combat my procrastination by pairing actions with thoughts). I need to get my hands dirty. So I signed up for an AI course related to my profession, even if I still feel really torn about using AI. Surprisingly, I enjoyed the course a lot (mostly because my cohorts are great people). In retrospect, I must admit that it definitely helps with my FOMO and gratifies my vanity of being ahead of my peers.

Ongoing ethics checks

Trying out the tools on the market naturally demystifies them. The AI tools are great at making things up and being sycophants. They speed up the 0-to-1 process enormously, which can grant you a fake sense of accomplishment. The real pain happens after the initial generation. That being said, AI tools are efficient in generating things that you don't necessarily care about, especially when your clients or upper management value quantity / metrics / outputs / closed tickets.

There are times that I feel I shouldn't use any AI tools because it's unethical. I don't want to be part of the AI boom any more. I don't want to contribute more user data to those greedy corporations. I want to live my analog life.

To cope with my moral distress, I wrote down a list of things to remind myself when to use, and when to NOT use, AI tools.

Bottom line (底线) - Support human creations

Green light - Outsource soul-crashing tasks to AI

My mind can't think straight right now

Am I justifying evil acts, given that I know how much negative environmental impacts and the expansion of digital colonialism are caused by Silicon Valley's AI gold rush? I don't know.

The air we breathe is infused with capitalism. Hot money dictates the direction of technology advancement.

To keep me sane, I will actively read criticism about AI, and continue to establish my own principles when it comes to AI.

I lost my flow in the middle of writing.
There's a lot more to say but my brain won't cooperate (脑袋木木的). Hopefully I'll come back to this topic in the future.
As always: #nobetawedielikemyjob

Writing BGM:
🎵あなたはかいぶつ (you are my monster), by TOOBOE.
🎵Lonely One (feat. Hikaru Utada), by Nariaki / 小袋成彬.

#ai #reflection