This is NOT a Challenge Submission Post 🥰
I was checking out Hacktoberfest Weekend Challenge submissions. I do this from time to time on DEV Challenges, even when I myself don't have time to submit my own projects.
This time was no different. I didn't have time to submit my own project, but had a couple of hours to checkout other submissions.
Then I was wondering, what would be my project if I had time! 🤔
Before I go forward, I just want to spit something out! Yeah, I could just ask Claude, Antigravity, or GitHub Co-pilot to generate the idea, build the implementation, write the submission post and submit — all by themselves! But that's not me. I would've revealed if I had done that. Otherwise, it wouldn't be fair to other contestants!
On the other hand, I wouldn't submit something, if at least some important parts of it wasn't my own. In other words, I'm not in favor of submitting completely AI generated works, even with the revealing! That's just my choice. I don't intend to pass judgement on others. If you want to do that, it's 100% up to you!
Now that it's out of the way, let me go back to the original story. Where was I? Oh, I wan wondering, what my own project would've been, if I had enough time for this Challenge.
So I was thinking that, and just to do a bit of brainstorming, I opened Gemini. I know, most people prefer Claude these days. But for these sort of brainstorming, Gemini works well for me, especially Gemini Notebook, A.K.A NotebookLM.
So I asked Gemini specifically to go to the challenge description page, and cook 10 unique ideas for me.
To my surprise, well, may be not a total shock, but at least I raised one of my eyebrows, when I saw Gemini's suggestions mostly matching many of the submissions! Some even almost verbatim.
Then I thought, well, it's Google. So, perhaps Gemini already indexed those posts, right?
Then I extracted all the submission entries URLs myself, and then asked Gemini again to check those posts, and suggest 5 unique project ideas that are different enough from those submissions.
This worked. I'm pretty sure these 5 ideas were unique up until that point! I was already sleepy at that point, so I want to sleep without going further down the rabbit hole.
The next morning, which is the date of posting this, after some routine works, being only few hours away from submission deadline, I checked for new submissions again.
And this time I got the shock! This time I noticed quite a few posts, very close to what Gemini suggested me, which were unique up to last night!
How did that happen?
Did Gemini suggest me the same thing some other people were working on, or brainstorming with it?
Did they come up with those ideas all by themselves, and it was just a coincidence?
Did the frontier labs share data among themselves, and people keep getting LLM cache hit?
I have no idea. But it's too much of a coincidence to just let go!
I remember the incident that happened a while ago, between a mathematician and OpenAI, about OpenAI's frontier model solving Navier-Stokes Millennium Prize Problem.
To summarize the incident somewhat, the mathematician claimed he was using ChatGPT with his work. So the LLM just used his research data, and regurgitated with some of it's own variations, and OpenAI just took it as the LLM's own invention.
If I consider that everyone is being honest in such incidents, then this is the only logical conclusion I can think of: when we are using AI, AI is using our prompts, works, conclusions, and spreading them around as either cache hit or part of it's training data, killing our own intellectual human uniqueness in the process.
What does that mean if my conclusion is correct?
I don't know. I leave the decision up to you. But I know exactly what I'm going to do from now on. But that's a post for another day. 🥰
Note: If you find a lot of spelling and grammar mistakes in this post, that's because I just typed it all in one go, without bothering to check with any AI tool. So, if something is unclear in the post, please just quote it and ask me in the comment. 🙏🏼
Top comments (0)