The founder of the Sublime app announced some integration with ChatGPT and Claude where you can share collections… honestly, I find it interesting and I’m curious to try that out later. I’ve continued to save things all though I need to add a lot more to it. After all, the idea behind sublime is to share things not for the sheer sake of collecting ideas (which is a bad habit I’ve already fallen into with it), but to connect these ideas for a greater purpose. I need to better organize the ideas, honestly. Put more tags on them at the very least.

This hearkens back to the voice note I made earlier. About “Prompting AI to help us ask better questions”

I said: I keep seeing people talking about how there’s this explosion of outsourced creativity and thought processes to large language models. I happen to agree that this is not a particularly great idea mainly because the AI tools are only as good as how you prompt them. I remember early on when I was learning more about ChatGPT and Copilot, one of the main things I was interested in was prompt engineering, and how you can intrigue the AI to go in directions that common everyday questions wouldn’t necessarily make it go. I need ideas to create prompts that are not exclusionary of potential connections that you might not otherwise have. 

However, I also agree with people that you shouldn’t just go to AI to ask: hey what does this mean? The better way to go about it is to make connections on your own. Provide some context. Do a little bit of digging beyond the surface yourself, then turn to AI to try and help you make sense of everything you’ve come up with. So, for example, I was reading today and I was as usual transcribing entire passages and then commenting on them. I would highlight my entire notetaking session, then throw that in chat and ask it to make sense of it. Today, it actually did a pretty good job summarizing my points and actually had some “insights” of its own. But, of course, AI cannot have insights per se. It simply is trying to give you the best possible answer it can based on the language you’ve given it. One thing I really want to do going forward is rather than looking to try and answer unanswerable questions, which just leads to further fatigue and burn out intellectually, I want to use the tools available to ask better questions. Therefore, how do we prompt AI which is not inherently creative itself to come up with questions that we should be asking? How do we create a framework around which we can actually stimulate our creativity further and not simply handover the reins of our explorations, even in part, entirely to chance?

…

Believe it or not, I think people underestimate the ways that AI can stimulate further curiosity. Perhaps it can’t be entirely original, but it doesn’t need to be if it’s helping point you in directions to further your thought process. But the initial work of discovery needs to be your own. It can be helpful to have AI create a summary of something that you’re reading, but you still need to read the work if it’s something that you ultimately decide is useful. What AI is best at doing is distilling what you already have come up with and provide further guidance on where you could go next. It’s an assistant. It can’t, nor should it, do everything for you, even if people are somehow building entire websites off of the back of AI by likely synthesizing the most popular articles online and book excerpts into something that sounds pretty damn smart, and all the human does is cite sources. 

AI is great at getting a lot of information in front of you, then parsing what it “thinks” is relevant. I’m well aware that its methods of doing this are not always optimal. What bothers me about this tech is that it makes plagiarism very simple… because is it really plagiarism if you take three or four different things, mash them up, then quote related sources directly? Isn’t that the same as writing a paper? The problem is, you as the author aren’t really doing the work… and the work is the discoveries that make something feel novel… if something reads as rehashed, it’s probably at least in part AI generated.

That’s not to say AI summaries are bad. I think they are quite helpful in parsing through videos and various things I could read. If I pick out a few points that I want to hear more about, I’ll go to the source material… but if the summary doesn’t offer me much to go on, I can file it away for later. It’s a slippery slope using AI “ethically” whatever “ethical use of AI” even is at this point… it’s so much engrained in internet content that it’s not going to ever leave. So that’s why I’ve learned how it works, see what it’s capable of doing, and see how it breaks, which is sadly, a lot. But that’s OK, you don’t want it to be perfect. In fact, sometimes the hallucinations and bizarre answers and curious word choices it makes are part of the fun.

Social media literally has altered our realities. Yes, over the past five years, the rate of people leaving social media has increased significantly. I don’t have numbers to cite at the moment, but anecdotally, it seems people are quitting by the many thousands every day, whether it be Facebook, Twitter, Linkedin… but a lot of their attention spans are still being held onto by TikTok and even YouTube to an extent. I need to be more careful with what I watch honestly, even if a lot of what I watch, like the sports card pack  openings, are just background noise. That and I’m intrigued how people waste their money.

Anyway, because so much of our attention is on social media, and so much of it is drama, and today’s “entertainment” is crap for the most part (yes there are exceptions) that most of the discussion around today’s films and music is about how “bad” they are. In reality, I think today’s music is getting better, believe it or not. People are tired of the same recycled garbage it turns out. But that’s for another discussion. 

The important point here is very relevant. The reason I don’t read news (what I see is usually by accident) is that there is nothing that it actually provides me that’s actionable. I don’t think that harping on how bad the current administration is, or complaining about the same societal issues we’ve had since Roman times is productive. (Longinus himself was complaining about the downfall of society in On the Sublime!)

Ever since I quit social media back in 2022, this has been a time of many serious questions that I’m still struggling to answer. When I tried out Blue Sky, and poked around on Pinterest and Tumblr here and there, it was more to see the state of these platforms. It’s all quite bad. I see it all for what it is; a massive time suck (although Tumblr, from the things that Anna Howard shares on Wild Geese, may not be entirely a lost cause, and even the toxic wasteland of X still has some interesting things that have been saved on Sublime.) 

All is not yet lost, but we’re teetering awfully close to a complete societal breakdown. And funny enough, it’s not because of social media; it’s just hastening the decline of the American capitalist empire. The best minds are too busy finding out how to keep lesser minds occupied and buying shit they don’t need and coercing “aspirational selves” for fun and profit, as previously mentioned. They’re not actually solving problems, they’re just slapping band-aids often enough to keep the appearance of function and status quo. But I’ve seen the state of things lately… Everything is getting progressively worse, and quickly.