site banner

Wellness Wednesday for April 19, 2023

The Wednesday Wellness threads are meant to encourage users to ask for and provide advice and motivation to improve their lives. It isn't intended as a 'containment thread' and any content which could go here could instead be posted in its own thread. You could post:

  • Requests for advice and / or encouragement. On basically any topic and for any scale of problem.

  • Updates to let us know how you are doing. This provides valuable feedback on past advice / encouragement and will hopefully make people feel a little more motivated to follow through. If you want to be reminded to post your update, see the post titled 'update reminders', below.

  • Advice. This can be in response to a request for advice or just something that you think could be generally useful for many people here.

  • Encouragement. Probably best directed at specific users, but if you feel like just encouraging people in general I don't think anyone is going to object. I don't think I really need to say this, but just to be clear; encouragement should have a generally positive tone and not shame people (if people feel that shame might be an effective tool for motivating people, please discuss this so we can form a group consensus on how to use it rather than just trying it).

2
Jump in the discussion.

No email address required.

It's possible whole movements will be established where people choose, perhaps for aesthetic reasons, to consume only books published before the advent of - say - GPT-4, to ensure they only read 'confirmed' human fiction.

This is already more or less my plan, with a cut-off of perhaps 2010. Not entirely due to AI though.

Since LLMs make (or will very shortly make it, depending on who you ask) 'AI' writing indistinguishable from human writing

GPT-4's writing is wholly unimpressive to me. I mean, it's impressive "for an AI" certainly, it's impressive in the sense that this sort of thing was unfathomable to me a few years ago - but in terms of actually writing a long-form piece that makes me think "hey this is really great, I want to read this", there's basically nothing there at all, among all the examples I've seen.

This is in contrast to the best examples of AI art, which definitely do look good and are pretty nice to look at, despite still having some flaws. This basically matches with my intuitions about what current machine learning techniques should be capable of, because "draw a spatially accurate picture of X in the style of Y" requires less creativity than "write an engaging and thought-provoking novel". Representational art is ultimately a technical problem (and I did recognize this fact many years before I ever thought that AI art would be possible, you'll have to take my word for it) - computers have been drawing pictures for decades. It's just that they used to require a massive precise description of vertices and colors in order to do it, but now we can give the description in natural language. Writing (the best writing anyway, the only writing that I bother reading in the first place) brings you immediately much closer to raw ideation - you can't just construct a product according to technical specification, you have to have novel thoughts at the same time, which is a much harder problem.

It's possible that there's a fundamental "creativity barrier" that the current LLM architecture will never be able to break through - creativity being that X factor that's involved in the production of genuinely original art, or novel scientific discoveries, or influential philosophical ideas. It doesn't seem to me that GPT-4 is any closer to breaking through this barrier than GPT-3 was. I acknowledge that I could be wrong though, and if GPT-5 is able to write a novel as good as Nabokov's Pale Fire then I'll update more strongly towards thinking that scaling transformers alone really will produce AGI.

I have a different, not necessarily conflicting perspective as yours.

Namely, that human writing is shit. No really, 99% of literary output is utter dross barely worth proofreading. Look at the amateurish garbage regularly posted on Reddit writing subs like Writing Prompts as an example.

What strikes you as good writing is the cherry-picked output of tens of thousands of would-be writers, and as such, you're not comparing GPT-4 to the average level of human writing, which it handily exceeds.

As of today, it can't write a proper novel since even a 32k token context window isn't sufficient, but I have gotten very decent prose out of it in my tinkering, especially if you prompt it with the names of existing good human authors.

In other words, GPT-4 is a General Intelligence in the sense that a typical human is a General intelligence, it's simply not strongly superhuman at the task.

No really, 99% of literary output is utter dross barely worth proofreading. Look at the amateurish garbage regularly posted on Reddit writing subs like Writing Prompts as an example.

What strikes you as good writing is the cherry-picked output of tens of thousands of would-be writers

I completely agree with you here. Sorry if I didn't make that clear enough. My point was that if all GPT is doing is beating the average or even the above-average writers then that's not really relevant to me, because I wasn't reading that type of writing in the first place. It needs to beat the superior writers to really have an impact.

I imagine that we would have some disagreements over how big the elite actually-worth-reading club is, and who gets to be admitted to it, but I agree with the basic point you're making here.

As of today, it can't write a proper novel since even a 32k token context window isn't sufficient

Actually it can, with recursive prompting. Predictably, the result is not very interesting. But in fairness, there would probably be an immediate improvement if you got rid of all of OpenAI's safety guardrails.

but I have gotten very decent prose out of it in my tinkering

Oh sure, it can write some good sentences. But writing a good sentence is like drawing a good (representational) picture - it's a purely local problem that doesn't have to deal with global conceptual structure. What makes a strong essay or a strong novel is the global structure, and the intricate relationships between local pieces and the overall structure. If there is an inherent limit to the "predict the next token" methodology, it will be found here.

Fair enough! I don't think we have any real disagreement at all then.

I've done my part, by writing a well-reviewed novel before GPT-5 can write better than me. It's going to stay up at least on archives, so I can forever bask in the miniscule amount of glory in provably being a good writer without AI augmentation 🙏

In terms of how the market will react in the future, I doubt draconian steps will take off, it seems more likely that AI will just subsume the entire market, leaving a small niche for people who claim that they're writing solo, often without showing proof, and reacting negatively to demands for proof.

I'm aware of people doing that, I just happen to believe it'll be a minuscule and ever dwindling share of the market, as is the case with most artisanal products.

People watch humans play chess because the product is the human matching of wits, whereas the majority of books sold seem to me like they're judged as products themselves, with knowledge about the author being a secondary concern.

There are obviously celebrity authors, but we also have blatantly ghostwritten works, pulp fiction, and plenty of genres that don't really give a damn about authenticity.

Right now, barring ghostwriting, it's difficult for any given author to significantly out produce another, they're still humans manually writing after all. But that's changing before our very eyes as I speak.

On further reflection, I think that authorship itself will become largely obsolete, outside of nonfiction and the like, people will very cheap custom order GPT-5+ to write on the fly something they'd like to read.

"Computer, write a salacious but not outright pornographic bodice-ripper set in a the Portuguese colonies in the 17th century starring a down-on-their-luck mestizo girl accidentally winning the heart of a heir in hiding. Oh, and make sure it has a happy ending, and give me 3 options at the end of each chapter for where to take the plot next."

This is almost inevitable, and will be easier in text than more multimedia heavy productions, even if the latter will only lag behind by a few months or years.

Authors, if they still exist, will be mostly brands offering a certain specific flavor of writing, not because they're sought for their own sake.

Back to your original point about demands for proof of human work, AI image generation is pretty much a solved problem. But in practise, a typical artist isn't assailed with demands to show receipts demonstrating they never resorted to using AI at any stage of the process, including digital art, at least not that I've noticed. I expect that literature will take a similar route.

On the longer term, I'm sure humans will be able to avail of cognitive enhancement to become more competitive. As I've said before, I don't want to just be someone suckling from the teat of a godlike machine intelligence, I want to be said godlike intelligence, writing incredible novels and designing beautiful works of art myself.

And reacting negatively to demands for proof

I'm sure there will be people who would love to deliver proof. I, for one, would love to pay for a provably man-written story, even if it's inferior to an AI-written one. The question is how do you prove it? You could deliver a handwritten manuscript, and I have no way of knowing if you didn't just copy it off GPT. Do we start locking authors in monasteries and not letting them out before they finish a book?

It's possible whole movements will be established where people choose, perhaps for aesthetic reasons, to consume only books published before the advent of - say - GPT-4, to ensure they only read 'confirmed' human fiction.

Can we backdate this to the advent of social media/smartphones?

A man can dream….