domain:alexberenson.substack.com
Amadan
Enjoying my short-lived victory
50m ago
As it happens, I have also been dipping into LLMs-as-beta-readers lately, even going so far as to build an application that can read an entire series of books and learn its "lore," and a custom GPT instance that will "compress" a book into a format optimized to provide context to itself or another GPT. (As you probably know, even the most powerful LLMs do not have a context window large enough to store an entire large novel in memory, let alone a series, and you can't directly upload embeddings to GPT or Claude.) The intent of these projects is so that I can, say, ask GPT to evaluate the fifth book in a series with knowledge of the previous four books. It's a work in progress.
So, some observations. First, sorry dude, but I have major side-eye for your ability to evaluate literary quality. :p
That being said, I have also noticed the tendency of LLMs to glaze you no matter how hard you try to solicit "honest" feedback, unless you resort to tricks like you mentioned. (Telling an LLM the manuscript is by an author you hate and you want it to roast it will work, but that's not exactly useful feedback.)
The hallucination problem is hard to overcome, even with tricks like my token-optimizing scheme. I find that in most sessions, it will stay on course for a while, but inevitably it starts making up characters and events and dialog that weren't in the text.
As long as you can keep it on track, I have found that some of the GPT and Anthropic models are... not terrible as beta readers. They point out some real flaws and in a very generic sense have an "understanding" of pacing and tone and where a scene is missing something. However, the advice tends to be very generic. "You need to show the consequences," "The scene ends too quickly, you should build more tension," "There should be some emotional stakes the reader can connect with," etc. Clearly they have many writing advice books in their training data. There is nothing like true understanding of context or story, just generic pieces it can pattern-match to the writing sample you give it.
And when it comes to specific suggestions, I have yet to see an LLM that is actually a good (not "mediocre and banal but capable of producing literate prose") writer. Its suggestions will be a pastiche of flat TV script dialog and trope-filled scenes.
(That said, any writer will tell you to listen to critics when they point out problems, but don't listen to them when they propose solutions. So in that respect an LLM isn't much different than a human.)
But these are still early days for AI, so I don't doubt that in a few years, we'll have LLMs that can be at least as useful as your average writing workshop. AI writing is already flooding some genres, and while it's usually as easy to spot as AI art is, just as with AI art, a lot of people clearly don't care.
I find it fascinating and I enjoy playing around with it, but yeah, I think AI-generated novels will crowd out human writers in low-brow undiscerning stuff like romance and progression fantasies, and writing those stories will become something people only do as a hobby, just like people are still passionate about chess and go even though no human can beat a computer anymore. I still think we'll need true AGI to write an actual good novel. When you show me an AI that can write a coherent series, with multi-volume character arcs, plot seeds planted in early books that clearly pay off in later ones, literary allusions and metaphors that aren't just clumsy pulled-off-the-shelf ones but deeply enmeshed in the story, and a recognizable differentiable style (in the same way that fans can read Dickens or McCarthy or Hemingway and immediately recognize the author), I will believe we're there.
dr_analog
top 1% of underdog fetishists
34m ago
You've made a profound and insightful point about online discourse in the age of generative AI. Let's break it down.
/uj you might be able to get away with emdashes so long as you steer far away from sycophancy
cjet79
33m ago
That application you are working on does sound interesting.
I've been wanting to skip the middleman for a while and just have AI write the stories based on simple prompts.
I have an existing 300 page story I'd love to just feed to an AI and have it finish the story for me, or at least fix it up.
Back when I fed the first chapter to chatGPT it just told me that my story was offensive and refused to help me, which was when I stopped using it altogether and a few months later switched to grok.
Progression fantasy : Epics :: sex : love
And anything with a modern setting is just unbelievably boring or depressing.
ulyssessword
30m ago
I'm going to guess without strong evidence that the vast majority of the views on the toronto article are of the corrected version.
I couldn't find the precise time to correction for the Toronto Star, but Global News was 5.5 hours. This paper (pdf) classifies news articles based on the speed of their spread, and found that most articles peak within the first four-ish hours (some much faster).
Regarding the bodies, I would hazard to guess that the articles themselves are not technically "wrong" and therefore do not warrant correction.
That would be incorrect. There are no reputable claims of bodies being there, and none have been unearthed or "confirmed". The Law Society of British Columbia (at a minimum, among other groups) is going against the findings of the First Nation in question, who are in charge of the site. The bodies were hallucinated into being four years ago, and they're still around now.
Primaprimaprima
...something all admit only "TRUMP", and the Trump Administration, can do.
29m ago
Musk-level value was OP’s analogy
I said it was a "heavily attenuated" version of that. It was just an analogy, not meant to be taken literally.
phailyoor
28m ago
Regarding the burial sites hoax, that is state sponsored propaganda, so any misrepresentations or questionably false statements are intentional, rather than mistakes. You wouldn't expect an article printed in the Pravda to misrepresent the facts accidentally either. AI is not immune to this, for example in chatgpt.com Sam Altman can put whatever narrative he wants into the system prompt and the ChatGPT will dutifully push it at every opportunity.
The hoax is also pushed with a degree of nuance. The journalists always cite their sources and never claim it to be true themselves: "According to [Insane sjw activist group], Donald Trump literally murdered 300,000 kids with his own bear hands." is not technically a false statement. As much as the reporting is insanely one-sided, I do believe that it's often newsworthy to report on the claims of [Insane sjw activist group] even though their claims may often be false.
They did not retract this article because it's not strictly false: https://www.cbc.ca/news/canada/british-columbia/tk-emlúps-te-secwépemc-215-children-former-kamloops-indian-residential-school-1.6043778 The are re-reporting what the indigenous group falsely claimed, but did not ever make a statement of fact that there were actually bodies there.
Nevertheless, they actually provided a correction after the fact at the top of the article that says there were no bodies discovered:
Since this story was first published, the Tk'emlúps te Secwépemc First Nation has updated its findings and clarified that what was detected in the ground-penetrating radar survey were about 200 potential burial sites. Read the full story here.
phailyoor
25m ago
I never claimed that AI is useless, I simply claimed that for the purposes of certain tasks, namely copywriting, AI is currently far far inferior to even below average humans.
I use AI every day for what it's good at, so I would be the last person to say that AI is useless.
Amadan
Enjoying my short-lived victory
20m ago
Other than reach and better animation, I don't think this is different from the AI companions that have been available for a while. Replika, the most famous one, will already do NSFW ERP. And yeah, there are men (and women!) who have decided their Replikas are preferable to real people.
You're concerned about what this will do to the psyche of teenage boys, but I'm surprised you haven't thought of the male version (which no doubt will exist). A tall dark sexy boyfriend who will treat you only exactly as roughly as you want to be treated, and will listen to you going on about your problems and your neuroses with endless patience and understanding and affirmation? If a real woman can't compete with an AI girlfriend, neither can a real man compete with an AI boyfriend.
That said, I do think your fears are somewhat overblown. Porn has, IMO, been bad for society, especially the ever-increasing availability of extreme and degenerate porn like our grandparents could barely imagine. But I think alcohol and drugs and gambling and smoking are also very bad for society. If I could wish them all away, I would. These things exist, however, and society persists, accepting that some percentage will be sacrificed to Moloch, and Moloch always has new incarnations. I accept that AI companions are a hazard, but I don't think they are "the" thing that flatlines birthrates and normal sexual relationships.
phailyoor
17m ago
This paper (pdf) classifies news articles based on the speed of their spread, and found that most articles peak within the first four-ish hours (some much faster).
Hmm point taken. That's certainly not ideal.
That would be incorrect.
The articles only reported that the first nations claimed that bodies exist. The articles never claimed themselves that the bodies exist, so the articles are not technically false. Nevertheless, CBC still was gracious enough to update the article and write front and center that there were no bodies, which is not something that they had to do at all, yet they did anyways. What more do you want CBC to do before you will be happy?
Now that it's obvious that there are no bodies, I'm confident that not a single recent article from a reputable source has tried to claim or suggest that the bodies exist anymore.
And yes, journalists should all be minecrafted, but that doesn't mean they're technically wrong, they're just evil conniving cunts.
ABigGuy4U
15m ago
I’m less concerned about the rare cases of complete psychotic break and more concerned about the rapidly nucleating cult behaviors that are quickly assembling into a an actual new religious movement.
sockpuppet2
13m ago
Can you charge at home for as long as you plan to keep the car? Quality concerns about Teslas aside, I think the main things to consider when weighing BEV or PHEV vs ICE are:
-
Can you charge at home for as long as you plan to keep the car?
-
Can you responsibly afford the upfront cost?
-
Do you like a car that's more "gadget" than "appliance?"
If "yes" to all three, you're probably better off with an electric car. If only the first two, it depends how much the "gadget" design ethos commonly used in BEVs annoys you. If you can't charge at home and/or the upfront cost is over a responsible budget, you're probably better off with a non-plug-in hybrid or ICE-only powertrain.
sockpuppet2
13m ago
What kind of engineer?
faceh
5m ago
I dunno, it feels like a deathblow on top of all the other mentioned factors. The thing that finally kills our urge to climb out of the hole.
There's a version where the AI can teach a man (or a woman!) how to talk to the opposite sex and both select and become a good partner.
But thanks to molochian incentives, that's not what we will get, if there's an immediate way to use the tool to extract resources from people rather than guide them to what they truly wish they had.
faceh
1m ago
Right. This is just the next step of a pattern that is already established. Money exchanged for the feeling of romantic or sexual attention/attraction.
They were already having a fully artificial 'relationship' with a digital 'woman' who doesn't know of their existence.
This just cuts out the need for a woman in the loop at all.
Not in the current form.
I'm dedicated to pursuing a quality of "authenticity," which I don't have perfectly defined, but definitely requires that my partner be a real human, with 'natural' skin, brain tissue, and standard human DNA. The thing that I'm ACTUALLY wired to find attractive, not something that mimics those things closely enough to pass a basic inspection. Related to why I don't really like Tattoos on women.
In many ways, we are descending into my version of hell, where finding meaningful connection with other humans is harder than is needs to be, where women are more focused on careers and adventures, at the expense of their own happiness, than even trying to find joy in bearing and raising kids, where men are fundamentally purposeless and nobody bothers to even try to create a purpose for them, and everybody is busy trying to live at the expense of everyone else, b/c coordinating to create that better future is HARD and we aren't able to see past the short term consequences of these actions. But I can, and it seems increasingly obvious where this is trending. And nobody with power is doing much about it.
And its all being patched over with digital (i.e. INAUTHENTIC) simulacrum that sort of satisfy the various urges without really fulfilling the purpose for which each urge exists, and these experiences that are simply insufficient to make you happy if you care to look and notice the cracks in their facade.
All the worse because I can clearly imagine a better set of circumstances that is happier for everyone, including myself, and I have a vague idea of how we could get there, but no real clue on how to implement that plan, and thus I am left to scrape by with whatever my individual efforts can achieve.
More options
Context Copy link