Good post! I often think about what Stephen King is his book On Writing wrote: “writing couldn’t really be taught, only learned.”
Anders Lindström
Karma: 80
Yes, I agree. There is a growing premium on being a native english writer these days to prompt/correct the LLM to sound human. The rest of us stumble in the dark when we use an LLM do get our texts in order.
Hey Gwern,
Haha, I am very curious why you felt like that reading my post? Is my Idea about the imminent death of vibe coding and the birth of software mining so crazy that it must be AI slop and that I should consider checking myself into a lunatic asylum?
Hey Karl,
There was no prompt approach that generated a one-shot post and that could be copy pasted if that is what you wonder? Yes. I have used LLM’s (Gemini and claude) extensively, both for language (english is my second language) and research.
Hello RobertM,
Don’t consider this a challenge to your decision, I just want to have some guidance on this. So please bare with me.
1. English is my second language, and I use LLM’s extensively to clean up the mess I create when I write any English sentence or paragraph. I usually reiterate every sentence, paragraphs, sections many times using LLM’s. In the LLM writing policy it says that “first-time writers are not permitted to use any AI text output in their submissions”, should this be taken literally, that no matter what and how that output was created?
2. Style. I like snappy one-liners and “bold” straight to the point language. That is something I aim for. I want it to read like a tweet or ad-copy (in the best sense). Is this frown upon and taken as slop?
3. Referencing. I have a feeling that the referencing is seen as overbearing and part of the “problem”. This is a forecast and I do not like to make claims without backing it up with references. For that reason I have a lot of references for this short piece. If the referencing style is an issue please provide some guidance on that?
4. A direct question on the content. Is there anything in my post that is factually untrue or something that is believed to be a hallucination, which then disqualified it as SLOP? Yes, the post IS speculative, which is the whole purpose, which I also mention in the beginning. But I have seen this trend and wanted to make a “solid” case for it (which I guess didn’t really go as planned...), so to the best of my knowledge I have checked every reference multiple times.
Regards/Anders
Of course hunter gatherers eat/ate carbs, but they did not base their diets on grains and beans. Fruits and berries “wants” to get eaten. But… how long is the fruit and berry season? A few months if you are lucky. The rest of the time animal foods is pretty much the only thing available. Sure you can chew on the occasional root, but how many calories will that give you?
My stance is that more animal foods in peoples diets would reverse some of the damage from the high carb ultraprocessed foods in peoples diets. I live in Sweden. The recommended weekly red meat consumption is 350grams. That is like one or two meals. To me that recommendation is madness when you look a insulin resistance and diabetes numbers. People need to get a bigger share of their energy from protein and fat, not less.
My point was not specifically that people up north eat an animal based diet but that ALL people everywhere do. We have been hunter gatherer for a couple of hundred thousands years, and farmers for a couple of thousands. To me it makes sense to look at what hunter gathers today eat (i know its not a perfect proxy but better than observing what people eat at McDonalds). As you can see, ALL present hunter gatherers eat animal foods, and they eat a lot of it. We need fat and protein, they are essential for us. Carbs are not.
“An alternate explanation could imply that people who live in the tundra do end up aging faster as a result of actual deficiencies.”
Do people age faster up north due to deficiencies? What is you source for that?
I am amazed how “they” managed to dupe almost everyone with the idea that humans are “made” to eat diets based almost solely on grains and tubers (high glycemic index) and that animal foods (glycemic index = 0) was never meant for us. Some food for thought:
(Source: https://www.sciencedirect.com/science/article/pii/S0002916523070582)
It would be much appreciated and very interesting if the moderation/editor team here at LessWrong could share their perceptions on how frequently AI is currently being used for posts and comments.
Your “Intrigue → Denial → Appreciation → Denial 2 → Fear”—cycle really hits the spot. I will filter future comments from AI pundit through this lens.
Thank you for an excellent post.
The results and studies discussed in the post further validate a feeling I have had about longevity for some time: that there is not much a person living a “normal” life with decent eating, exercise, social, and sleeping habits can do to significantly extend their lifespan. There is no silver bullet that, from a reasonably “normal” health baseline, can routinely give you 5 or 10 extra years, let alone 1 or 2 years.In this regard, current science has failed, and I think the whole longevity research community needs to reassess the way forward. No matter how much pill-swallowing, cold-bath-taking, HIIT-training, and sleep-optimizing they (we) do, it does not really work.
I wonder, in the aftermath of the US bombing of Iran’s nuclear enrichment facilities, how this will affect future possibilities of performing a ‘Yudkowskyan strike’ on datacenters. Will this effectively drive advanced AI development underground to nuclear and bunker bomb proof locations?
The book will probably be a treat to read, but since “someone” (Open AI, Anthropic, Google, META et al.) apparently WILL build “it” everyone WILL die. Oh well, we had a good run I guess.
Design of the frontpage: No. I am fan of the white clean design with the occasional AI generated image. The current iteration reduces readability a lot.
“Employees at OpenAI believed…” — do you mean Sam Altman and the board?
If this information is accurate, it speaks volumes about how flawed their alignment predictions might also be. If a company with vast resources and insider access like OpenAI can’t predict the capabilities of competing firms (a relatively simple problem with objectively knowable answers), how can we expect them to predict the behavior of advanced AI models, where the unknowns are far greater and often unknowable?
Since a lot of people disagree with this, please tell me what a score of 100% mean or say 50% or 37%? I am not writing this to provoke, I am genuinely interested to know.
Feedback: I really like the breakdown of each companies stance on safety, but please skíp the percentage numbers. Its just silly to give a score based on guesswork.
Brace for impact. Everyone with a desk job should start to plan for alternative exit strategies.
Angry that doctors had spent years teaching her to delay treatment by dismissing her concerns.
Sorry for your loss, but thank you for reminding us how precious life is.
The quoted sentence from your post is I believe the main reasons why doctors (doctors not surgeons) will be one of the first high-status professions to be replaced by AI in a couple of years . If you can just get comprehensive blood work done (you could draw blood in a drop-in booth in a mall and send to a lab) and my some pictures taken of your body and then have a conversation with an AI about the symptoms you experience, the need to go to a psychical grumpy stressed doctor is no more needed or even something you would like to do when you will get much much better and consistent results from the AI doctor who treat you with respect and dignity and do no check his or her watch every minute.
And should we still think that a mere text message could increase yields for the hundreds of millions who work in farming?
Maybe. If you can use your phone to take photos of soil samples for image analysis paired with data which crop you intent to grow and weather data, why would it be far fetched to assume that an AI could give not you good advice on irrigation and fertilizers etc via text messages?
When I browse X for Astra examples I realize we are still in the boring “LP to CD” era. Most of the stuff is about recreating old stuff “just because we can”, and that not many people (hardly anyone) actually have genuinely novel ideas.