I occasionally read articles on Nate Silver's substack but I'm still missing the breadth of 538conbined with the trademark data-driven analysis.

Those are not contradictory: a company's inference costs can increase due to deploying more models (Sora), deploying larger models, doing more reasoning, and an increase in demand.

However, if we look purely at how much it costs to run inference on a fixed amount of requests for a fixed model quality, I am quite convinced that the inference costs are decreasing dramatically. Here's a model from late 2025 (see Model performance section) [1] with benchmarks comparing a 72B parameter model (Qwen2.5) from early 2025 to the late 2025 8B Qwen3 model.

The 9x smaller model outperforms the larger one from earlier the same year on 27 of the 40 benchmarks they were evaluated on, which is just astounding.

[1] https://huggingface.co/Qwen/Qwen3-VL-8B-Instruct

For standard questions I feel like it doesn't matter too much what you use. When it comes to multi-step searching + reasoning flows like look for alternatives, fetch pricing, feature lists, compare etc, the differences are larger because of the engineering glue and prompting around the pure LLM inference which makes the tools more or less powerful.

Moderation for sure helps, would there be ways to make it scalable with less manual supervision? Or a system that would organize people with certain rule-sets to distribute them into suitable sized groups?

I do agree with your statement that "Good discussions evolve naturally and also randomly", let's say now your platform becomes popular. It will attract players that will want to exploit that either to sway opinions for their own gain, and I believe that this is becoming increasingly cheaper to game and simulate whole crowds. So the limits are mostly with this in mind.

Indeed perhaps the term social platform is vague and different "optimal rules" could be different for social platforms that is a mega-forum, a network for friends, or just generic post sharing.

I'm wondering if there is some sort of taxonomy of these rulesets or levers that exist? Or a review paper on what has been tried and what effects they had? There are so many possible ways to structure online social interactions.

I wonder if it would be possible to simulate this to understand what behaviors will emerge if you set certain types of rules. It is certainly difficult to create coherent personalities with LLMs that act in realistic ways but I wonder if one could get an approximation.

Perhaps what I have in mind is also not best described as "pleasant", but also something that is net-positive for society, where as a whole society is better off having that than not. This is arguably the case for HN but not necessarily for some of the bigger ones out there.

- Desirable: Most people actually want to be on it and find some use or pleasure in using it

- Non-toxic: I also added this one because some people might enjoy being on a toxic platform, this is not what this is about

- Less exploitable: Difficult to manipulate which is increasingly important in the age of cheap LLMs, but can also be a tradeoff with desirability as barriers are erected to prevent bot manipulation / vote brigading.

Taking Hacker News as an example of a desirable, non-toxic and less exploitable social platform. I believe several attributes make it so:

- Voting: Upvoted content rises up, contributes towards desirability / non-toxicity of the content

- Strict rules / moderation: Keeps the content on topic, constructive, friendly, more pleasant to parse. Contributes towards desirability / non-toxicity and also makes it less exploitable as manipulation can be detected.

- Novelty / surprisal: This third one is somewhat special, it is less a mechanistic property of the platform but a content focus choice. It contributes towards desirability but I believe also towards lower exploitability: it is more difficult to fake novel or thought-provoking content.

Now I do realize I could have phrased the target properties differently and semantics can always be discussed ad infinitum so take the spirit of what I mean rather than the exact wording.

What I'm specifically asking HN here is what sets of rules / mechanisms could have the above as emergent properties?

Is Hacker News a desirable, non-toxic, less exploitable platform also because it focuses on novel, thought-provoking content or could there be mechanistic rules that would allow having these properties for any kind of social platform?

I'll include some possible mechanistic rules that crossed my mind that each have their flaws:

- Member verification (ID / Credit card): Less exploitable but likely very undesirable for many

- Vouching: Start with a kernel of trusted members, include only members vouched for

- Contribution limits: Members can only contribute / vote n times per day / week

- Active discussion limits: Not everyone involved in the same conversation e.g. have two people discussing a topic, have a system for others to "raise hand" to participate in the conversation

- Exposure limits: Your post can never reach more than n random people, it has to be actively reshared by someone to spread further.

Comments URL: https://news.ycombinator.com/item?id=46312236

Points: 4

# Comments: 5

https://www.youtube.com/c/Charlesenv%C3%A9lo

Typically LLMs don't produce usable embeddings for clustering or retrieval and embedding models trained with contrastive learning are used instead, but there seems to be no mention of any other models than LLMs.

I'm also curious about what type of clustering is used here.

However I don't believe that PG's article meant to address the elephant, but rather was a meta-level thesis on how he sees debates being shut down by orthodoxy, and for that he does suggest what he thinks would be a possible solution.

Perhaps the thesis could have gained in being more balanced to as you say "avoid giving tacit permissions for the extremists on the other side"? On the other hand, does one always have to shield one's expressions with disclaimers and is one not free to share thoughts however raw in order to express, discuss and learn, update our beliefs?

There likely is a bigger responsibility when one has a larger audience to avoid misinterpretations, but ultimately I believe as long as there is a rational and nuanced discussion to take the good points and have a productive debate, it should be okay.

How can we create incentives to have a more nuanced discussion?

Having read both essays I don't see them necessarily in disagreement. pg criticizes the performative and orthodox nature of some social justice activists' behavior, however it doesn't seem that the author's behavior here is performative at all.

Perhaps we should just avoid these terms like "woke" and just say what we mean to avoid this societal dissonance? I feel like decent rational people can talk past each other depending on how they have been exposed to the term.

This bias can also exist for lower-paying jobs, however I would guess proportionally there might be more 80-hour/week type high responsibility jobs in the higher paying brackets.

Do Large Language Models learn world models or just surface statistics? - https://news.ycombinator.com/item?id=34474043 - Jan 2023 (174 comments)

Wouldn't this mean that capital is flowing in which should lead to more hiring? Is the job market response delayed or are there other factors?

Error handling when you get rate limited, the token has expired or the token length is too long would be problematic, and from a security point of view it requires your DB to directly call OpenAI which can also be risky.

Personally I haven't used that many Postgres extensions, so perhaps these risks are mitigated somehow that I don't know?

RepRisk's goal is to drive transparency and accountability of company practices. As a leading ESG data provider, we monitor media reports worldwide to provide a comprehensive overview of companies' behavior.

RepRisk, based in Zurich Switzerland, has been in the ESG business since 2007 and values intellectual honesty, humility and openness.

We are looking for a talented CTO to drive the technical vision of the company.

Apply at : https://www.reprisk.com/careers/chief-technology-officer-zur...

You can accumulate the gradients of multiple batches before doing the weight update step. This allows you to run effectively much larger batch sizes than your GPU would allow without it.