JacobAsmuth

Marco econ 101 disagrees :) Ask your favorite LLM to explain the "yield curve" to you

How has your experience been with Gemma 4?

You need to compare cost per task buddy boy. Cost per token doesn't tell you much when you don't know how many tokens a model will use to accomplish a task

Could it be that they have to serve their models to billions of users?

Just you. This is typical market competition in a fast moving field.

(Rate limits * capability of the model) / cost

It's also about 15x faster.

Is Kimi 1/5th the price? Artificial analysis has it at the same price as GPT 5.6 Sol while performing worse.

This is huge. I've always wanted to use Git in the terminal but never been happy with the underlying language that other TUIs were written in. Now that I know that I'm using developer-managed memory I can be much more comfortable and confident changing between branches, pushing, and pulling, and even merging code. Thanks Simone!

Agreed 100%. This guy thinks there's a limit on the demand for intelligence. You think that Fable 7 which can run a billion dollar corporation on its own has no consumer demand just because we have fable 5 at 9k tok/s? Who do you think will be the biggest customer of such a model? Fable 7, obviously.

The assumption that LLMs can't answer this question. You can pay approximately $0.000001 to correctly answer this question in approximately 50 milliseconds using Google's cheapest model.

Are emotions actually salient parts of text or is that just because you have a human brain which is tuned to recognize emotions?

You're using "I find it easy to recognize emotions in text therefore it is a simple task", but we know for a fact that some tasks which are easy for humans are hard for LLMs, like counting objects in an image, while other tasks are easy for humans and easy for LLMs, like adding single-digit numbers.

It's not readily apparent to me that precisely modelling emotional state of the characters in a piece of text is the second and not the first, which you seem to assume. In fact, the work as presented seems to indicate that it's a class much closer to the first.

Huh? This is from May. You guys should read the more recent notes. https://hypr.land/news/update56/

Oh, you're talking about "AGI"! In the 90's we started using the term, you should catch up!

Not a single LLM will do that. Why lie?

Would you expect to find concepts related to emotions evoked by thinking those thoughts present? Or meta-descriptions of thoughts? Sounds like you're just post-hoc rationalizing to me.

Gemini 3.1 Flash-Lite with web search disabled through the API and zero thinking gets it just fine.

Hate to break it to you but Gemini 3.1 Flash-lite with grounding disabled (through the API) gets this first try, even with thinking disabled.

That would make sense if the author specified a specific year's "year group" - e.g. if they had said "I spoke to the 6th year group". But they said they spoke to "THE year group" - which is very confusing to me, even more so after your explanation.

First line of the blog the author states they spoke to the "year group" - what is that? I'm not familiar with the term.

I think instead they should sell super hackers and get their product banned instantly and go bankrupt

Anyone interested in this should check out my Qed project I've been working on, a formally verified web frontend. https://github.com/JacobAsmuth/qed

You'd say that Google Maps hasn't improved much in the last 20 years?

Sounds like Jane Street would be interested in the Lean framework I've been working on for formally verified frontends. https://github.com/JacobAsmuth/qed

Fable can beat pokeon red in 50 minutes with only visual input. https://www.youtube.com/watch?v=Ty_50J84fMY

This is explicitly not what Dario asked for in his blog. Care to quote his post for me where you feel that he asked for this?

No no we should push the limits until a bioattack happens, then when those people are all dead we comment angrily on the hackernews thread and say that someone should have seen this coming

Reading comprehension failure on display here from maxall4.

"Nobody I know is talking about censorship" is a certified HN banger.

How does this instillation happen?

Are they exploiting zero days to somehow force the phone to install an app unbeknownst to the user? Are they physically stealing the phone while the user is unaware, installing software, and then returning the phone?

I don't even know how spyware ends up installed on a modern phone without the users knowledge.