gpt5

5 ms·

31.
by gpt5 1mo ago
Both versions of DeepSWE (1.0 and 1.1) are likely not that meaningful anymore. Whether through models progression or through contamination.
32.
by gpt5 1mo ago
We should just consider the pelican bench as saturated and mostly meaningless.
33.
by gpt5 1mo ago
The pause on visa appointments does not apply to H1-B and L1 as they are nonimmigrant visas. It applies to family based immigration visa, diversity visa, and EB-1/2/3 when you do processing abroad. (You might have known that, but
34.
by gpt5 1mo ago
They tell you, and allow you to opt out in certain plans.
35.
by gpt5 2mo ago
I think the average person is definitely intelligent. It's just that we are very prone to mixing intelligence with knowledge in addition to also forgetting how long it took us to learn things that are so automatic for us we see them as
36.
by gpt5 2mo ago
That's just whataboutism that ignores the order of magnitude scale differences of the subsidies that were provided in China in the original comment.
37.
by gpt5 2mo ago
Apple makes their own connectivity chips. It was a big blow for Qualcomm.
38.
by gpt5 2mo ago
UniFi UDB-Switch has throughput aggregation for 5GHz + 6GHz.
39.
by gpt5 2mo ago
Whatever you think the fourth amendment is supposed to do, doesn’t match reality. Despite your ad-hominem attack.
40.
by gpt5 2mo ago
That’s not true. US citizens maintain their constitutional rights at the border. That has been tested in court. This is not the case for non-US person.
41.
by gpt5 2mo ago
To be fair, it doesn't seem like the post here is attracting high quality and thoughtful comments either.
42.
by gpt5 2mo ago
I'm sorry, but the Fourth Amendment does not protect against the intentional destruction of evidence. He could have refused search without a warrant, intentionally destroying evidence is a different legal matter.
43.
by gpt5 2mo ago
I don’t get the sentiment of classifying it as a felony. OpenAI’s model found security breaches in HugginFace’s system (it wasn’t even OpenAI running it, as it was a 3rd party evaluation company that didn’t secure it well). OpenAI collabora
44.
by gpt5 2mo ago
A little meta - I want to point out demagogic/populist comments like these that try to clear all nuance and brush a topic in black and white tend to come from a really small portion of the users here, but the same user (whose account i
45.
by gpt5 2mo ago
I don’t think it’s 90%, but also - fake content doesn’t apply just to LLM generated content. When you go the places which are clearly at the target of bots such as commercial products, geopolitical issues, and politics - the ratio becomes b
46.
by gpt5 2mo ago
You are mixing supporting AI (which HN is mostly against, but more supportive than other places), and thinking AI is dumb, which HN has been pretty persistent on for years. I mean - just looked at this downvoted comment essentially saying t
47.
by gpt5 2mo ago
I mean, the AI skepticism is still so prevalent today in HN, despite it solving decades old math problems, hacking into companies, making software engineer no longer code, etc. (all this just in the last year). So nothing really has changed
48.
by gpt5 2mo ago
Data centers don't really cost less in China. Perhaps even more due to trade restrictions and the difficulty of smuggling the chips in. Electricity is a small part of the bill.
49.
by gpt5 2mo ago
It actually is showing in public benchmark if you know how to look for it. For example, in Terminal-Bench 2.1, GLM 5.2 received 78%, while GPT 5.6 Sol received 88%. Then Terminal-Bench 3.0 came out (where the questions are new), and GPT 5.6
50.
by gpt5 2mo ago
I noticed it myself. It's not just that the models are faster, but that by becoming more capable, they can take on larger tasks, which require larger compute.
51.
by gpt5 2mo ago
That's not true. An agent in a loop can test itself, review, verify and iterate as much as needed. That's one of the primary reasons more capable models tend to have a higher success rate. I don't disagree that multiple tests
52.
by gpt5 2mo ago
A smarter model would know how to communicate with you correctly, and not just throw jargon it has just invented at you without explaining it.
53.
by gpt5 2mo ago
What was the change?
54.
by gpt5 2mo ago
You can see the Pareto Frontier well in DeepSWE's chart here - https://deepswe.datacurve.ai/ ChatGPT 5.6 Luna on the right (cheaper) cover most of the frontier, with a point for Deepseek flash, and higher performance o
55.
by gpt5 2mo ago
It's private companies making a bet on the AI boom. What governments do (which is the actual content of the article) is betting on economic growth to exceed the growing debt.
56.
by gpt5 2mo ago
Poor countries have a significantly higher rate of corruption, which means that the government itself is part of the crime syndications. It is true that we see correlation between richer countries and lower level of crimes - across all la
57.
by gpt5 2mo ago
Funny how you said "it turns out" when the whole point is that we don't understand LLMs - we just empirically see what they are good at. Claiming that you understand LLMs is similar to saying that you understand how our biolo
58.
by gpt5 2mo ago
They definitely are - the OP claimed that we are reaching "the end of the road for LLMs", based on absolutely no data and some handwaving on pareto distribution. We absolutely don't know enough about LLMs and intelligence to
59.
by gpt5 2mo ago
iPhone are in practice more reliant on central servers than ever before. Except for some games, if you take a random person's iPhone it becomes almost useless without internet connection. Which is exactly my point, it's not about
60.
by gpt5 2mo ago
Most people are already used to rely on the internet on basically everything. At best, they download a tiny chunk of entertainment from it when they go on a plane, and as soon as they land they immediately abandon that offline chunk. In add

More ›