fakwandi_priv

For that claude problem in specific about 95% of the time it will follow the rule that a plan file needs to be made before any changes to the project are allowed. Haven't had to ESC it in a while now during a question, things definitely improved in 5.5.

Slightly off-topic to this model card but has anyone noticed changes to GPT6 Astra about a week after 6.1 came out?

I feel like 6 Astra just turned a lot lazier during planning and strangely enough claude has become a lot more assertive with Opus 5.5 while it used to be very lazy.

It’s really obvious during planning, it used to be that claude would just give a best effort guess and handwave things but now it will actually write small tests, launch playwright to test a hypothetical and even look things up in github during a question. This was something codex’s models were really good at.

I’ve read so many stories about people claiming models getting nerfed after a new release and now it seems I can finally join them.

Then that check accounts history and tell me if you see a pattern or not, this one is really obvious.

Because both are bots, to prevent other people from wasting their precious mental energy on reading text which may or may not contain useful information but certainly are verbose.

Edit: downvoting them now can also help prevent future astroturfing which these accounts are made for, at least I hope.

I tried to integrate conductor in a project I’m working on but when starting a workflow the idempotency key was ignored, so sending the same key twice started two workflows with different id’s. I found that pretty strange but after prodding claude and linking it the github project it decided to look through the code and apparently that request param disappears somewhere in the plumbing. Somewhat surprised the article doesn’t mention it. This was the self-hosted version, the cloud version might be fine.

I switched over to Temporal, and so far pretty happy about it.

> On October 16, 2026, the Ministry of Commerce and the Ministry of Science and Technology issued Announcement No. 48 of 2026, publicly solicited opinions on the revision of the China Export Prohibition and Export Restriction Technology Catalogue (hereinafter referred to as the Catalogue), and implemented temporary management measures for pre-export registration of relevant technologies. Recently, the spokesman of the Ministry of Commerce answered questions from reporters on relevant issues.

https://m.bjnews.com.cn/detail/1792137269129878.html

Apparently globally recognized top 10 models (so Deepseek 4.1 Flash and GLM 5.3 Flash) would have fallen under this export registration.

I recently started playing Expedition 33 and was impressed right away by the quality just in the first 10 minutes, hours later and the game just gets better.

I count 32 people on their team photo + the dog, and even this seems like a somewhat sizable number of people to manage. When I see news of these lay-offs it’s terrible for those people but at the end of the day I feel like those companies never should have reached those numbers. I fear for Larian

Most of this thread seems neutral/skeptical with just that post being obvious. On the other hand the volume of jev threads we have been seeing these past 2 weeks..

I’m a bit saddened this thread has so little activity, 55 comments as of posting, here even though there are half a million people watching the spacex stream. Just watching the launch, the amount of technology and people involved is incredible.

Prompting claude 5.5 has more appeal it seems.

> There is no additional fee for using prompt caching. Input tokens, whether served from the cache or processed fresh, are billed at the standard input token rate for the respective model.

So what he’s saying is correct, there is no separate cache pricing, which by normal standards should be 10% of the cost, which can become exceedingly expensive for anything other than single turn. The way they are stating this is of course strange..

I’ve also noticed this twice now in the autogenerated chat title, recent example: “Explore book locking liveewer”.

Googling it returns no matches but I think it was supposed to be “live viewer”?

What’s wrong with wiki.gg?

I have noticed the same thing starting with 5.6 when editing my last prompt inside the vscode codex plugin, I’ve seen the model’s thinking respond to the edit with a remark.

Slightly bummed out about it because in the past you could try different situations during a planning session and it wouldn’t pollute the cache but now it does. I’m not sure if forking the conversation has the same problem.

> According to Windows Latest, the high memory consumption is due to the fact that Weather is not a fully native Windows application. Instead, it is essentially an MSN Weather web app built on Microsoft's WebView2 framework.

I wonder how this approach is having such an high reduction in memory usage compared to what’s stated in the article. I would assume both use the exact same WebView here.

The only explanation here that I can think of would be that the “Add an application” starts it under an existing Edge process which shares it’s memory instead of completely isolated.

> - Math completely fails in longer contexts

Not sure what longer contexts we're talking about but didn't we have an old math problem optimized, which even the LLM itself was surprised about, just a week ago? Something which wasn't possible 6 months ago.

When I'm reading reviews of plans created by an agent especially on security boundaries it's suggesting huge matrixes to test even the very obscure situations, but then I'm also reading things like this and I just don't understand. Are we even using the same tools?

You're guesses are getting somewhat closer each time, almost there.

I've been an avid fan of codex for the last few month's but finally hit the weekly limit so I've wanted to try out claude code before biting the bullet and going for the 200 dollar codex sub.

Obviously in hindsight it would be unfair to Anthropic to judge them on an unstable day so I'l leave those complaints aside but I hit the session limit way too fast. I planned out 3 tasks and it couldn't finish the first plan completely, for that implementation task it has seen a grand total of 1 build log and hasn't even run any tests which already caused it to enter in the red territory of the context circle.

It was even asking me during planning which endpoints the new feature should use to hook into the existing system, codex would never ask this and just simply look these up during planning and whenever it encounters ambiguity it would either ask straight away or put it as an open question. I have to wonder if they're limiting this behavior due trying to keep the context as small as possible and preventing even earlier session limits.

Maybe codex's limits are not sustainable in the long run and I'm very spoiled by the limits but at this point CC(sonnet) and Codex(5.4) are simply not in the same league when comparing both 20 dollar subscriptions.

I will also clearly state that the value both these tools provide at these price points are absolutely worth it, it's just that codex's value/money ratio is much better.

So what was the word you changed?

AMD had an upgrade path with the 5700x3d, assuming you’re on AM4.

Just reading now that they went out of production half a year ago which is a shame. I was very impressed being able to upgrade with the same motherboard 6 years down the line.

I run the same config but it tends to fly through those commands on the weekends, very noticeable difference. I wouldn’t be surprised that the subscription users have a (much) lower priority.

That said I don’t go beyond 70% of my weekly limit so there’s that.

This is very unrelated but I worry about posts like these.

HN is a great place to get genuine thoughtful discussions compared to a big portion of the rest of the internet. Reddit used to be the place for finding genuine experiences for products with subreddits like buyitforlife (or more specific) but now these and other subreddits are filled with bots and marketeers promoting what they sell and hammering their competition in the same thread.

Some bots are already here but I fear when the marketeers come.

It used to be Hola VPN which would let you use someone else’s connection and in the same way someone could use yours which was communicated transparently, that same hola client would also route business users. Im sure many other free VPN clients do the same thing nowadays.

From what I can read Swift gives you a stack trace which is good. At the moment I’m using Go where that stack is only generated where the panic is triggered, which could be much higher up. Makes it a lot more unwieldy to figure out where an error happens because everyone uses:

> if err != nil return err

Might not be the best fix but other than disabling memory I changed the setting ‘ChatGPT personality’ to ‘Robot’ and I’ve always had straight to the point answers (so far).

> The total wall clock time for the build was measured. The smaller the better. I always did one build to prime the caches and discarded the first result.

The article is worth the read.

I have heard people on both ends of the spectrum:

- LLM's are too limited in capabilities and make too many mistakes - We're still in the DOS era of LLM's

I'm leaning more towards the the 2nd, but in either case pandora's box has been opened and you can already see the effects of the direction our civilization is moving towards with this technology.

The point the article is trying to make is that it reduces boilerplate, wouldn't be surprised if this gets added to TS in the next year of two.

Refreshing to see a potential conflict[2] get solved in a civilized manner, rare these days. Thanks for including it.

Why is this being downvoted? Seems like a valid concern to raise if you find two pieces of software somewhat having the same functionality.