“ Claude Fable 5.1's writing is generally a step up from earlier Claude models, with fewer stock phrases and less unexplained jargon. In some cases, though, its prose is denser than Claude Fable 5's: sentences run longer and there are fewer paragraph breaks.”
I cancelled my pro max Claude subscription last week; codex is much more succinct. I am curious if this is getting better.
I don’t think Anthropic realizes that humans have a token limit too and it can be exhausting to read Claude’s output. Prose density is not the same thing as succinctness.
One thing I've noticed and HATE, is that when you increase thinking-effort, that seemingly increases response-length. Meaning that X.High is longer than High, which is longer than Medium, etc.
Which is kind of the inverse of how people work; a really smart person can condense difficult ideas into simple[r] terms. Whereas people who struggle speak a lot but say very little.
High/X.High do seem to deliver better quality results, but it sometimes feels like needle-in-haystack extracting that from the word vomit.
With LLMs, you're still mostly read things "off the tip of the tongue". A better comparison is observing a smart person talking to themselves while working on a tough problem.
EDIT: also there's a reason the dial is called "effort", not "smarts".
I don't think smart people generally solve problems by talking through reasoning steps at a mile a minute. They clear their mind and let the solution come.
Of course I don't know if there's really a way for this to be molded in current LLM's (sounds more like diffusion)
It's so bad I made my own chat client for Claude, so I can attach steering prompts in conversation. They are applied just at the end, before the last LLM response, then removed and response kept.
I just go over the comments with Gemini 3.1 Pro at the end which has a much more normal "voice" and it doesn't lose nuance as a cheap model would. I don't care so much about what Claude writes during the debugging as I just do all the cleanup at the end instead of at every commit.
The higher the effort the more things Claude checks, and it's eager to tell you about all of them
See, this insight it had early on looked like a red hering for a while, but then turned out to be load-bearing. And that's not just a difference in semantics, it changed the whole conclusion (spoiler: it didn't). And Claude is very eager to tell you about this exciting journey
Sometimes Opus 5 (high/xhigh) feels like I'm dealing with the programmer equivalent of Zeno of Elea.
Every time, without fail, it would get me 90% of the way there and then leave a small note, exception, or deferral. When instructed to address that, Opus would somehow take nearly the same amount of time as the first 90%. And then it would finish with yet another deferral. Repeat ad infinitum.
You can sometimes get around it using the `goal` directive provided you are not subject to the constraints of mortality.
They got that from Anime seasons. Every prompt has yet another cliffhanger to keep you hooked. But the Season II story arc where Claude-chan fights the NsPasteboard boss battle on the journey to the UIViewMainController, I thought that was pretty intense. I guess I just gotta keep watching my terminal to see what happens to the main character input - rooting for him to survive the next season, but you know they always kill off the good input characters early.
Yes and the last bit is always mysterious and inscrutable. I have to think way too hard to figure out what the actual problem is. I’ve noticed it does a lot of explaining the mechanics of the problem it found, but almost never explains why it’s important until I ask.
And the worst part is that this little problem will keep sneaking into the context of future sessions, unless you spend the time to fix it. Even if it isn’t important, I’ll sometimes have Claude fix it so it will shut the F up about it going forward.
i think they took a huge bet that speaking like a ted talk was going to be a vast popular differentiator in their offering, i don't think they anticipated that people were going to make fun of it, that it could become a meme..that it could get in the way of getting stuff done and result in cancellations.
it's downright exhausting to read claude, the language style was a regression imo.
I wonder if I can make a tool for it to write messages back to me, say that it can only speak to the user through tool use, and then put a hook on that tool to prevent any of the Claude-isms
Me too. And it does it so often, that I've added a stop hook that detects "honest*" in its response and forces it to regenerate without the banned word.
"Humans have a token limit too" - that's so good and it explains so much of the fatigue that myself and colleagues/peers have about Claude in particular.
I think it's not just token limits - I think it's because it's so _dense_.
You get a week of research and debugging and testing compressed into a few pages. Even if it's explained well, it's just so much information.
And since it's AI, I'm constantly second guessing "is that really true?" and it's exhausting.
I've developed a habit of adding into my prompts "please keep your response concise and succinct" or "I'm trying to cram, please only provide the minimum level of technical detail necessary to understand this topic"
I find it helps immensely but it'd be nice if I didn't have to do that.
I don't understand that complaint, although it seems to be a common one. The whole problem with the way models talk nowadays is that they are succinct to a fault, going to the extent of coining new buzzwords and misusing existing ones. What I want to see is a shift towards plain language.
why so many people add 'please' when asking machine to do something? Was there actually research that when you SCREAM or curse it follows your instructions better?
P.S. Although my wife insists that I should stay polite in case AI overlords remember how I treat them ...
I'm polite to LLMs. It's not for the models it's for myself. If I start being rude to models then I might accidentally start being rude to other people as well.
Probably because polite people are already in the habit of saying please when typing out requests in chat. We're not consciously thinking about it, regardless of whether a human or machine is on the other side.
Not to go all ying/yang about it, but just to give a parallel: https://en.wikipedia.org/wiki/Loudness_war - you kinda need silence to draw a contrast with what's meant to be loud.
Separately, my boss confided in us that he's super abusive with his agent, wondering if we are too (no, lol). While I try not to read too much into this (which he doesn't make easy), I also can't help but not really notice a whole lot of amazing agentic delivery differences from his side. On the contrary, while the passion may improve his agent's performance, I'm not sure if it doesn't decrease his, upending the entire theatre.
I think about removing please/thanks, but then I accidentally add them back in during some edit/rewrite of the prompt... It's just how I'm used to asking for things
It only sticks to the instruction for maybe 3-4 turns. This is why when Anthropic released "concise output style" feature in claude code, it basically spams the model's context with "be concise" system reminders every other turn.
ya, but now i have this system that ive build that works with any ai agents, putting boundaries, gates every single time and it generate memories from the runs so it can inject them as needed.
> Prose density is not the same thing as succinctness
Can't agree with you more. I review 2-3 PRs a day from my team of eight data engineers. Most of my team members use Claude to write SQL, dbt and Python code. Some of them use Claude a lot, some less so. I can easily tell when I review the code that is mostly Claude generated vs. the one that is not. In dbt models where we have a lot of biz logic in intermediate layers, that's where I really have a difficult time following Claude-generated comments. So much jargon copied over from other adjacent dbt models (yet inconsistently), and the prose is super choppy (for the lack of better word).
After reading a looooong sentence/comment line, I still can't figure out what it really means. Had to always re-read the line 2-3 times (sometimes, more) to sort of understand. Reading code, however, is so much easier and usually, I just skip to reading the code and then come back to the comments. :D
Amen. I would trade some stupidity (say ten points on any benchmark) in exchange for a version of Opus or a similar model that actually gave me direct, concise answers.
You should try setting claude code to opus 4.6. With the style instructions I set in my user CLAUDE.md it does exactly that. It's like night and day: Opus 5 gave me a page and a half of word-vomit, yet the exact same task and prompt with 4.6 and I got maybe 100-150 words total, entirely readable.
x2 on opus 4.6. still works great, and it's fast. opus 4.6 is where i hope local llm's get to someday, that's kind of my personal benchmark for where "local is more than good enough i dont need these idiot large-scale service providers"
Yes, and they will work ... for like two turns, after which Claude will go back to its usual wall of text.
And yes you could add context (memories, rules, CLAUDE.md entries, etc.): they won't help (for long). Same for hooks that remind Claude to be concise: it gets "attenuated" and starts ignoring any such instructions quickly. There's also writing guidelines ... but they're basically just more context with slightly higher weights (ie. Claude will still ignore them).
I've even gone so far as to make a hook that identifies long responses and requests shorter versions (which is challenging in itself, as you need to run another lower-powered model to evaluate how long is "too long", as what's "long" when the expected answer is one line is different from what's expected for a ten line answer). However, that just shows you the long version, then some hook text, then (10-15 seconds later) it shows the short version. So I created a proxy that hid the long version/hook text for me ... but I had to abandon it because all that used up so much usage I was running out.
I'm fuzzy on the details, but Caveman somehow "hacks" Claude in a way that gets past all that ... but it takes things too far in that direction, with "cave man" speech that sucks.
I switched to using Codex for the last two weeks, and while the prose has been better, there have been a lot more technical oversights. I'm now having fable review codex commits and it finds deep issues. I'v also done the reverse where opus/fable do the work and then I have codex revise all of the prose prior to reading anything myself. This has also been effective; I'm not sure which is the better approach.
Just the other way I was thinking that if I asked "What does Lamborghini do?" the only correct way to answer is a single sentence "Which Lamborghini are you referring to?".
But LLMs will fail at this question: they will tell you about Lamborghini's latest car and mix some history in it. Just try.
Which is the wrong answer anyway, because there's at least two major companies called Lamborghini, one making cars, one making agricultural equipment and at least one famous person (Elettra) with that family name.
This very simple test/question makes me realize how much do I hate LLMs in a sense: while I agree that the answer it gives is the most plausible for 90% of the users, it's ultimately both wrong and long. And that 90% compounds.
But there's no "correct" answer in my eyes than "who are you referring to?". Possibly without listing all the possible Lamborghinis.
This is ... unnecessarily pedantic. Anybody in my social universe who asked me that question would undoubtedly expect "they make cars".
If you're picking nits, why not focus on the word "do" and (wrongly) expect an answer like "Lamborghini (either of the two main companies of that name) does not 'do' anything - the companies employ humans who 'do' things. Lamborghini is a legal entity established to allow humans to 'do' things, such as make cars, or agricultural equipment."
Shared context is a thing. Reducing every conversation to first principles is not always required. Get a grip.
My biggest frustration with Anthropic with Opus being too verbose is that they tried to put this on users. It’s pretty clear that Anthropic employees don’t use the day-to-day models that everybody else use. They have access to the next tier model so they don’t see the problems that everybody else is dealing with.
You can change CC's output style (https://code.claude.com/docs/en/output-styles). You can also put style notes in your global claude.md. I've instructed claude to treat me like I have adhd, get to the point, and be succinct, ... More or less eliminates the problematic prose.
I took time to figure this out after Fable spat out "...then stays purely as cascade-debugging provenance rather than load-bearing arbitration."
My experience with output styles for long-running sessions is that Claude starts to forget the terse output style by the middle of the context window. Obviously I don't know if 5.1 suffers the same fate but I ran into this issue with both Opus and Fable 5
That sentence is fine; it’s tolerably annoying. As a long-time HN reader, HN is full of this kind of performative erudition and I’m already used to it. Fable probably learned from the worst parts of HN.
I just used it to do a review of a ~100k SLOC codebase that Fable 5 / Opus 5 largely built, cost like $2 and caught some good stuff, but more importantly, it communicated very directly and was pretty light on bizarre metaphors. No "let me read the source before opining" type verbiage launched at me. Honestly night and day for me vs before.
Same, currently on a mix of Kimi Vivace (K3), GLM Max (5.3 and 5.3 Flash) and OpenAI Max (Sol and Terra mostly).
I will say that Kimi feels nice but slow, GLM feels faster but has limited tokens (even off-peak) and OpenAI is nice and fast but has limited context (258k shows up in Codex, really).
Neither of them are perfect, but I prefer their type of prose across the board to what Opus 5 and Fable 5 kept outputting. I'll probably check out Anthropic again in a year, but for now I need a break from its brand of slop. Oh also all of the other ones allow usage in OpenCode with their subscription plans.
> In some cases, though, its prose is denser than Claude Fable 5's: sentences run longer and there are fewer paragraph breaks.”
This sentence reads like Claude wrote it. Perhaps it did, or perhaps Claude has learned to write like the folks who work at Anthropic?
(Had I edited this, I would have said that a colon is not the right separator here. The second clause does not _explain_ the first, per se, bur instead expands upon it. Consider instead: "In some cases, however, its prose is denser than Claude Fable 5's, with longer sentences and fewer paragraph breaks.")
Today, Opus talked about "rotation slabs" in relation to logging. (and not log rotation). I didn't even bother asking what that was supposed to mean and switched over to Sonnet.
Same here. I still have access until my account churns but Anthropic has huge issues comparative to everyone else with token / usage burn down. K3 Swarm also delivers better results than Fable at a fraction of utilization. The Pro plan is definitely not worth it anymore and if I do want to burn some money I can always just leverage the API. But Anthropic went from simply amazing last year to a dumpster fire in less than 6 months for my use cases, anyway.
> In some cases, though, its prose is denser than Claude Fable 5's: sentences run longer and there are fewer paragraph breaks
that feels like they just blocked words like load-bearing but can't actually fix the real problem. The insane word slop density and run on sentences was the real reason it became annoying to work with claude, colored with way too many analogies and pointless linguistic comparisons.
You know you can just... change this right? What a weird thing to cancel over. I have in my global CLAUDE.md something to the effect of I have ADHD and give me succinct responses with headings and lists etc, works great.
Big ole pool of very fast ram that can be accessed by the CPU and GPU. Lets you run larger models. AMD does the same thing with Strix Halo. I have a 128gb machine at home, and have had difficulties running 120b models, but 70b and below run pretty well.
It’s more that it’s a very parallel architecture than fast.
The LPDDR5X under the hood is slower than what you’d find in a graphics card vram of years ago
That’s why you get consumer macs with 512gb while GPU makers are reluctant to give you more than 16gb unless you pay dearly. It’s not the same kind of mem
> It’s rather amusing to me to read comments like this, and then simultaneously whenever a Chinese company or team releases open-weight models or whatever there is a giant round of applause, America is so behind, and there’s nothing but positive things to say about the intelligent, creative, and well-intentioned Chinese engineers
It is absolutely applied accurately. You're commenting on the alleged hypocrisy of people simultaneously criticising American open-source while praising Chinese open-source, and then attributing your perception of hypocrisy to the website as a whole. The reality is the behaviour you've observed comes from completely different individuals, not some kind of HN hivemind. Your comment is such a typical case that it could go in the wiktionary page as the example excerpt teaching people what the goomba fallacy is.
> My normal bill for running this whole site is around $90 a month. During one bad spike month, it jumped about 500%.
This is D1 - which has very surprising costs. you may just want to drop D1 and move to a static site. There’s no reason your site should cost this much.
There's a lot of people who host their site at extremely expensive places and then do everything they can to minimise unneeded traffic - instead of just moving to a cheaper host. Vercel is another popular extremely expensive host.
I'm always flabbergasted when I see what people pay and how much effort they need to invest to keep their cloud websites from eating them alive.
My allegedly more complicated VPS stack needs an afternoon of attention every two years when a new Debian major release is necessary, and costs have been predictable for 15 years, no matter what happened traffic wise.
The predictability of costs is underrated too. You don't want your hosting solution to automatically scale up to $20,000. You want that if it's overloaded it's simply overloaded.
Thank you! I need to tighten up my KV compression, which is actually carrying a lot of D1's load otherwise. We also had some bad queries some months, as the pages and database grew, that were counting the wrong things (or extremely inefficiently) and those have since been fixed.
For personal stuff, I use it with AnythingLLM. It replaced any Google search for me. For coding, I run opencode though I have been debating switching to Pi. I would argue it’s at Sonnet 3 level.
I would recommend a registrar that would explain to the customer why they would not want a .me domain for anything critical unless the person lives in Montenegro and trust the Government of Montenegro to maintain a good and trust worthy registry.
Otherwise just use which ever registrar is cheapest and who you think will handle any quirks or shenanigans that registries may do to domains you own, and which own system and processes hold high enough standard for you.
ccTLD hacks are both semantically incorrect and geopolitically unstable.
The .io TLD will likely be phased out in the future due to geopolitics, and all the companies who decided it was more important to signal how hacker jargon aware their startup was will have to go through the very difficult process of changing domains.
In order to log into IRS.gov to get a code to pay my USA taxes, I had to verify my USA ID via a private company called ID.me, whose domain name AND company name are now forever tied to the whims of the government of Montenegro.
The biggest difference is the role of ICANN and their willingness to regulate the management of a TLD. With ccTLD they have an official policy to be hands off and not dictate what a country will do with their top level domain. If the US government would start to mess with Verisign and how the registry handles domain, then ICANN is within their own policy to just move it somewhere else.
The government of Montenegro could, just like many other ccTLD, decide tomorrow that every registrant must be a citizen of Montenegro. Many countries do this today, and it is no big deal because it is the country of and ccTLD that dictate how their domains should be operated. They can raise prices by a factor of 100 if they wanted, decide on some form of ID for registration, or dictate that you must have a company located in the country. ICANN has no objection to any of that.
ICANN is a nonprofit in California. It doesn’t really have any power on its own. It only exists because everyone from browser developers to server operators to consumers agrees we need a central place to organize stuff.
If the US justice system issued a warrant telling Verisign to do X, what do you think the courts would do to ICANN if they tried to actively stop X from happening? At best, they would be politely asked to stop, at worst, they would get felony charges for obstruction.
It's very different. If the USA soon starts cancelling politically inconvenient domains, European ones will be safe. Just like Nazi propaganda domains would be censored in Europe, but are safe in the US.
Every domain has a country. It's as if every non-ccTLD was actually underneath .us. For legacy reasons .com .org etc were grandfathered in. gTLDs are also under .us for corrupt reasons.
Politically inconvenient domains are already routinely seized by the US and there are international agreements in place that allow cross-border takedowns.
Why do you think t.me got taken down? Montenegro doesn’t care about Telegram.
Never use the same company for your hosting or CDN and your domain, and avoid cloudflare in general.
That's because if they don't like your website being on your CDN, and they suspend your account, you'll lose your domain. If your domain is at Porkbun you can change it to point to a different IP address.
And avoid Cloudflare because they're centralising the internet.
I chose Porkbun because it's a small company with good prices, a good vibe, and all the tools that I need. Cloudflare was never going to be on the table because I don't want to feed the beast that is already swallowing the entire internet.
My understanding is that both GoDaddy and Namecheap used to do domain front running[0] at the time I was registering my first handful of commercial domains, so I've always avoided even using their search engines.
My bank automatically blocks payments to Namecheap. When I had domains with them, I had to call and give prior approval for the exact amount I would be paying. My bank claimed it was because of a high number of fraudulent charges.
Namecheap got bought by private equity fairly recently, so i switched away from them. Wouldn’t recommend starting with them just in time for the enshittification to start.
Their prices had already been going up for a few years before that, which finally pushed me off them starting around August of last year. I'm about to swap my final few domains over this month before they renew.
Porkbun has been great so far. Easy to use, refreshingly minimal, and good prices.
FYI - Gandi was great, but they got bought by private equity a few years back and the price skyrocketed and service went downhill super fast after the buyout.
Never had any trouble with them, but also moving away from them is unnecessarily hard (the code sometimes takes a day to arrive) and they cover the entire interface with their paid hosting stuff which makes them a poor registrar. I ended up on them because of Google Domains selling off but got off them because very annoying to use.
I'd honestly be careful with squarespace. They are owned by private equity, advertise on countless YouTube channels, and at the same time their core market is under a looming threat from the AI companies.
You need your domain registrar to be stable and predictable. Their profile is not that.
reply