today at 8:28 PM
The way Claude did it is fight entropy with entropy.
"Add a static composer into the HTML" <- This seems like something can be done with SSR?
"For faster navigations, we kept the composer mounted between conversations" <- Your SPA should cache this between pages, why fetching it every time? Or you need better routing for your react components.
"cheap first-character check before the regex" <- Should we cache compiled Regex instead?
I think even 1.3 sec to load the front page is unacceptable. Something need to be reworked from basics (SSR, chunk-based rendering) to solve the problem. Focusing on invidual benchmarks may miss the opportunity.
today at 10:10 PM
Even Next.js, for all that people are unhappy about its quirks, complexity, and random undocumented behavior and bugs (God help you if you ever want to try and actually use parallel routes as documented), can do all of this SPA stuff out of the box. Use `<Suspense>` appropriately on data loading, and everything static (including e.g. purely input-output components like editors) will load once and be re-used forever, and your Suspense-wrapped items will show a loading placeholder and only re-load when you intentionally re-trigger the data loading (which you can push down all the way to the level of individual buttons if you want).
today at 8:47 PM
The amount of complexity added for the gains is depressing. I am confident a human and about 5 minutes with chrome debugger would yield better results with a fraction of the complexity at a fraction if the cost and in a fraction of the claude baby sitting time.
Reading this shows the authors have a profound lack of fundamental understanding on how to effectively optimize in the web domain.
This isnt claude being bad but how wild it is watch people from the cutting edge of ai brag about pretty mediocre gains.
today at 8:59 PM
When people talk about AI being able to handle everything I keep wondering, have these people built anything complex, novel or serious? Just because people can see a website or a simple app improved, does that mean that all code can be handled by LLMs? It's like people are totally forgetting a whole category of careful, well-thought out programming for the critical parts.
today at 9:25 PM
Yes, what you are seeing are amateur developers that barely understand the tools they are using either giving LLMs poor instructions, or totally taking whatever it says at face value, then not bothering to put in further effort.
If they just kept prompting it, or maybe used a different thinking level, it could have identified and solved this problem. Sometimes an engineer would look at a system and say "the current approach isn't delivering the desired engineering requirements. Maybe we need to rethink".
Either engineer or LLM could take that sentence and run with it. OP of the article clearly can't do either.
today at 9:35 PM
You're missing the point where: in complex systems, sometimes optimizing code is both a high effort undertaking, and can totally not pay off. Having done hundreds of such exercises on our software over the years, it's liberating to have an idea of how to make something faster, being able to validate it without the fear of having to throw it all in the trash if it fails after days of work. What is still important is being able to provide proper guidance - we even built new tools to allow an AI agent to analyze memory usage in more depth, and instructions on how to benchmark in cloud environments where shared CPU usage and VM reallocation happen all the time and confuses the AI all the time with measurements
today at 9:04 PM
I visited https://claude.ai/ over a mobile tethered connection from my laptop the other day and was pleasantly surprised at how quickly it loaded.
(That said, I just had a look in Firefox and it loads 20.78 MB of JavaScript (6.84 MB compressed) so I expect they could make it a bunch lighter if they kept trying.)
today at 10:30 PM
A few times now, I gave Claude all my IronPython scripts and asked it to make to see if it could improve the code. It proceeded to break nearly all of them.
today at 7:59 PM
How about you make Opus 5.5 actually work?
I had it try to prepare a code review for me. Not only did it refuse, it refused to even tell me what the prompt (written by another Claude!) was. Why?
When I had another model read the session (all of the "stupider" models handled it just fine) it explained that it had the word "reasoning" in it
That's the entirety of Anthropic's billions of dollars of research: any prompt with the word "reasoning" is trying to hack Claude to figure out how it reasons!
A model like that should never have gotten out of QA, let alone been released.
today at 8:13 PM
I understand the frustration but shows a lack of critical thinking. Esp when you start with 'How about ..'.
This blogpost is about frontend performance. It'll be akin to you commenting on a swift blogpost saying 'How about Airpods noise cancellation'. Sure both are Apple, but they are wildly different teams.
today at 8:30 PM
This is a techie discussion forum.
In such a forum, it seems to me like it's fair game to point out that the company patting itself on the back about how great they are at programming (as evidenced in the article above about their 3x speed improvement) ...
... can't even make their latest model handle basic English without refusing to work.
today at 9:32 PM
Not that I’m defending Anthropic but OP is right, you’re complaining about the taste of a pear on a blog post about roses with the excuse that you’re in a biology forum. Sure, they both come from the same family, and sure, they both fall under the purview of biology, but they’re not the same.
today at 9:34 PM
today at 8:55 PM
This issue is mentioned on the Opus 5.5 post [1] from Anthropic (no idea if it has been added after your rant):
> Don’t ask it to show its reasoning in the reply
>
> What to do. Remove requests to reproduce its internal reasoning in the reply from your prompts and instructions.
>
> Why it matters on Opus 5.5. A request to reproduce its internal reasoning in the reply can be declined. It’s one of the flag categories.
>
> How. Ask Claude for what you need instead, for example, “Explain why you chose this approach in three sentences.”
today at 8:20 PM
Probably it thinks you're doing some sort of system prompt exfiltration/distillation attack. Also what even is the workflow you're trying to have it do? It's doing code review but you're having it read some other AI models prompt/session history? Are you doing code review or like session history retrospectives?
today at 8:00 PM
seems a bit weird to complain about the model issues in a post about the harness/sites?
today at 8:23 PM
> When I had another model read the session (all of the "stupider" models handled it just fine) it explained that it had the word "reasoning" in it
Did it explain it did it hallucinate?
today at 8:40 PM
This happened at the classifier level, there was no "Claude thought X about it (hallucinating or otherwise)": this was a glorified regex deciding Claude couldn't work on a prompt (a code review prep) because it contained a string ("reasoning") it didn't like.
It's more or less the same mistake we've seen Anthropic make repeatedly with it's brain-dead regex-based Fable/Mythos gates.
today at 9:00 PM
> because it contained a string ("reasoning") it didn't like.
How do you know? How would the stupider model know?
today at 9:27 PM
Because it’s in a page accessible from its search tool presumably https://claude.dev/blog/getting-the-most-out-of-opus-5-5/
today at 8:19 PM
I’ve had the same thing occur five or six times over the past week; they seem to be attempting to prevent anything resembling chain of thought extraction.
Every single time it triggered, it was due to a prompt written by their own model in a dynamic workflow. The self-serving nanny oversight has to go.
The fact that they label model distillation as an “attack” is genuinely hilarious after they “distilled“ their models from all of our work, and continue to do so.
I believe AI is here to stay and an incredibly powerful tool, but these companies, and especially Dario and Altman, are the very last people I want to see in charge of it.
today at 8:31 PM
> “distilled“ their models from all of our work
They distilled all digitized human knowledge and artifacts and they're now complaining about someone copying their outputs saying it's a "national security concern."
I'm not sure about how to classify that. Hilarious? Pathetic? Sad? Hypocritical? Hyperdramatic? All of the above?
today at 8:07 PM
[flagged]
today at 8:11 PM
Again, I used a slightly older model in the same series, Opus 4.6. It read the same prompt without any problem whatsoever. Also, a (non-Opus 5.5) Claude wrote the prompt in the first place.
Opus 5.5 literally refused to work OR EVEN TELL ME WHAT I'D "SAID" when it read that prompt.
Nothing to do with skill or the user at all: same exact prompt, three different models ... two worked, one didn't.
today at 8:10 PM
the skill issue is "having to use a cloud model to do work of any value"
might as well offer your life to a king to work in their fields.
today at 8:25 PM
> $500k engineer: [X] feels slow. Make it faster.
> Claude: On it... Done.
> $500k: Can you make it faster still?
> Claude: On it...
today at 10:22 PM
Claudehart's law says that when a measure is picked as an indicator, then Claude will start to game it.
today at 8:48 PM
You removed the load-bearing seams didn't you
today at 7:40 PM
This writeup legit coincidentally matches the asking-agents-to-make-code-faster-but-with-constraints-to-stop-agents-from-breaking-things writeup I posted on Monday: https://news.ycombinator.com/item?id=49803085
Front-end UI optimization is slightly trickier than optimizing strict algorithms, but I found that prompts to the agents to build tooling to track visual regressions are more than sufficient. The main issue (at least with GPT models) is that you have to be very explicit about the use of padding/margins/negative space.
That said, for my front end projects from scratch, I'm staying away from front-end JS frameworks and seeing how far and fast I can get with just HTML/CSS/vanilla JS shenanigans now that agents can wield them effectively.
today at 8:24 PM
Less a post about performance and more about their Claude tag product. I guess it makes sense that there's not actually a ton of technical detail being moved into given probably opus is the only one that knows what all the dragons were lol - but cool workflow I guess
today at 8:08 PM
I used Opus 5.5 today for the first time hoping the writing would be more bearable and it SUCKS.
Why can't they fix that
today at 8:10 PM
Opus 5.5 writing is much more concise than 5.0
today at 8:17 PM
but Opus 5 is absolutely terrible. I'm still on 4.8 when I use Claude. I still think 5.6 Sol is the best available model outside of Astra/Fable.
today at 10:31 PM
[delayed]
today at 8:18 PM
MAYBE, but the bar is so low... It's not that it's good in any way
today at 8:11 PM
The juvenile nonchalance with which some Anthropic employees seem to be talking to their AI (wacky, sick, cook, ...) is truly bizarre.
today at 8:23 PM
That’s just how people in their 20s casually communicate now. It’s mostly from tiktok
today at 10:13 PM
That's literally just how kids talk these days. Are you really that unaware of modern slang?
today at 10:00 PM
They've never had a job at a normal company
today at 9:38 PM
"You know what we want. Let’s go"
today at 9:55 PM
Perhaps they can hill climb a native app instead of the electron and react native nonsense we have today
today at 8:13 PM
Okay, now fix the WYSIWYG markdown parsing in the chat input!
Paste in a stack trace, then try to put it in a code block. Add a newline above, then add the opening triple backticks, then arrow down and add closing triple backticks at the bottom. Opposite congrats -- you have ended up with raw triple backticks at the top, plain text stack trace, and your cursor in a brand new code block at the bottom starting where you tried to close.
Realize you want to go put code span backticks around some identifiers you wrote out earlier? Best make sure to insert them in the blessed left-to-right order, or else opposite congrats again -- you'll end up with a mix of raw backticks and code span treatment for the text between your identifiers.
If Claude can discover novel CRISPR enzymes, surely it can make a rich text markdown editor, no?
today at 8:21 PM
Here's a simpler bug:
Claude Cowork still can't persistently reference a local dir, broken ever since they moved to their cloud project system, breaking many non-technical people's workflows.
today at 8:48 PM
Since I didn’t love the technical approach here (once one takes the humans out of the loop, there’s more ambitious things that can be done), I do appreciate the process. It’s not until the last sentence when process inspiration is revealed: “Special thanks to Boris Cherny for encouraging us to be more ambitious.”
today at 9:19 PM
I dare you to try and get heavy CPU cache-level performance optimizations you might see in tried and true HFT code written this way.
today at 10:27 PM
I'd be surprised if it wasn't possible
today at 9:17 PM
Sorry! I don't see that much difference. I would be more impress if you can make TTFT and ITL faster.
today at 8:41 PM
Love it, the kids rediscover plain HTML and optimisation.
today at 8:40 PM
The most important question is how much more unreadable the code became after all this "ratcheting the benchmark down". If you unroll a loop it will perform faster, but making changes to such unrolled code will be a mess. Will this make them ship slower overall? I'm sure at least half of it was just poorly written React code, but the other half?
It's the same problem as overfitting in model training. If you're not measuring something it will get sacrificed.
Or, perhaps the code quality literally doesn't matter anymore and we've reached "code quality escape velocity" where you can code as much slop as you want, the next generation of models will clean it up faster than the slop generates?
today at 9:23 PM
Came here to post this. AI rationalists love thinking about paperclip maximizers, but don't seem to care when it turns their own codebase into paperclips. Or to rephrase, turns their whole engineering org into meat proxies, slowing down engineering productivity in the long term because understanding is drained out of the staff and flushed down the drain every time they close a Claude Code tab.
today at 9:25 PM
today at 7:55 PM
Great writeup really appreciate the detail on what actually worked vs. what didn't.
today at 8:06 PM
Step 1 - make a website that takes 3 seconds to load a blank page.
Step 2 - bring it down to 1 second and pat yourself on the back.
today at 8:19 PM
Correct. A web page is fast by default. Everything that is added makes it slower.
today at 9:19 PM
Sure, but some things increase the delay much than others. More static text, for example, is pretty fast. Elaborate JS, not so much.
today at 9:06 PM
In Brazil they say you can make things better by first putting a goat in the living room. It's going to shit everywhere, chew, droll, and make a horrible mess.
Take the goat out and your life will improve right away!
today at 9:29 PM
Also described as a traditional Yiddish folktale (a rabbi suggests responding to stress by moving a series of animals into the home, and then at the end they're removed and everything seems great by comparison!).
today at 9:25 PM
60+ yo Brazilian here and I have never, in my entire life, heard this saying before. Sounds something a Brazuca would come up with after the 17th round of Kaiser though. Pretty on brand.
today at 9:24 PM
A life without a goat sounds horrible
today at 9:02 PM
Faster is great, hopefully the quality remains.
today at 9:25 PM
It can certainly measure your token spend rate.
today at 9:27 PM
”With that approach, we merged more than three thousand changes…” Why are they writing this? That’s terrible marketing all around, it means they let it go so far, with so little care, that they needed 3000 changes to make it into just a mediocre website performance wise (sure, electron app too, but still).
today at 8:04 PM
AI written slop. They need to upgrade to Opus 5.5 or switch to OpenAI for writing.
today at 10:08 PM
My user experience was significantly worsened just by their referring to "user journeys". I mean, I definitely don't want to go on a journey with Claude or ChatGPT (admittedly my go-to). I want result quickly and be done.
today at 8:38 PM
With a simple trick: make it very slow first
today at 8:01 PM
Let's try this again if you want an instant 10x speed up:
Claude rewrite Claude Code from TypeScript into Rust. Make absolutely no mistakes.
today at 8:38 PM
Not sure why you're getting downvoted, the Rust based tui's absolutely smoke Claude Code.
today at 8:49 PM
It's an overdone Reddit-esque comment that adds nothing to the discussion.
today at 7:47 PM
[dead]
today at 8:19 PM
[dead]
today at 9:10 PM
[dead]
today at 9:09 PM
[dead]
today at 7:56 PM
tl;dr if you make absolute dogshit software that takes 4.38 seconds to stabilize its first paint you can make really nice headline claims by "optimizing" it later
today at 8:02 PM
Unironically true, can someone question, what is Claude desktop app doing that needs 500k+ lines of code?
Agents complicate something that should be much smaller and simpler and then agents speed it up adding more complexity. I suppose functionally you may say this is fine but aesthetically it is hideous!
today at 8:36 PM
the codex desktop app bundles libreoffice's libraries. i wouldn't be surprised if claude does something similar.
today at 8:27 PM
TempleOS had kernel, compiler, 2D and 3D graphics drivers + libs, and tons of games and applications and the entire thing weighed in at less than 100kloc.
AI coding agents of today definitely bloat things up [egregiously].
today at 8:35 PM
On the other hand applications on TempleOS are expected/assumed to be perfect and have no bugs. It makes things a lot simpler =)
today at 8:10 PM
There is nothing even functionally fine about this. The me who has programmed for a 4mhz computer with 128kb of RAM is crying inside. It's a fucking trivial interface for writing and displaying text and sending HTTP requests. You could have that displayed ~instantly on an 80s home PC. Now we have home computers that can execute somewhere between billions and trillions of instructions per second and yet a task that should take <10ms takes 4500ms. Our industry has become an absolute embarrassment.
today at 9:03 PM
Thank you. The apps created like this are absolutely an embarrassment. But to me the bigger horror is if people are going to be convinced to develop the critical libraries and infrastructure in the same way. Then we'll have bloat and rot in the deeper levels and everything will become exponentially worse and unreliable. It makes me so sad that people can't see this.
today at 8:11 PM
The Claude CLI does a lot more....
today at 8:14 PM
This article is about the graphical interface, not the CLI.
today at 9:36 PM
The Claude graphical interface does a lot more...
today at 8:10 PM
Most line-of-business software has a lot of optimization opportunities. Making software optimized takes up time that can be spent building features. The fact that you can just make things go fast without having to take time away from feature building is actually pretty awesome.
today at 8:12 PM
Not taking 5 seconds to load text is a feature. 10x more valuable than whatever other shitty feature you're thinking of piling onto your monstrosity. Like, actually tangible valuable to users and consequently your business; Google has already done the studies at scale that demonstrate how every 100ms of delay has an observable impact on usage statistics and user retention.
today at 7:43 PM
[flagged]
today at 8:32 PM
[flagged]
today at 8:36 PM
Bot comment?
today at 7:47 PM
today at 7:47 PM
This was a fantastic read. Lots of useful info in there for your own projects.