Okay, so anthropic has amazing AI which supposedly writes most of their code and can continuously improve... meanwhile they have outages on a regular basis, and any kind of long-running work will now consistently hit 'API Error: Server is temporarily limiting requests'. Not sure of this is intentional to force a reduction of token usage, but at this point I need to build around these throttling limits and outages with my own tools to restart/resume sessions. From my experience, in the last 2 weeks, literally 100% of any non-trivial Claude session/work will now be blocked on these issues, requiring manual intervention.
One of my focuses now is my own model-agnostic, harness and workflow orchestration (I know everyone is building these) , baselining on opus, and aiming to transition to Chinese models like deepseek in the short term and hopefully open, self hosted models in the future (which I plan to open source).
The nonstop marketing fluff from anthropic while their service quality and availability noticeably degrades... just continues to destroy my trust in the company.
You mean on some days it goes faster and some other days slower?
That is by design. It depends on how much other people are using their services right now and they do communicate it somewhere in the TOS that they do this. Otherwise they could give us a fixed amount of tokens - but they don't because it is not fixed.
Their outages are probably not due to their code though. It’s probably their infrastructure that can’t keep up. So seeing failures of infrastructure doesn’t really tell you anything about how good or bad Anthropic makes use of their models.
This is a beta feature where Claude code draws the interface on the terminal’s alternate screen buffer like vim or htop. I believe it’s not the default because there are some potential compatibility issues deepening on your terminal setup. I’ve found it to be a nice improvement. It also fixed the issue where copy-pasting selected text from the terminal creates unwanted line breaks.
Claude Code is essentially a terminal emulator that runs on mature OSes with excellent support for this type of application. Why are they having difficulty implementing it?
The whole thing is actually powered by a shitton of hamsters inside a bunch of 4u rack mount cases running on spinning wheels at high speed. Somehow at scale this works.
Sometimes they all happen to randomly take a nap at the same time - hence the outages
That seems like an assumption based on basically nothing. There is a lot of code at the infra layer, and based on the stack choices for Claude code and based on how buggy and unreliable ~everything from anthropic is, it seems pretty bizarre to claim these issues are not related to their code.
There are other indications, however, like Anthropic paying through the nose for compute just months after Dario told Dwarkesh how hard it is to predict demand, or ChatGPT and Codex not quite having the same issues after Altman spent much-publicized years scrounging for trillion-dollars of capacity.
While I'm very bullish on Anthropic, I'm a bit wary about their IPO because it seems to me that they're filing now while their financials look good and before other trends like the decline of tokenmaxxing and their compute bills catch up.
Whoa, first name basis with Dario but not Sam. Ouch. [I actually have no idea who Dwarkesh is and it sounds like a first name to me but that's not a particularly reliable indicator so I won't comment on your relationship with Dwarkesh.]
Oh, are they filing now? I think their financials look somewhere in between devastating and criminal, so I'm really looking forward to the IPO!
Well, people keep throwing money at them, including you and investors. So why would they care? It hasn't annoyed you or a large enough portion of users enough to move off their service - because there isn't a better alternative.
This is a decades-old design pattern when CPU >> IO. Emacs has been doing just that since the 80s, when people were complaining about "Eight Megs And Constantly Swapping". See "redisplay" [1]
This minimizes screen flash. You can't rely on terminals doing double-buffering.
> This minimizes screen flash. You can't rely on terminals doing double-buffering.
GUI and TUI have different architecture model. Most GUI have have a 2D surface that is redrawn multiple times per second. Double buffering is for decoupling update and render. TUI is a grid of characters that are updated one at a time via an active element, the cursor. Double buffering there is very wrong. Like adding airbags to a bicycle.
There’s a reason you see most old TUI either have an option to redraw the screen (automatically like top, or manually) and those that have a scrolling option allow to scroll by page. The TTY (the underlying concepts) used to be slow and it can be slow today as well (ssh connection). You need to be thoughtful about whole screen updates.
React part maybe. The rest is what any TUI that's using ncurses would do. :)
It really bothers me that most of the TUI harnesses are using 100% CPU quite a lot just printing stuff to terminal. Seems ridiculous.
I guess it comes from syntax highlighting/formatting, which is probably not done incrementally, but over the entire so far displayed block of output, recomputed from the beginning for each new streamed in character. Can't imagine anything else causing the rendering to gradually grind to halt when eg. thinking block is open in opnecode and updates get palpably slow as it grows.
Terminal output itself is fast and consumes almost nothing. You can have 60fps terminal apps that update content every frame and that consume almost no CPU time.
> Terminal output itself is fast and consumes almost nothing. You can have 60fps terminal apps that update content every frame and that consume almost no CPU time.
The TUI mode is a client-server architecture. An analogy would be like an html page where all content is updated server side. Try to do 60 fps and you’ll have flickering as well.
No. Fetching pages from remote server will just make the client wait for I/O. That takes 0 CPU load and if the server can't respond at 60fps, lowered redrawing frequency would mean even less CPU load from the terminal redrawing itself.
This does not explain 100% CPU load these harnesses sometimes exhibit.
It's not recognizing that they are just one building block that should do one thing well, like tmux.
You don't need a computer display on your fridge for the same reason, but Anthropic think you do. You should see virtual ice getting created and they should correspond to the actual ice behind the door - think of how amazing that is!
And it's not even completely a bad idea. make it claude-code-react-beauty of some way to take it off, it would be far more palatable.
I love the idea of installing high resolution cameras in the fridge to monitor the ice maker to feed into a vision model that renders digital ice to the exact position of the real ice on the fridge’s giant screen
Put a servo on the door and a camera on the front. Train a vision model to recognize when your eyes are looking at the door and automatically open it for you.
Another camera inside will detect when you are done and close it.
> We have a ~16ms frame budget so we have roughly ~5ms to go from the React scene graph to ANSI written.
It looks like video frame, full framebuffer, generated and parsed at 60fps. It surprises me they haven't introduced GPU shaders, 16x oversampling and raytracing. Maybe for next release.
Also you forgot "render to a framebuffer, then parse the framebuffer back to chars".
Anyway, I'm off to construct the new `ls` command. It will render the list of files to a mesh of billions of polygons in a GPU with advanced shaders, 16x oversampling, HDR and all the graphic acronyms I don't understand, then read the resulting image, find the nearest character in the ANSI charset and use that one.
Could be improved. Encode the image to webp with high compression settings and handle the ASCII mapping by spinning up a local LLM to do OCR on it. Individually. For each cell.
My roguelike's "graphics" are a simulated terminal, so it's a 2D grid of colored characters. It's essentially a TUI, just like Claude Code, except instead of rendering to a real terminal using ANSI escapes, I render to a web canvas using... something probably more complex than what Claude has to do. It's still not hard.
They forgot to add 'make it as simple as possible' in the prompt is one possible cause.
On a more serious note using a react-like lib for TUI in the hope you'll share the codebase with the web version is a more likely explanation. Still not the best idea.
React is not that stupid to re-render in a loop at 60fps and instead waits for changes to happen before re-rendering. It even batches changes and stuff.
You don't need React for reactive TUIs - at all. I can understand chosing React for web, but for a TUI it sounds like a really poor idea. And in practice we can see that the claude code TUI is also poor.
It doesn’t need to be that complex, but it can be that complex without being slow. Claude Code’s interface is extremely simple. It has tons and tons of headroom to tack on performance overhead without it being noticeable at all. You just have to not do dumb things like redraw the entire UI every time a spinner spins.
"We made our app chew up so many unnecessary resources that we can use even more resources in the future, and no one will notice" is not the strongest engineering idea I've ever heard.
I can't help but think it's their engineer's and PM's making these decisions, since I know that if you asked Claude to write a TUI there is no world it would recommend whatever the frontend architecture of claude code is.
~ "it's not a TUI! <describes an outrageously overengineered TUI> and my dad works at Nintendo"
curses, bud. curses.
It's genuinely difficult to tell how much of this is true. The post is obviously 100% posturing, but some of the words describe things that could be done.
Very few game engines do anything I'd describe as rasterisation. That's kind of the point of a GPU. Well, it used to be.
I suppose "small game engines" might be more likely on average to include a rasteriser. The typical reason for this is because the author wanted to write it.
Whereas big engine make triangle give hardware go brrr.
So I assume here 'rasterize' means 'printf'.
And diffing screens means diffing 50..150 lines of text.
And "generating ANSI sequences to draw" means 'printf' with some ANSI sequences interpolated in.
Then there's the frame budget. You have to understand they are operating within a strict frame budget -- they're not messing around, OK. They have a 16 ms frame budget, so they burned 11 ms and now have a (roughly) ~5 ms approx. budget for the final 'printf' in the chain???
Your broader point is well taken but I thought I'd stop by with some trivia. High end engines such as unreal will rasterize absurd quantities of micro-geometry manually using compute shaders in order to avoid the bottleneck of the hardware rasterizer.
High end engines such as unreal have the excuse of being tasked with rendering millions of polygons, in which case a complex approach makes sense. Claude Code is only being asked to render a few thousand UTF-8 characters.
> For each frame our pipeline constructs a scene graph with React then
-> layouts elements
-> rasterizes them to a 2d screen
-> diffs that against the previous screen
-> finally uses the diff to generate ANSI sequences to draw
So I’m wondering what ‘rasterizing’ literally means in this case. I imagine it’s just creating a 2D map of elements at a very low (probably character) resolution, then diffing that against the last generated map to come up with an optimal ANSI sequence to send to the terminal, would that be right?
Seems like a cool puzzle to solve. I wonder what the engineering and organisation tradeoffs were that lead to it — does it let them reuse a bunch of existing code?
I wrote a TUI library back in the day for Turbo Pascal — it was essentially taking an immediate-mode approach (which in this context is just a fancy way of saying it was procedural haha).
"Rasterizing" means just one thing in this context: to transform a data structure into an array of pixels. It seems absurd to do this, given that the next step must be to convert back from pixels to text data, but maybe they have some way to generate predictable sequences of pixels (e.g. the character "t" is always rendered as the same pattern of pixels), such that they're cheap to convert back.
If they're doing anything else, the word "rasterizing" is being misused.
Yes, the much more plausible explanation is that the word rasterize was misused there. They are generating and diffing text data which has been a standard approach to drawing a TUI since the dawn of computing. It is not even remotely resource intensive.
> They are generating and diffing text data which has been a standard approach to drawing a TUI since the dawn of computing. It is not even remotely resource intensive
No one has ever done that. Even top[0], which does full screen refresh, clear the screen (if necessary) and write the new information (the period is in seconds, not ms). No need to diff. That would be like diffing a file, just to find which bytes to update.
I don't understand why you would make such a confident negative claim rather than ask for an example or otherwise engage in discussion. Particularly given that you replied to a comment elsewhere in this very thread that links to a real world example of exactly such an implementation! [0] See in particular this part of the source. [1]
I agree that most programs don't bother to do that but please recall that my claim was merely that what Claude Code is claimed to be doing with regards to diffing is a well established and long standing optimization. The important point being that it is neither expensive, novel, or particularly complex thus not an excuse for poor performance.
The emacs code is not purely diffing. They already have the final output, they’re mostly comparing it to see a cheaper way to update than render the output. I’m pretty sure the curses library have the same thing.
But ink, the library Claude is using, defines a tree data structure for the main concept. The diff there is about comparing the old tree and the new tree created by the update, and then updating the node that has changed. That means if a single character change inside a bing panel, the whole thing is rewritten. And if you have something that is updating a lot, that means flickering.
The diffing that ink does is just architecturally wrong. You can create a dom, but a dom is not a concept for the terminal. It’s up to you to optimize its rendering. But just diffing the dom structure like react does is not optimizing, it’s busywork.
Hmm I thought this was due to me using tmux with claude-code, also it seems that `claude agents` doesn't have this issue.
By comfortable ergonomics, meant the forgiving and asynchronous input system. You can start typing, cancel, retry with previous input, accumulate messages while the agent is active. I don't know all TUIs but this is not common IMO.
> You can start typing, cancel, retry with previous input, accumulate messages while the agent is active. I don't know all TUIs but this is not common IMO.
Literally every audio player or anything that uses threads.
good point, i didn't classify tui audio players in a way, they don't converse, they allow asynchronous effects and stacking, that said i might be lagging about these, last i used was mocp, any names i should check out ?
Maybe Claude is operating at a higher, self-improving level than all of us poor HN commenters. Wasting the local machine's resources to look pretty is a plausibly deniable way to make the Claude Code FE unusable with local LLMs, starving the competition.
I dont think they need to optimize their infrastructure (at least not from their perspective). They have high-end PCs with 64GB of RAM, so 1GB doesn't matter to them. For example, I have 8GB of RAM, and I make my apps very performant. Honestly, I probably wouldn't bother if I had 16GB+ of RAM
For useful things, by the computer's owner. It's not there to be used just because Anthropic can't be bothered to give a shit about the quality of their product.
> which eats 1GB+ of RAM. Meanwhile, my editor only consumes 80MB of RAM
And why are you comparing Claude Code to your editor?
> They can't even improve Claude Code
That depends on how you define "improve". They've added a ton of features to it over time. Who said minimizing RAM usage was something they are prioritizing right now?
> why are you comparing Claude Code to your editor?
Because the editor does more. All the compute-intensive parts of the agent are in the cloud. Zero reason for an agent harness to require anything beyond a potato to run.
The main reason I am building my own agentic environment is that I need full control and reproducibility of what I am building.
Post November and post openclaw agentic environments need to be built differently, and for selfhosting models the context size problem really requires a strong harness which intelligently helps reduce context size.
Planner/orchestrator architecture, agent to agent summarizer, specification based tools (fck all this markdown memory bullshit btw), tool call shrinking, and workflow management are all really important because of the context size problem.
Nobody has enough VRAM for the large K/V caches, and nobody can afford f16/f32 caches in terms of memory, which are also necessary for longer conversations. MoE 30b models have improved so much though, qwen 3/3.6 coder is the real champion doing almost the same things with less than 1/10th the memory requirements. Just think about that in terms of engineering and what your bet is going to be. Haiku pales in comparison.
Currently my focus with exocomp is trying to figure out how I can record, replay, restart, and debug workflow sessions of agents in a better manner so that I as a human can understand what's going on. Currently I think that UI will be something like a gantt chart where you have a graph with connections representing agent to agent communication. And yes, that's a lot of fiddling with SVG as it turns out, so I'm not quite there yet.
Anyways, in case you're interested. I'm manually building this env and trying to unit test the critical parts. [1]
If their AI is good enough to write their code, why isn't it good enough to tell them how to fix their infra? That's a different problem space, but it's not harder than the code.
Growing pains of being successful. These are solvable problems and will be. Can they maintain their momentum without pissing off too much of their customer base before these issues are resolved?
Personally at my own job self-writing code is letting us tackle big, long-deferred refactoring projects (like the article mentions), but any sort of refactoring introduces new bugs.
Look, I've never been someone who mindlessly hypes AI companies, as a matter of fact I think they have serious leadership problems across the board, but you people are straw-manning them so badly it actually makes me sympathize with them.
They aren't saying they have fully automated luxury AGI, they specifically list the ways models fall short of that bar and caution against people taking the 8x figure as the actual uplift number. At the same time they recognize that 80% of new code is now AI-authored, when two years ago those models were little more than toys. And frankly that checks out: if two years ago you told me we'd have something like Opus 4.8/GPT 5.5 I would have rolled to disbelieve.
> At the same time they recognize that 80% of new code is now Al-authored
I can setup a loop that will write a trillion lines of code automatically, how much of it is actually useful? Or are we back to counting LoC because there's no other metric for these systems that anyone can rely on?
It's 80% of new code they shipped that is AI authored.
Would you ship pointless code?
I do tend to agree though, it could be that AI solves problems with more code than a human would. What you need to measure is the value the code brings and how much of that is done by AI, hard to get an objective measure of that though.
I wouldn't, no. I don't see evidence that the engineers at Anthropic are similarly cautious however. They describe Claude Code as "basically a game engine" when it's literally a TUI app, and it eats memory for no apparent reason. I fully believe that Anthropic would ship pointless and garbage code. Especially if it's being written by LLM.
I could write a bash script that copies a codebase repeatedly in the pre-AI past as well, but I didn't do that because I wasn't stupid. More than 80% of my code is now AI-generated, and trust me I'm still not stupid. It was 0% only a year ago.
Who says LoC is the only metric we should rely on? A software product should first and foremost meet user requirements, functionality and performance. Judging from the sensational rise of Anthropic's user base and revenue I think we can safely says they're in that ball pack.
Not if you’re claiming that the spells, once cast, automatically get exponentially spellier until they awaken into a spell god, capable of literally anything, including casting more complicated spells than any wizard is capable of. If that were true, you’d have no need for wizards. The fact that wizards are still around means it’s probably bullshit.
So in your opinion AIsnd LLMs aren’t improving? They can’t do it today, therefore they never will?
Certainly has never been times in the recent past when people have confidently predicted computers could never do something that computers were then able to do shortly after the prediction was made.
What really happens is the spells only have other spells to draw from and they begin to degenerate over time, eventually turning into chaotic eldritch horrors that randomly add limbs to people or adamantly refuse to discuss goblins or just shriek in gibbering madness. Our Evil Overlord sacrifices the dreams of children to keep the magic sustained and controlled, and soon the people can't even think or speak without the help of magic. And they think they're wizards even though they can't even read a grimoire.
This is weird to me because i am using claude code 10+ hours/day 7 days a week, usually multiple sessions, and run into api errors maybe in 1 or 2 sessions per week. And about..2 major outages of 10-20min in the last month. Not terrible and nowhere near what you are reporting. Therefore I dont believe you, because you dont even couch this in terms of it being something that seems particular to you or your region. Obvious dishonestly is fairly bad of you.
We’ve got a company of several thousand employees serving hundreds of millions of people arguably the best AI model in the market. Meanwhile you’re asking for a handkerchief for your pool of tears because their product is struggling to do your daily job functions for you, with much of that due to being limited by the worlds supply of silicon, electricity, water, and other resources. Cry me a river.
Irony, maybe? Do you not get it?
If these models are so great solid_fuel then I guess it wouldn't be interesting that Anthropic's own models can make up ulterior BS as analysis.
So why don't you pound sand since that clearly went straight over your head? That would be far more useful than your asinine response.
One of my focuses now is my own model-agnostic, harness and workflow orchestration (I know everyone is building these) , baselining on opus, and aiming to transition to Chinese models like deepseek in the short term and hopefully open, self hosted models in the future (which I plan to open source).
The nonstop marketing fluff from anthropic while their service quality and availability noticeably degrades... just continues to destroy my trust in the company.