"it doesn't silently go unnoticed", "would be silently inert", "instead of silently overwriting", "you can never silently overwrite"
The biggest tell for me is overuse of the term "silently". "quietly" is another one you often see from Claude in particular. Models love adverbs for whatever reason, whereas a human writer would use them in moderation for emphasis or prefer terms like "by accident".
In my experience, "silently overwrite" appeared regularly in technical writing long before LLMs were a thing, because it's a useful concept to be able to point at. "Silently go unnoticed" is kind of redundant, though.
Accidental things and silent things are very different. Accidental means you didn't mean to do it, silent means you don't know you've done it (or might not if you want to get picky, you could notice).
I often see replies to AI-generated posts being pointed out here asking what makes it obvious. Is it that difficult to notice the indicators? Is it mostly undetected by English-as-a-second-language speakers, people inexperienced with generative AI, or is it something else?
Have you ever “wired” anything to anything else when developing software? No, because software doesn’t involve wires, but LLMs are quite convinced that it does.
Depends on your background. If you're an EE dabbling in writing software, using the verb wire for connecting a couple of libraries together rather than "pipe" or "glue" seems entirely reasonable to me.
I can't quite tell you, it wasn't something specific, Claude's writing is just a specific sort of punchy. The "directly on X - no Y, no Z, no A" and the "this is the part you no longer have to do" just smell a lot like Claude. Also "removes that layer entirely", "they map cleanly onto each other", it's all just Claude.
It's how you see a painting and you know it's by Picasso, let's say, or you read an author and you know it's Hemingway. Everyone has their own unique style, and so does Claude. It's just that Claude is the most prolific writer in human history now.
Yes - the Claude ACP is nice, as I like to have a view of the code while chatting. Using just the terminal for dense/long running work feels like a handicap imo. It would be great if it supported more commands though!
I feel like it doesn't support some of the commands that manage Claude itself so think `/mcp` `/plugins` etc. Most of the common ones are configured to work though from what I've seen but the ones that do more configuration of Claude seem to be blocked.
That is likely a drawback to their ACP wrapper scheme, it helps exposes IDE functionality but they have to keep up with Claude Code functionality in the other direction. VSCode's Claude code plugin is just like using the CLI.
Entirely right it's a limitation to the ACP side. They're in the middle of adding functionality where you can have terminal/CLI threads and ACP threads too. https://github.com/zed-industries/zed/pull/54729
Context usage is in an open PR now! https://github.com/zed-industries/zed/pull/54881 give it a week or two depending on if you want to use stable vs preview releases. I haven't tried pasting images yet either but I have used their context menu that lets you add images.
It works 'well' with Claude Code, but you're going to be missing a lot of features. There's no display for sub-agents/teams, no ability to clear the context without starting a whole new thread window, no ability to view the current context or usage, etc. There's also no built-in ability to view or change the model's current effort level, which I think is a current limitation with the SDK.
I tried it for a bit and it was definitely usable and I got a few features built out, but I eventually moved back to using CC in the terminal. I'm sure they're working on it, though.
Does "local Ollama" or OpenRouter count? I fell into using Zed because there was zero sign-up friction when trying to set up a connection to a local Ollama LLM. Literally "drop-down, select the model you want"
Once I got that running on my machine it was also easy (literally a drop-down+ API key) to switch and explore using models on OpenRouter.
I used to run Claude code on terminal on zed. But the memory usage would balloon eating all my ram 128gb and have to kill the session every other day. I moved back to vscode. I don’t know if they addressed it
They both work very well. My only complaint is that you can't use them to generate commit messages using the Git Panel (have to do it inside the Agent Thread instead), and /compact seems to have issues so you will often need to start a new thread (this wasn't an issue in the past so hopefully it will get fixed soon).
It lacks a lot of features, but IMO feels less "busy" than the terminal version, which I like.
Very recently Zed also gained support for parallel sessions, which is nice.
In general it's very obvious that a lot of effort goes into improving it, and it gets better with every release.
I haven’t heard this quote before, but I am copying it here because it makes so much sense:
Arguing that you don't care about the right to privacy because you have
nothing to hide is no different from saying you don't care about free
speech because you have nothing to say. - Edward Snowden
I actually set the time limit to 0 minutes , then I restarted the app in iOS and right now I do not see any shorts. Let’s see how long it would take for them to reappear again
it goes much further back than that, think it was xp
the issue is that it sucks, it's at least 10x slower than 7 zip, maybe more, showing lots of files/folders freezes the explorer gui on w10 and it only supports .zip (which could've been changed on w11, never used, never tried)
there was a German startup (Sono Sion) some years ago that had a very similar idea and had even prototypes [0], but they filed for insolvency 2-3 years ago. Now I think they pivoted to something more boring and practical - putting solar panels on vehicles. I hope aptera gets to production though
There’s also Lightyear. But so far the Zero (formerly One) ran into the ground after a barely existent production run, and the Two I’m not sure exists in any way.
It is interesting that 2 years later the same reviewer changes their mind a bit:
> Now, a few years after writing that review, I must return to say that as much as I hate this book, it's probably the best textbook that I have. I constantly return to it to reteach myself basic concepts or math. The problem with the text is that in order for it to be useful, you pretty much have to already understand the material. It's a dense, technical manual that, when paired with an easier to understand text such as Griffiths, grants tremendous power. Don't get me wrong, if there is a hell, I personally hope John David Jackson is burning in it right now, but I also have to tip my hat to him
