Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

I think the charitable interpretation is that the author is referring to particular use-cases which stopped being served by CLIs.

Heck, just look at what's happening at this very moment: I'm using a textarea with a submit button. Even as a developer/power-user, I have zero interest in replacing that with:

    echo "I think the..." | post_to_hn --reply-to=45742461


He mentions this as bad example of UX: "search web history for a doc about atlas core design"

I have opposite view. I think text (and speech) is actually pretty good interface, as long as the machine is intelligent enough (and modern LLMs are).


A hybrid will likely emerge. I work on a chat application and it's pretty normal that LLM can print custom ui as part of the chat. Things like sliders, dials, selects, calendars are just better as a GUI in certain situations.

I've once saw a demo of an AI photo editing app that displays sliders next to light sources on a photo and you are able to dim/brighten the individual light sources intensity this way. This feels to me like a next level of the user interface.


IMO the best fusion for this kind of thing is:

1. There's a "normal" interface or query-language for searching.

2. The LLM suggests a query, based on what you said you wanted in English, possibly in conjunction with results of a prior submit.

3. The true query is not hidden from the user, but is made available so that humans can notice errors, fix deficiencies, and naturally--if they use it enough--learn how it works so that the LLM is no longer required.


Yessss! This is what I want. If there is a natural set of filters that can be applied, let me speak it in natural language, then the LLM can translate that as good as possible and then I can review it. E.g. searching photos between X and Y date, containing human Z, at location W. These are all filters that can be presented as separate UI elements so I can confirm the LLM interpreted correctly and I can adjust the dates or what have you without having to repeat the whole sentence again.


Also, any additional LLM magic would be a separate layer with its own context, safely abstracted beneath the filter/search language. Not a post-processing step by some kind of LLM-shell.

For example, "Find me all pictures since Tuesday with pets" might become:

    type:picture after:2025-10-08 fuzzy-content:"with pets"
Then the implementation of "fuzzy-content" would generate a text-description of the photo and some other LLM-thingy does the hidden document-building like:

   Description: "black dog catching a frisbee"
   Does that "with pets"? 
   Answer Yes or No.
   Yes.


>charitable interpretation

Theres very little thats charitable in the article, so I met it at the same place.


But the article is coming from a decisively antagonistic angle though.

>This was also why people hated operating systems like MS-DOS, and why even all the Linux users reading this right now are doing so in a graphical user interface.

What is the significance of "Even all the Linux users"? First of all it's probably incorrect, because of the all quantifier. I've went out of my way to look at the website via the terminal to disprove the statement. It's clearly factually incorrect now.

Second, what does hate have anything to do with this? Graphical user interfaces serve different needs than text interfaces. You can like graphical user interfaces and specifically use Linux precisely because you like KDE or Gnome. You can make a terrible graphical user interface for something that ought to be a command line interface and vice versa. Vivado isn't exactly known for being easy to automate.

Third, why preemptively attack people as nerds?

>There were a tiny handful of incredible nerds who thought this was fun, mostly because 3D graphics and the physical touch of another human being hadn't been invented yet.

I mean, not only does this come off as an incredible strawman. After all, who wouldn't be excited by computers in an era where they were the hot new thing? Computers were exciting, not because they had text interfaces. They were fun, because they were computers. It's like learning how to drive on the highway for the first time.

The worst part by far is the unnecessary insinuation though. It's the standard anti-intellectual anti-technology stereotype. It creates hostility for absolutely no reason.


The first time my driving instructor took me on a high way.. after a few minutes he looked over at me and said something like "isn't this exciting?" The sarcasm did not register with me because I indeed found it exhilarating even though we were just casually cruising in traffic.


A TUI client (with some embedded CLI) would perfectly work for HN.


That's still not a command-line interface. An ed-based mailer is a command-line interface: what you're describing sounds more like *shudder* vi.


Some TUI programs can embed a small cli. Like Midnight Commander, and others. Or externally call commands and shells, or even pipe the output. Ed itself, vi, slrn...

Also, some commenters here at HN stating that the CLI/TUI it's just the fallback option... that's ridiculous. Nvi/vim, entr, make... can autocompile (and autocomplete too with some tools) a project upon writting any file in a directory thanks to the entr tool.


That just doesn't register with some people.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: