People keep asking for this w/ Codex, too, and I really regret that both labs seem inclined to listen.
If you ask a thinking/research type question in 'Chat' versus 'Work' mode in these products -- say, something complex about politics, or do a multiturn business strategy, or want to work thru a new concept, you get very different answers.
The harness, steering, etc. in the chat/reasoning products is so much better for this type of question (that doesn't require code as a primary substrate).
Even as someone who mainlines like, 7 coding agents at all times, I regret that productivity fever will mean the regression of think-first-act-later AI UX.
I hope not Chat is still all you can eat and Work uses Codex tokens. I can see why they would want people to think they wanted this without thinking about how it currently works. If Chat starts eating tokens then there is not much reason to use Chat other than most people don't live in the command line like developers. Claude Chat / Cowork already eats tokens either way so not much of a change really. I have never used Work since if I want access to local files the CLI is a much better interface combined with an IDE, but can see the appeal to non-developers.
This is my main reason for not using Work. I was tempted to use a few "scheduled tasks" in ChatGPT, but once I found out they used Codex tokens I stopped creating new ones.
I built a terminal control plane and added an MCP server that runs on my Mac. It lets MCP clients create terminal sessions, send messages, read the contents, etc on my Mac.
The terminal sessions are running Luna on XHigh (on Pi) and GPT 6 Pro on ChatGPT.com is the main orchestrator.
Since ChatGPT and Codex have different limits, I can have 6 Pro orchestrate the Luna sessions. Using voice mode to talk back and forth about the software factory is really nice. Will be open-sourcing soon.
I recommend using Work sparingly because sometimes it is the better choice, but I do ration it. With Claude I simply don't use the web/app chat at all (and I have no idea what Cowork even does ngl, never used it) I think there's a Claude Design? No idea why I would need it (mostly, why I would spend tokens on it lol)
Claude Design is good at making a bunch of UI pages in a single design look consistent, and it's better at making an initial starting UI than just asking Claude Code. Think of it like Anthropic's version of Figma. It honestly works pretty good and I use it often and feel that it saved me a lot of time. It's pretty token heavy though.
GPT Chat seems smarter for business questions or planning that does not require iteration or wrangling data sources.
But Work gives the model a computer instance it can use for programmatic data access, API munging, etc.
Success cases I’ve had:
chat - read the reports in my Gmail and put a counter proposal together.
Work - cross reference all the options with commute time using traffic aware map data. Ground it on drop off / pick up times from each option’s stated rules and hours. Create a table with commute and idle time break downs.
The fact that it doesn't work well today doesn't invalidate the user need. Not everyone wants a terminal style interface. Traditional UX makes more sense especially if say you are accessing existing plugins and say kicking off a scheduling task - no way would I want to do that via a chat interface if I have the option.
I see Work as the mode where you can generate and edit files without manual intervention.
I tell chat to recommend me some books based on some of my favourites. Chat does it. I read some of them, liked some, disliked few. I give chat that information, it recommends me based on those, now I don't know which of the entries are relevant or not.
I don't want lock-in, so I try my best to tell it to generate copy-able MD syntax that I have to manually paste to an editor.
With Work mode, I tell the model to just maintain an append only markdown file, where it writes "still recommended", "not recommended any more", etc. beside titles. It does this after each turn.
This is just an example. Work - I see it as a mode that can maintain, update, and edit a file.
I agree that they should be separate modes, but presenting it as a toggle in the app is a terrible user experience.
The model itself should figure out how much “power” it needs to run the query. I’m just asking a question. Whether you need to do a simple internet search or spin up a VM and run scripts is up to you.
> The harness, steering, etc. in the chat/reasoning products is so much better for this type of question (that doesn't require code as a primary substrate).
I had this same thought recently. Is it possible to 'recreate' ChatGPT/Claude Chat easily in Codex or Claude Code? I assume it would roughly be just swapping the system prompt. Anthropic seems to publish theirs but I don't think OpenAI does
LLMs are inherently deterministic. The way everyone deploys LLMs leads to non-deterministic results, but there's nothing† stopping providers from offering deterministic evaluation if they choose.
All the sources of randomness are under the control of the provider, even if today's deployment structures mean providers introduce extra randomness due to the concurrent nature of the evaluation. Serialise the computation, feed it from a pRNG, and you have a fully deterministic result. But providers don't want to offer a deterministic result, and especially not one as fragile, expensive, and inefficient as a full serialisation would be.
†: For variants of "nothing" that include cost and deployment challenges.
This is technically true, but when people talk about randomness, its not only about same input-> different output, like temperature>0 and the things you said.
Its also about very similar inputs -> different outputs. Even with everything you said, yes, same input would result consistently into same output, but sliightly different input and you might get completely different/semantic answer.
There's deterministic enough, and then there's computer science non-deterministic. If I ask for a Todo app, I'm going to get a Todo app, even if the buttons get moved around and the background color of it is brown instead of purple if I ask today vs 6 months ago. If the AI completes the phrase "the capital of France is..." with anything other than Paris, something has gone more wrong than usual.
In one of the earlier iterations of this they at least made the UX distinguish Codex vs. ChatGPT which I really enjoyed.
I must admit I spent a little time in ChatGPT this weekend, thinking I was in Codex. After some unexpected behavior (disappointed) I realized my mistake.
The fact that people asked for it means that there is a demand for it. Maybe you aren't the target audience? One size can't fit all. People just have to adjust and go.
Yes the point is that they wouldn't have to adjust unless they are merged into one. Precisely because one size can't fit all, yet they insit on one size now.
It would appear no one else submitted the posts in question, so the most obvious answer is the account is finding interesting articles that no one else has found. It would be one thing if it won out of many dupe posters, but that doesn't seem to be what's happening.
You aren't serious about high-frequency karma ("HFK") unless you colocate your equipment with the routing for your sources so you can beat your HFK competition to a submission by at least 200ms.
We will have reached HFK-ception when the bots are fighting to be the first to submit "How I used AI to build a High-Frequency Karma Farming Bot on Hacker News."
Noob comes from newb, which comes from newbie. Some argue it comes from "nubile" but I don't personally buy that etymology. I distinctly remember "newb" followed by "noob" but that was eons ago.
They are all valid front page submissions by the looks of them and they appear to be the first ones submitted. It wouldn’t be hard to do if you cared enough to configure notifications for frontier announcements. If someone were using technology to influence submissions or misuse HN, that would be a concern but I don’t see any evidence of that.
To directly answer your question, my guess is they have an alert set up for various places on the internet, and are able to post the links faster than anyone else. Maybe that part is automated.
AI product updates probably tend to rank very well here. If you look at the domains the user is submitting, they’re pretty normal: twitter accounts of tech CEOs, big company blogs, etc.
If I saw a bunch of blogs with affiliate links or landing pages for slop products I’d be more suspicious.
I question all the time what dictates the posts I see on the front page every day, and whether or not it's truly based on organic interest. i wish i could say i up/down vote frequently but i do not and i wonder if that's true for most people.
You can make your post on top in some services by paying 5 dollars on fiver. I asked before about HN, and someone said it is hard to do... Maybe they found a way?
These kinds of updates always have this romantic scenario of someone having Claude develop a presentation or something on their way to work between multiple devices, which actually feels a little sad and does not align with what happens in my life at all.
I get to work, go straight to the meeting room, and pull up what it made to present.
It's barely coherent nonsense. Lots of irrelevant details, buzz words, wrong charts or confusing phrasing. Obviously LLM output.
I read it out.
When I'm done, I get a question about one of Claude's incorrectly inferred details.
The shame instantly kills me.
This is a scenario I've seen play out with coworkers. Except that last part, instead of dying or owning up to the mistake of trusting LLM output they waffle. Their shame circuit is broken.
Someone I know does this, and seems to think it's a flex, because he bragged about it being "the first time I've actually seen this presentation" a few times during the presentation.
At my previous company, a distinguished engineer did this. Used generated slides he hadn't looked at until presenting to the entire staff+ engineer group. The slides weren't good and he mostly ignored them.
A Canadian politician did this during a speech in front of the house/parliament or whatever. In the middle of the speech the LLM wrote "Here's a more natural, flowing version of that section that reads like a legislative speech" and he read it verbatim and then continued on, not even realizing what he just read.
Previously having a deck was basically a signal that said “I took some time to prepare for this meeting and put together some coherent talking points” even if the deck itself was just a bunch of of text outlining what the speaker was saying.
Now that people are generating decks with AI, they’re basically worthless. I don’t need a bunch of bullet points you didn’t take anytime to think about on the screen while you talk about something you didn’t prepare for.
Why even have a meeting? That's just a waste of time when agents are much better at reviewing plans anyway. You don't even need presentations anymore. Executive officers can just create a review agent with a "special sauce" prompt (essential to justify their high salary) and auto sign off on a plan if it passes the crucial test: "is this plan good?" Finally we can have true business automation and people can go back to much more important things, like looking at KPI reports and being impressed or unimpressed.
That is incredibly frustrating for a coworker to do that. Even if you do use an LLM to generate all the slides it is your job as a presenter to go through and make edits and actually practice the presentation.
If you are just presenting it for the first time after seeing it you are going to just be reading off the slides, which is a waste of everyones time.
Defaults matter a lot. Before you had to intentionally be lazy to put together garbage and call it a day. Now your boss is telling you to use AI to "speed up." The tool you're pointed to presents like it can make the slide for you and claims to do so. Every signal is screaming that you're supposed to be a lazy fraud, and people will, on average, take the path of least resistance.
I'm attending a conference that I also attended last year, and it's night and day.
This year, a good 40% of workshop slides are fully LLM generated, incomprehensible, and almost not matching what the speaker is talking about at all. Last year, while there were some poor presentations, standards were much higher.
Because nobody actually knows what anybody else’s job consists of. But everyone is familiar with how jobs work in the ‘romcom job universe’ where workplace stakes are generally driven by ‘I have this really important make-or-break presentation due Friday!’
I suspect it’s because writers don’t normally have jobs and the only time they are ever in offices it’s to deliver a pitch for a project.
These are modern infomercials. Concrete pains ("planning a vacation is hard!"), shallow solutions that don't stand up to real-world use, and, in the end, a promise left unfilfilled. (And a thing in your drawer that takes up space.)
honestly, one day i will understand why these guys go so hard on planning vacations.
one of the best parts of vacations is actually planning it! trying to find restaurants, organizing your days, trying to fit activities -- that makes you look forward to your time off.
really, i want to automate the boring parts of my life (did i really pay rent this month?) not the things that make me happy.
Social media has fried some people’s brains so much that it is incredibly anxiety inducing to make many typically mundane or routine choices, for fear of missing / missing out on something.
I think they're grabbing at universally-understood frictions that people associate with pleasure or anxiety … or, most powerfully, identity.
We all know the feeling of wanting to be successful, or to go on a great vacation. Showing these scenarios creates strong emotional response in the audience - we're now the people winning at work, going on a great vacation. This is especially important when you're selling something dull. See: insurance.
"Find a primary care physician in my area who is taking new patients and accepts my insurance. Give me your top three candidates, and book an appointment within the next six weeks with your top choice."
idk I thought the same until I got personal amex support at work, now it’s my roofline on how useful/trusted an agent could be. Now I book all my travel through them, and it’s an absolute blessing
Exactly! Imagine if it booked like a 3000 dollar 1st class flight because it thought it was providing you with "the most comfortable and hospitable experience as possible."
I've found basically all of the release presentations are completely irrelevant and seem to target a very narrow subset of the population. They're always reserving something like a restaurant or a sports venue or generating a new slide deck for their business meeting. Perhaps these companies would do better if they could make a use case for regular people. You know, the 90% of the folks out there working normal jobs and not spending infinite money. I don't think I know a single person who has made a restaurant reservation more than once or twice (mostly never) in their entire life. Maybe I'm the weirdo here, but I can't relate to any of these demonstrations.
I regularly make restaurant reservations and create slide decks for my work, and both of those seem quite common among people I know. What would be a normal people use case for Cowork in your view?
I'm probably about 90%ile income for my household but live in a HCoL area and have two children in daycare. I rent, don't own, and I have enough that I can buy a coffee without thinking, but will be sending my kids to public school, can't just up and go on a vacation.
I make restaurant reservations 7-8 times a year? My wife and I go on dates and you simply aren't getting a table where I live if you don't. Fyi we typically spend $120 or so on those dinners.
Here in Bangalore, you can usually book directly from the Google Maps listing. Or sometimes click a link that will open in Swiggy/Zomato, and you just book from there.
At least in this city, reserving a table using the apps you already have on your phone is faster than using an LLM. Maybe these products are for people who know about the existence of Claude and ChatGPT but not Google Maps?
If I'm booking a restaurant I'll probably be discussing the date/time options with the people I'm going with, and I can't see how adding a chatbot to the mix is meant to help?
I'm retired and write code as a hobby, not presentations or docs, but it fits how I use exe.dev to make a code change. I will have a conversation with the AI, then say "Please go ahead" and close my laptop to do something else. Since I use web browsers exclusively for this, I can check on its progress from my phone, tablet, or laptop.
(They have iOS and Android apps too, but I've never bothered to install them.)
VPS is production. Linux dev box is my local big box. Macbook just lets me move about and do iOS development. Orca lets me run agents either on linux box or locally on the macbook depending on how I'm feeling, and Orca handles all the worktrees & sub-agents for me (via Opencode). Everything runs in a tailscale network, ssh connects all the things.
My github repos all have CONSTITUTION.md files that keep multi-agent dev grounded. Production runs trigger github actions that automatically check logs daily, file bug reports, etc. I launch Orca and literally just type "checking in" and the CONSTITUTION.md file governs the scanning of github issues and prioritization of which issues need to be addressed. Another process prioritizes and bundles them. When I type "checking in" the Orca worktree handles the launching of sub-agents that fix things. Occasionally I weigh in with an opinion or pick a recommendation. If I'm in the mood to pay attention I'll ask it to find another round of bugs and let's keep going. Often I just let it close after the first round. When I tell it "done?" it does a full regression and a production box review. Code auto-deploys to production twice/day. I don't regularly push to production manually.
On Fridays the CONSTITUTION.md kicks off a strategy and roadmap review when I type "checking in". As I have ideas I add them to the roadmap in one of the worktrees. Generally nothing happens until they're reviewed on Fridays together, unless I push one manually.. which happens.
Hermes runs on the production box. I have a few skills and quick commands that lets me check in on production runs and the status of things. It's read-only by design, basically my version of a dashboard. If I have a thought, idea, research link, or question it'll write to a github issue and we'll deal with it in the next check-in.
Orca lets me do all this remotely from an iPhone, and I talk to Hermes via Telegram. I generally don't INITIATE new work using either of these. I report issues/ideas and let the process do the things. But I can, and have.. I just tend not to. The whole thing churns, and so my inclination to "jump in and do a thing" is less than before. My inclination now is to toss things into the machine and let the machine work the schedule. I'm at a point where I could easily automate 80% of this and do my manual things 1x/2x week and I think progress would stay steady. I'll get there at some point, but I like the level of engagement I'm participating in now.
Occasionally I use Claude Code as a watchtower review of things, or do a wholesale code review, or do a review of logs. But deepseek is much better at building this machine -- I built something manually over time and through a ship of theseus process it got ugly. Deepseek reviewed the situation and wrote the CONSTITUTION.md and designed the processes independently. It favors deterministic scripts for process but launches them via LLM and monitors for exceptions; also, no memory system which also disqualifies Claude from being at the center of this. Memory plugins didn't work well because instructions start accumulating all through the chain of instruction files (AGENTS.md, memory, etc) and that gets very undeterministic very quickly. Instructions are written once, reviewed/audited intentionally.. it's best not to let LLMs learn and write their own dynamically (yet).
I've since used it as a template for other projects I've started -- they work the same. Currently building the GTM agent that will handle all SEO, marketing, keywording, etc for my projects -- it'll be a significantly autonomous Hermes agent. Don't really need Claude for any of this and my Deepseek bill is $50-100/mo.
A lot of dev tools now support remote execution, even those used in corporate environments. And for orgs that need locked down dev instances under regulatory control, these tools fit that bill too. I don't think I'm operating outside of norms that are used across all kinds of industries. It may not be the most common config yet, but these tools are building for customer bases that exist.
Hi, this is my team! Happy to answer any questions.
There's a lot in this launch, but the core idea is to simplify the product while giving users access to more capabilities. You no longer need to know ahead of time how much work a conversation might involve. If you're at your computer, Claude can use your local files and apps. If you close your laptop, Claude can keep working on its own computer.
This launch also lets you use Claude Design, Claude Docs, and Claude Slides directly from conversations. That's possible because we made Artifacts much more powerful: whenever Claude makes you an app, website, design system, or anything else, it can deploy an artifact with multiplayer features and databases.
As many of you probably know from your own work, giving users more power while making the experience simpler is really, really hard. It took many iterations to get to this version. We're far from done, but I expect people will be able to do much more while having to think about it less.
From a user experience standpoint this sounds like a clear win. Congratulations.
With AI safety top of mind so much lately, I can't help but notice the announcement does not address this.
With "Chat" mode, there was a user expectation that session had only limited capability to produce unexpected side effects, read sensitive files, etc.
With "Cowork" mode, it seems like more powerful capabilities have been on by default, requiring deep settings and safety understanding to disable if desired.
Merging modes feels like it is removing a simple and easy to understand risk management tool. How does the combined mode help users understand, manage and feel confident about what risks they are accepting?
It's not a bad idea, however if Cowork, Designs, Docs and Slides are still different products, it'd be great for the user to know they are being routed to that particular product automatically because of what they were asking for in their prompt.
There are a lot of people using this outside of the coding space, most of my friends use these to create decks and things like that to present. Mind you they never double check information and numbers.
Claude Code has almost always been able to do a passable job of this. I wonder if this new version will need less cleanup. Claude Code would write python scripts to generate the deck. It at least formatted cleanly with templates.
How are you determining how much oomph Claude gives a particular question? When I ask Claude Chat to e.g. summarize a document, it does this and returns a summary inline, whereas cowork does a bunch of tool calls and intermediate steps, eventually writing a document to disk. How does new claude know how deep to go?
Why is Claude Code still proprietary, while Codex is Open Source? What would it take to fix that? This is a major reason to pick OAI models over Anthropic (because while you can use an open agent with any model, the AI labs charge much more for that interaction style).
I liked the ability to explicitly only chat, without the possibility of Cowork activating since I have never wanted that. It's unclear how I can still guarantee that now.
Will there be a Claude Sheets too? I find myself using Claude to work with spreadsheets more than almost any other document type. It does okay now, but feel like it could be even better, especially with visualizations and formulas.
What sort of issues do you see with spreadsheets? I almost never use Claude or Cowork. Just Claude Code. When I've generated or modified spreadsheets it tends to build python scripts for it which makes regeneration trivial. I have to fight a bit on formatting sometime, but I find if I format it the way I want I can get CC to read the file follow that in the future.
I have a question about project memory. Last time I tried using Claude to organise my chats into Projects, it turned them into isolated silos. I asked a different chat to refer to what we'd discussed in the other chat, and it said it wasn't allowed to access it.
This was pretty frustrating. By trying to organise my chats, I actively made them worse. ChatGPT at least gives you the option to have either open or closed memory. Is this being considered?
I rarely use Claude or Claude Desktop. Almost 100% of my usage is Claude Code because it ties into all of the private infrastructure I've built around it. I've had decent success using CC for generating Office documents of various formats. It tends to reach for a python script and library to generate them and the output is typically clean and easy to apply a template to. I have had issues where I need to manually modify the files to demonstrate format and style, but that's also something that can built into context after with references to templates.
Where do you see Claude doing better than what I'm seeing with Claude Code in these circumstances? Are there any reasons for me to leave my pretty heavily customized CC environment for the other versions?
As part of all the work being done I really hope the sidebar gets more user control. There is a bunch of stuff that I don't intend to use myself but take up first class space (designer is always there, I never need to bring up artifacts, etc). I'd like to reduce the sidebar to what I use most often.
Specifically around the work UX. Trying to bring up scheduled tasks is really painful and required a lot of clicking around instead of just seeing the task results under my project. I hope this is being looked at.
Lastly, and this is a real nit. Let me turn off the "tips" when stuff is being worked on. I'm already paying, you don't need to force a rotating feature advertisement into the interface.
Can you standardize with all of your tools on a common registry key and settings paradigm and same plugin/skills paradigm and common method for enterprise delivery? It’s a pain in the neck getting all of this set up and deployed throughout the org.
In the web interface, are there any plans to support attaching local directories through File System Access API and accessing internal (as in, not reachable over Internet) MCPs, assuming correct CORS setup, of course?
Thanks for participating in the discussion Felix. A big issue for me - not sure if others suffer from this - is that it’s next to impossible to track work in progress, especially with when there are artefacts involved or reports etc. Things that require multi-day work of back and forth and reflection.
Chat is a horrible interface for this, having to scroll up and down across lengthy conversations to try and pick up a thread.
I’m really sorry but I really fail to understand why this massive push to “chatify” everything. A task with its own context and multiple chats feel like a much better approach to work. Even if the work is spawned from a main chat window - eg “Claude we need to work on XYZ” and it creates a task to track this piece of work.
Further, recurring tasks again are really really hard to manage with a chat interface - which of the 10 chats has that question that the recurring task raised?!
And of course all this UX debt will likely stand in your way of building reactive items - ie spawning a task in response to some outside event. Think an inbound email being handled by a prompt that triggers a task and creates a draft response ready for my review and approval. Very hard to do with chat windows.
A bit meta, but this is one of the first posts in Anthropic's official blogs I've read that has distinctively Claude-style prose, thought less prominent than usual:
What Claude makes doesn’t need its own place either.
Enterprise admins choose when to turn them on
the frustrating part was deciding where a task belonged
What they’d started in one also didn’t carry into the other
And the unnecessary school-project level verbosity:
You can choose how Claude checks in with you. By default,
Claude asks before taking an action. If you’d rather let it
keep working and check in only when something needs a closer look,
you can turn that on. You keep the final say.
I am a technical user and I literally last week was trying to work out what the difference between ChatGPT Chat and Work was and couldn't figure it out (other than they bill you differently)
Is there a mature centralized platform, self hosted, that sync all your “AI stuff” across multiple devices? So regardless of the harness, it sync the api keys, providers, skills, and seasons for starters with other authenticated machines? Say you used codex on macos_1 through this platform, and later you used opencode on linux_3, it syncs everything to the new machine, no ssh to centralized machine, no git, none of that duct tape solutions, which they work but cumbersome.
The mistake here is thinking or treating AI tools as special or different in any way. We’ve had tooling that does this for decades. Ansible, puppet, saltstack, NixOS/home-manager, and dozens of other solutions. Even better, writing the configuration is extra simple now, since you can just ask your agent to do it.
True but that’s not the same, think of it like a git but for AI related, I just open whatever harness, I authenticate to that centralized server, and it pulls everything and continue from where I was done last time, seamless and secure. Right now personally I have multiple machines, and I use key managers for the secret API, git for the skills, and sessions? Manually imported/exported. The alternative is using ssh (or herdr or others) and remotely accessing a specific machine where it has everything you need, but it’s not always reliable and better to avoid single point of failure too if anything goes wrong.
There are several projects that injest coding sessions from lots of agents, but for the chat side I haven't found much, except replacements for Claude Desktop. Jan, for example.
Nobody really knows how to product-ize any of these LLM interfaces beyond just chat.
They don't want a traditional UI with buttons and forms and labels because they want the interface to be "chat". The problem is that "chat" is tedious. And the turn-based, linear nature of the chat interaction model makes it even more tedious and unproductive.
I don't think it's that anyone wants the interface to be chat. I think it's that the underlying technology is inherently words in and words out. It's similar to how devices with capacitive screens are most naturally going to support tapping, dragging, and pinching.
It's funny to me that instead of moving existing tech to a word in word out paradigm, people are dropping millions getting the chat to work with the existing paradigm.
Probably inevitable, but seems like a lot of disruption could happen there.
I mean, that's the first adoption phase where you don't start from scratch but keep using the old tech with the bolted on LLMs. It would take some time for devs to build the primitives around the new tech.
My personal hot take is that the product people (engineers too probably) at these companies are just straight up lazy. Yes, a new UI paradigm is hard, but it's been painfully obvious that chat just absolutely sucks. It's also obvious that some DSL-ish thing is possible, something that just does token juggling and the end-user sees some UI behavior.
I know this all sounds abstract. I've been mulling over it for the past year and it's very hard; and LLMs are super janky and inconsistent so it's 100% not trivial. So in some sense I understand why a lazy bottom-of-the-barrel "chat interface" has become the de facto standard.
For the abstract "you can do anything" product, you want an open ended interface, but if I'm, eg, renting a car, is a free form text box really the way to go? The rental company knows exactly what information they need from me before they'll rent me a car, so a free form text box so the computer can ask me questions one by one vs a traditional form, seems suboptimal.
> How in the world would I interface with claude other than sending it messages?
The protocol will of course be `text -> text` for the typical LLM (though some new models are structured, as we saw yesterday with Jev). Even so, the user doesn't need to be exposed to the protocol (almost ever). It's not like I'm crafting a POST request to send this form.
There's a lot of room for window dressing, and I look at AI like I look at the touch-screen. It was a fun technical toy until people (mostly Apple) poured in a lot of serious effort into "ok, how do we make this pleasant to use, intuitive, and genuinely useful?"
The great confusion for Anthropic and OAI is how to break out from pure chat without signalling you in fact intend to eat all of your customer's business too.
Technology folks don’t really understand people and what they need.
This always happens. This is why woz needed Steve.
Steve Jobs is sorely missed tbh. For all the shit he got - he was a true visionary. He lived at the intersection of technology and the humanities… he kept preaching this. And now we are seeing why.
> The moment someone figures out a new modality for LLMs is when we’ll see the next hockey stick.
Honestly, that fills me with fear. LLMs exist to make money to their companies, and said companies are not gonna turn around and say, "you know what, go are going to make an android for each elderly person, that can not only help them with their medications, but that can actually make their medications, tailored to their biologies." Instead, they are going to go for the low-handing fruit of "you know Bob, the guy who makes jokes in meetings but who is grumpy about delivery timelines? Well, we are going to make an android to replace Bob. MetalBob will make even better jokes. The blue model will be able to explain in excruciating level of detail why timelines aren't reasonable. The red model will walk through the cubicles with a whip to ensure everybody keeps working all the time, and nobody goes to pee."
One thing I would like is a tree-like chat structure like reddit / HN. Many times I abandon the direction things have gone but would like to resume at some ancestor or sibling response.
I solved this with task management and got work trees.
The downside of course is branching in got, but I usually don’t go of course more then a handful of tasks.
I had to engineer my own ticket management to keep opus on target. It’s been great for managing work, history, audit trails and commits are tagged with the task id.
I implemented subtasks to deal with the way Claude likes to stage its own objectives.
Then I made enforcement logic in the task manager so tasks can’t be closed out until reviewers have consensus on the same sha.
The adhd that is Anthropic demanded I build it and now I’m knocking out issues faster than Batman.
>Each conversation is a Yjs CRDT tree. It can branch into sub-threads (recursively), and you can drill down, backtrack, edit, undo/redo, and inspect everything: tool calls, approvals, and the raw context JSON going to the model, etc.
Before OpenAI dropped ChatGPT, nobody knew chat would be such a hit. Several labs had versions of these things, Google engineers were getting finessed by a rudimentary model just like in the movie Ex Machina that came out 10 years prior which was satirizing Google
But I don't agree that they aren't product-ized. There are many applications doing calls to LLMs behind the scenes and are hits, leveraging structured data very heavily and not having conversations with users at all. I would say that there is a predictable scope creep from executives to surface a conversational aspect though. We need to bring representation to that so we can point to some other best practice to push back
API calls are expensive and slow, so I’ve shied away from integrating them into my personal projects. And of course there’s always the risk of hallucination.
IMO no one knows what LLMs are capable of yet and the goal post keeping moving every few months that building any specific UIs right now risks rendering them obsolete or too slow.
e.g. We went from somewhat smarter code autocomplete, to asking chatgpt copy paste, to cli agent running inside your project, managing session, to GUI to manage that, to projects where you talk to a "Chief of Staff" agent that manages other sessions, to who knows what's next.
I think the right interfaces for LLMs right now need to be very simple and easy to change/evolve. And chat still seems to be the best default solution.
I wouldn't call copy & paste code from a webui of chatgpt either traditional or boring. I'd call it tedious, error prone and guaranteed to get poor results. There is much better tooling and harnesses to leverage now.
We’re working on this problem at my company. All of our competitors have churned out a bunch of generic chat based agents (basic harness with a few tools connected and different system prompts) to do basic stuff you’d used to do from a wizard or even just a couple of buttons in a form. We’re steering away from this by building full ‘surface’ agents with A2UI and an initial surface message. The users just see a normal looking web application but under the hood it’s all driven by a harness. UI updates are effectively just new events received by the client. Events are either generated by an LLM directly, or deterministically - it’s up to the dev.
I very intentionally have all memory turned off for chat. The amount of times I want to discuss an approach for it to pull memory out and have that steer the decision making is so obnoxious
Reminds me few months ago I asked it about a TV model specs, then later I was doing a UI for something, response was “I modified xyz and now should works perfectly with the HDR in your TV model!!”
When casually a friend asks you to ask your Claude something about topic you chatted about earlier and then Claude brings back a secret you didn't want anyone to know yet.
Yes you can set up a project and then ask a question, but this is tedious.
I turned off Anthropic properly and switched all of that over to OpenAI yesterday. (I use other models for other things too, especially DeepSeek in Pi).
Honestly aside from the voice it uses you wouldn't notice a difference. Switching costs are low, vote with your wallet.
> I turned off Anthropic properly and switched all of that over to OpenAI yesterday.
Same, just a while longer ago.
I much prefer how OpenAI models write to Anthropic's, will probably revisit Anthropic in a generation or two. Context size is more limited, but no critical forgetfulness due to compaction so far, though I also like having plan files around both for future reference and improving chances of success at long form work.
Also tried out Kimi K3, was nice but slow (and apparently routed some requests to Claude anyways), GLM 5.3 was faster and still pretty good but the token allowances were kinda limited.
I would like to have alternative to chat - model and provider agnostic (BYOK), but with history, project, maybe memory, with good search and quality tools.
I know openwebui and i dont want to host it.
codex (their app) is pretty good and lots of banked resets, luna is cost effective and Astra seems better than Fable for many tasks. Most important one is less refusals and I can use it the way I want without the fear of getting banned. I do like Claude Code a lot when it comes to pure coding use cases but the work often touches outside code and I don't want to keep switching
> I do like Claude Code a lot when it comes to pure coding use cases
The harness is great, but opus is such an arrogant little prick that spews out unintelligible word salad. Opus 5 is so bad at communication it amazes me that somebody green-lit it. It's absolutely awful.
The fact that this isn't an acknowledged regression (and Fable 5.1 isn't much better) leads me to believe that people at Anthropic actually like Opus 5's output.
Good to see AI providers finally starts understanding how bad was the user experience. This is not just a Claude issue but same issue exists in ChatGPT.
While it is in the right direction, there is more to be done. I personally think that they should separate Chat and Coding into two separate apps. Both are completely different use-cases. As a developer, I have need for both but I don't like them in the same app. For example, I consider my chats as knowledge base like Obsidian that I want to keep it organized and browse in the future. On the other hand, once I done a coding task, I don't really need that chat anymore. Having coding tasks in the app feel like polluting my regular chats. This frustration led me build my own Chat-only App https://www.vinaa.ai/ where I keep my regular chats in one place organised into projects and folders incl. the ability to auto-organise chats. App is still in early days and I am experimenting a lot to make it a real AI knowledgebase.
The problem is that knowledge workers just need the simplest way possible to automate their work. Currently I feel like Anthro is overcomplicating this. If I'm an accountant, I want to open Claude App, describe my workflow for balancing some books, then have Claude design a reusable WORKFLOW that it runs whenever I ask. The workflow only needs 2 primitive skills to do anything: /work, and /review. Compose these together with guidance docs into workflows, then workflows and call other workflows. It's all just callable skills.
/balance-books hey claude here's the books. Go balance them
Claude then uses the guidance docs and various subagents to review the books, send off parallel workers with cheap models, then review with a more expensive model. Done in the repeatable, correct order with independent review every time. As new models come out, your workflow structure stays the same. You just delete some prescription from the guidance docs.
I still use Google Workspace for personal and work stuff, but for solo agent workflows, I think I have some use cases that would help me personally. I can let agents spin up working documents that I can review, comment on, and iterate over asynchronously.
For example, I’ve spent the last few weeks building an internal ticketing setup for my agents so they stop trying to fix every tangential issue they encounter. I wrote a skill where subagents bubble out-of-scope issues up to a parent orchestrator, which decides whether to log a ticket.
Currently, that setup mirrors between a local SQLite database and an Obsidian vault: agents write to SQLite, which syncs to Obsidian; when I edit notes in Obsidian, it syncs back. Native Claude docs mean I could ditch the Obsidian middleman for remote review—editing tickets from anywhere while keeping the core SQLite state local to where the agents execute.
That makes sense since you treat it as your central productivity manager, but for me all of Claude/Codex/etc are not productivity tools, just heterogeneous compute /for/ the productivity tool :P
1) If you want to make Claude usable how about addressing the fact that OpenAI 20x lasts all day on Astra, yet Claude 20x pips out in 20 mins of agentic with Fable 5.1, which is totally useless.
And unlike Sol: Opus like yoda, anything for unusable he is.
2) The first AI company to reduce their token pricing such that Astra/Fable can run for less than $150/day from credits would see a 5x usage increase and 20x spend increase from me. It’s currently around $500 for Astra and $5000 for Claude.
3) They seem to recently have addressed the fact that their safeguards were OTT so kudos to them for realising and now Fable is usable for wire and memory level programming.
Claude has a jagged UX if you support non-devs. The surfaces before today in the desktop app were:
Chat, cowork, Claude code.
Inside Claude code you’ve got cloud environments and local.
Let’s say as a normal desktop user you wanted to just automate something normal in your work, going to a website (maybe some internal app at your company) getting some data. Depending on what surface you used this will either not work, not work well (cowork), or work quite well but possibly be blocked because of bot controls (local Claude code) or again not work (cloud Claude code)
It was nuts. You can clearly tell these were different teams and products mashed into the same app. Super confusing for non technical users. Hopefully it’s a little better today
I would argue they need 1 product owner for the app surface so I can’t see their org structure so clearly. Those jagged edges and users hitting blocks depending on what section of the app they were in are design failures, I know how a computer works and when using the Claude app it wasn’t always clear to me what the right combo of things to press to get my desired outcomes.
Yeah I don't disagree, ChatGPT/Claude both could really stand a UI/UX pass (and a perf pass for Firefox's sake lol). I was more expressing the perennial fear of power user features being jettisoned in favor of the lowest common denominator.
For me it’s still jagged, but chatgpt never had the weird “cowork” abstraction with the Linux machine.
I don’t talk to non-technical users that use the chatGPT app for work so I’m not sure if that’s been a positive change for them, but I would imagine if that was put in place for Claude it would be better for them. Chat is instant answers, different model, faster; work is more detailed, thinks more, has access to all the tools. This is what I see the users doing, but there is a third mode they use (deep research, exhaustively find me the right answer to a crucial question) that is underserved right now.
All of this is moot to me as a dev, I use terminal, herdr, pi… so I don’t see the jagged edges myself. But regular users do
I really appreciated that the users that I administrate had an option where Claude was not going to get carried away and orchestrate workslop, I will have to completely re-write my usage guides after testing how prompts are routed now.
Feels like the same "let us do _all_ of the thinking for you" messaging that Microsoft has with Copilot.
It's funny how they keep trying to add more interfaces and everything keeps coming back to chat. And of course it would, natural language is the simplest interface there is. Anything they add beyond chat, that isn't placed within chat, that isn't some artifact of chat, is friction.
I really hope ChatGPT doesn't follow suit. Chat and Work serve 2 distinct tasks, and unifying Chat and Codex limits will cause the users to avoid using Chat in order to save the tokens for Codex.
That's the natural outcome. AI will eventually collapse into a single, universal entry point, much like Google Search became the default gateway for the web.
My point with above is to criticize their strategy, I have been losing faith on Anthropic and feeling like they are throwing everything at the wall to see what sticks in a more desperate fashion rather than careful one. This sort of decision to me, alongside all their other initiatives recently with shipping agentic PaaS rather than enabling others to build with their tools has been a miss on thoughtful leadership. Sorry for the snark take, thank you for moderating.
A change for the better. The distinction made it unnecessarily complex for users. The ChatGPT desktop app inherited the same separation. Having to change between ChatGPT for day to day questions, to Codex for everything else always seemed too cumbersome.
People keep asking for this w/ Codex, too, and I really regret that both labs seem inclined to listen.
If you ask a thinking/research type question in 'Chat' versus 'Work' mode in these products -- say, something complex about politics, or do a multiturn business strategy, or want to work thru a new concept, you get very different answers.
The harness, steering, etc. in the chat/reasoning products is so much better for this type of question (that doesn't require code as a primary substrate).
Even as someone who mainlines like, 7 coding agents at all times, I regret that productivity fever will mean the regression of think-first-act-later AI UX.
I hope not Chat is still all you can eat and Work uses Codex tokens. I can see why they would want people to think they wanted this without thinking about how it currently works. If Chat starts eating tokens then there is not much reason to use Chat other than most people don't live in the command line like developers. Claude Chat / Cowork already eats tokens either way so not much of a change really. I have never used Work since if I want access to local files the CLI is a much better interface combined with an IDE, but can see the appeal to non-developers.
This is my main reason for not using Work. I was tempted to use a few "scheduled tasks" in ChatGPT, but once I found out they used Codex tokens I stopped creating new ones.
I already burn through Codex too fast.
I built a terminal control plane and added an MCP server that runs on my Mac. It lets MCP clients create terminal sessions, send messages, read the contents, etc on my Mac.
The terminal sessions are running Luna on XHigh (on Pi) and GPT 6 Pro on ChatGPT.com is the main orchestrator.
Since ChatGPT and Codex have different limits, I can have 6 Pro orchestrate the Luna sessions. Using voice mode to talk back and forth about the software factory is really nice. Will be open-sourcing soon.
I recommend using Work sparingly because sometimes it is the better choice, but I do ration it. With Claude I simply don't use the web/app chat at all (and I have no idea what Cowork even does ngl, never used it) I think there's a Claude Design? No idea why I would need it (mostly, why I would spend tokens on it lol)
Claude Design is good at making a bunch of UI pages in a single design look consistent, and it's better at making an initial starting UI than just asking Claude Code. Think of it like Anthropic's version of Figma. It honestly works pretty good and I use it often and feel that it saved me a lot of time. It's pretty token heavy though.
I like the web chat UI for quick one off questions where I don’t want it to have the context of a local project I’m working in.
The OpenAI folks have been clear that they are merging them by the end of the year, obviously TBD what they do with quota.
GPT Chat seems smarter for business questions or planning that does not require iteration or wrangling data sources.
But Work gives the model a computer instance it can use for programmatic data access, API munging, etc.
Success cases I’ve had: chat - read the reports in my Gmail and put a counter proposal together.
Work - cross reference all the options with commute time using traffic aware map data. Ground it on drop off / pick up times from each option’s stated rules and hours. Create a table with commute and idle time break downs.
The fact that it doesn't work well today doesn't invalidate the user need. Not everyone wants a terminal style interface. Traditional UX makes more sense especially if say you are accessing existing plugins and say kicking off a scheduling task - no way would I want to do that via a chat interface if I have the option.
Agree, and it allows ordinary users to have a taste of both worlds, instead of being like: "eww, what is this black box and why doesn't my mouse work"
I see Work as the mode where you can generate and edit files without manual intervention.
I tell chat to recommend me some books based on some of my favourites. Chat does it. I read some of them, liked some, disliked few. I give chat that information, it recommends me based on those, now I don't know which of the entries are relevant or not.
I don't want lock-in, so I try my best to tell it to generate copy-able MD syntax that I have to manually paste to an editor.
With Work mode, I tell the model to just maintain an append only markdown file, where it writes "still recommended", "not recommended any more", etc. beside titles. It does this after each turn.
This is just an example. Work - I see it as a mode that can maintain, update, and edit a file.
It's very convenient.
I agree that they should be separate modes, but presenting it as a toggle in the app is a terrible user experience.
The model itself should figure out how much “power” it needs to run the query. I’m just asking a question. Whether you need to do a simple internet search or spin up a VM and run scripts is up to you.
> The harness, steering, etc. in the chat/reasoning products is so much better for this type of question (that doesn't require code as a primary substrate).
I had this same thought recently. Is it possible to 'recreate' ChatGPT/Claude Chat easily in Codex or Claude Code? I assume it would roughly be just swapping the system prompt. Anthropic seems to publish theirs but I don't think OpenAI does
https://platform.claude.com/docs/en/release-notes/system-pro...
I sometimes get very different answers if I ask the chat product the same question multiple times.
Isn't that expected since LLMs are inherently non-deterministic?
LLMs are inherently deterministic. The way everyone deploys LLMs leads to non-deterministic results, but there's nothing† stopping providers from offering deterministic evaluation if they choose.
All the sources of randomness are under the control of the provider, even if today's deployment structures mean providers introduce extra randomness due to the concurrent nature of the evaluation. Serialise the computation, feed it from a pRNG, and you have a fully deterministic result. But providers don't want to offer a deterministic result, and especially not one as fragile, expensive, and inefficient as a full serialisation would be.
†: For variants of "nothing" that include cost and deployment challenges.
This is technically true, but when people talk about randomness, its not only about same input-> different output, like temperature>0 and the things you said.
Its also about very similar inputs -> different outputs. Even with everything you said, yes, same input would result consistently into same output, but sliightly different input and you might get completely different/semantic answer.
That's "chaotic" rather than "non-deterministic".
There's deterministic enough, and then there's computer science non-deterministic. If I ask for a Todo app, I'm going to get a Todo app, even if the buttons get moved around and the background color of it is brown instead of purple if I ask today vs 6 months ago. If the AI completes the phrase "the capital of France is..." with anything other than Paris, something has gone more wrong than usual.
LLMs are chaotic pure functions. Their input is usually randomized.
In one of the earlier iterations of this they at least made the UX distinguish Codex vs. ChatGPT which I really enjoyed.
I must admit I spent a little time in ChatGPT this weekend, thinking I was in Codex. After some unexpected behavior (disappointed) I realized my mistake.
The fact that people asked for it means that there is a demand for it. Maybe you aren't the target audience? One size can't fit all. People just have to adjust and go.
Yes the point is that they wouldn't have to adjust unless they are merged into one. Precisely because one size can't fit all, yet they insit on one size now.
How did this new account `vertigoruntime` get three posts on the front page, all in the last day?
- This post
- The DeepMind Institute https://news.ycombinator.com/item?id=49727659
- Mistral X Mozilla: Private, Multilingual AI Browsing https://news.ycombinator.com/item?id=49723408
based on their account submission history they have a 25% hit rate (!) Seems very, very weird.
It would appear no one else submitted the posts in question, so the most obvious answer is the account is finding interesting articles that no one else has found. It would be one thing if it won out of many dupe posters, but that doesn't seem to be what's happening.
This is valid.
1. each one is a first-party vendor announcement, submitted very close to the announcement and under the vendors headline.
2. Titles are "verbatim" vendor headlines. No editorializing which follows HN guidelines to the letter.
3. Someone; a script/agent is watching vendor feeds and X accounts then submitting within minutes of the posting.
4. Altman/Navier–Stokes post got [duped] so prob not inside thing.
High frequency karma farming.
You aren't serious about high-frequency karma ("HFK") unless you colocate your equipment with the routing for your sources so you can beat your HFK competition to a submission by at least 200ms.
We will have reached HFK-ception when the bots are fighting to be the first to submit "How I used AI to build a High-Frequency Karma Farming Bot on Hacker News."
Then a week later: "Paul Graham announces a 100000$/month high priority access plan to the HackerNews via YC API." :)
YC already announced something along these lines.
re: https://www.ycombinator.com/early-access
Oh my... I should be more careful in my predictions :)
Their scraper is physically really close to the internet building!
I just have RSS feeds of news in AI circles that may be cool, built an entire tracker around it - helps me pick up new stuff on time !
What made you just start posting on HN? Welcome and thanks for the links!
this post became a great "Show HN" for it
Any chance you could drop the set of feeds? Would be very interested!
Totally unrelated but I google'd their username to see if there was a clue, and I had no idea this feed existed:
https://news.ycombinator.com/noobstories
...does noob here just refer to a new (green) user? Or is it something else?
Noob comes from newb, which comes from newbie. Some argue it comes from "nubile" but I don't personally buy that etymology. I distinctly remember "newb" followed by "noob" but that was eons ago.
Always thought it was "new" but made to sound like "boob" because it's funny.
And not related to Safire's "nattering nabobs of negativism," the originals of whom SF's Nob Hill eponymizes.
Technically “new user” I think.
maybe someone wrote a classifier (or LLM wrapper) to find posts that will rank on hn
They are all valid front page submissions by the looks of them and they appear to be the first ones submitted. It wouldn’t be hard to do if you cared enough to configure notifications for frontier announcements. If someone were using technology to influence submissions or misuse HN, that would be a concern but I don’t see any evidence of that.
To directly answer your question, my guess is they have an alert set up for various places on the internet, and are able to post the links faster than anyone else. Maybe that part is automated.
AI product updates probably tend to rank very well here. If you look at the domains the user is submitting, they’re pretty normal: twitter accounts of tech CEOs, big company blogs, etc.
If I saw a bunch of blogs with affiliate links or landing pages for slop products I’d be more suspicious.
> putting my tinfoil hat on and awaiting for replies
I question all the time what dictates the posts I see on the front page every day, and whether or not it's truly based on organic interest. i wish i could say i up/down vote frequently but i do not and i wonder if that's true for most people.
You can make your post on top in some services by paying 5 dollars on fiver. I asked before about HN, and someone said it is hard to do... Maybe they found a way?
These kinds of updates always have this romantic scenario of someone having Claude develop a presentation or something on their way to work between multiple devices, which actually feels a little sad and does not align with what happens in my life at all.
This is a nightmare scenario to me.
"Claude, prepare me a presentation on XYZ."
I get to work, go straight to the meeting room, and pull up what it made to present.
It's barely coherent nonsense. Lots of irrelevant details, buzz words, wrong charts or confusing phrasing. Obviously LLM output.
I read it out.
When I'm done, I get a question about one of Claude's incorrectly inferred details.
The shame instantly kills me.
This is a scenario I've seen play out with coworkers. Except that last part, instead of dying or owning up to the mistake of trusting LLM output they waffle. Their shame circuit is broken.
Someone I know does this, and seems to think it's a flex, because he bragged about it being "the first time I've actually seen this presentation" a few times during the presentation.
It was all awful.
I feel like we need to normalize walking out of meetings like that. If you have no idea what you’re presenting why is anyone in the room?
At my previous company, a distinguished engineer did this. Used generated slides he hadn't looked at until presenting to the entire staff+ engineer group. The slides weren't good and he mostly ignored them.
I would immediately stand up and walk out of there. That's my time you're wasting!
It’s not your time anymore. It’s the corporations time ;)
That is incredibly disrespectful of them _unless_ they prepared in other ways or are some expert who doesn't really need slides.
I don't really know what the right response is, though. Walking out would be seen as way too hostile
“I came in unprepared. Please hold your applause.”
A Canadian politician did this during a speech in front of the house/parliament or whatever. In the middle of the speech the LLM wrote "Here's a more natural, flowing version of that section that reads like a legislative speech" and he read it verbatim and then continued on, not even realizing what he just read.
Previously having a deck was basically a signal that said “I took some time to prepare for this meeting and put together some coherent talking points” even if the deck itself was just a bunch of of text outlining what the speaker was saying.
Now that people are generating decks with AI, they’re basically worthless. I don’t need a bunch of bullet points you didn’t take anytime to think about on the screen while you talk about something you didn’t prepare for.
Why on earth would you present a presentation that you haven't reviewed ahead of time?
I wish I knew, but I've definitely seen it happen.
"What does this bit mean?" "I dunno."
Why even have a meeting? That's just a waste of time when agents are much better at reviewing plans anyway. You don't even need presentations anymore. Executive officers can just create a review agent with a "special sauce" prompt (essential to justify their high salary) and auto sign off on a plan if it passes the crucial test: "is this plan good?" Finally we can have true business automation and people can go back to much more important things, like looking at KPI reports and being impressed or unimpressed.
That sounds incredibly frustrating. It's bad enough in code reviews, but is that approach to work being tolerated in meetings these days?
Because you think ChatGPT will do the job for you, and sadly, the quality drop doesn’t universally hurt people
That is incredibly frustrating for a coworker to do that. Even if you do use an LLM to generate all the slides it is your job as a presenter to go through and make edits and actually practice the presentation.
If you are just presenting it for the first time after seeing it you are going to just be reading off the slides, which is a waste of everyones time.
PowerPoint Karaoke used to be a fun distraction; now it's literally people's job.
This made me laugh out loud.
All the Apple Intelligence commercials were shamelessly this too. Felt like such poor branding for apple
> It's barely coherent nonsense. Lots of irrelevant details, buzz words, wrong charts or confusing phrasing. Obviously LLM output.
Plenty of that even before LLMs too though.
Defaults matter a lot. Before you had to intentionally be lazy to put together garbage and call it a day. Now your boss is telling you to use AI to "speed up." The tool you're pointed to presents like it can make the slide for you and claims to do so. Every signal is screaming that you're supposed to be a lazy fraud, and people will, on average, take the path of least resistance.
I'm attending a conference that I also attended last year, and it's night and day.
This year, a good 40% of workshop slides are fully LLM generated, incomprehensible, and almost not matching what the speaker is talking about at all. Last year, while there were some poor presentations, standards were much higher.
Because nobody actually knows what anybody else’s job consists of. But everyone is familiar with how jobs work in the ‘romcom job universe’ where workplace stakes are generally driven by ‘I have this really important make-or-break presentation due Friday!’
I suspect it’s because writers don’t normally have jobs and the only time they are ever in offices it’s to deliver a pitch for a project.
"Plan a trip to Italy next month and book the flights"
I see a version of that all the time and would never, ever do it.
These are modern infomercials. Concrete pains ("planning a vacation is hard!"), shallow solutions that don't stand up to real-world use, and, in the end, a promise left unfilfilled. (And a thing in your drawer that takes up space.)
honestly, one day i will understand why these guys go so hard on planning vacations.
one of the best parts of vacations is actually planning it! trying to find restaurants, organizing your days, trying to fit activities -- that makes you look forward to your time off.
really, i want to automate the boring parts of my life (did i really pay rent this month?) not the things that make me happy.
> one of the best parts of vacations is actually planning it
For some. Others hate or fear it. For others it just feels like work.
Imagine the fear these people will feel walking in to a vacation planned by … a ROBOT!
Social media has fried some people’s brains so much that it is incredibly anxiety inducing to make many typically mundane or routine choices, for fear of missing / missing out on something.
I think they're grabbing at universally-understood frictions that people associate with pleasure or anxiety … or, most powerfully, identity.
We all know the feeling of wanting to be successful, or to go on a great vacation. Showing these scenarios creates strong emotional response in the audience - we're now the people winning at work, going on a great vacation. This is especially important when you're selling something dull. See: insurance.
"Find a primary care physician in my area who is taking new patients and accepts my insurance. Give me your top three candidates, and book an appointment within the next six weeks with your top choice."
Claude: Would you like to go to Bermuda instead?
idk I thought the same until I got personal amex support at work, now it’s my roofline on how useful/trusted an agent could be. Now I book all my travel through them, and it’s an absolute blessing
Exactly! Imagine if it booked like a 3000 dollar 1st class flight because it thought it was providing you with "the most comfortable and hospitable experience as possible."
[dead]
I've found basically all of the release presentations are completely irrelevant and seem to target a very narrow subset of the population. They're always reserving something like a restaurant or a sports venue or generating a new slide deck for their business meeting. Perhaps these companies would do better if they could make a use case for regular people. You know, the 90% of the folks out there working normal jobs and not spending infinite money. I don't think I know a single person who has made a restaurant reservation more than once or twice (mostly never) in their entire life. Maybe I'm the weirdo here, but I can't relate to any of these demonstrations.
I regularly make restaurant reservations and create slide decks for my work, and both of those seem quite common among people I know. What would be a normal people use case for Cowork in your view?
I'm probably about 90%ile income for my household but live in a HCoL area and have two children in daycare. I rent, don't own, and I have enough that I can buy a coffee without thinking, but will be sending my kids to public school, can't just up and go on a vacation.
I make restaurant reservations 7-8 times a year? My wife and I go on dates and you simply aren't getting a table where I live if you don't. Fyi we typically spend $120 or so on those dinners.
And does having an AI call the restaurant instead of you doing it sound cool to you?
I’m not the person you asked but I’m also confused by this use case.
Most restaurants have a web UI which is very simple and easy to use.
Here in Bangalore, you can usually book directly from the Google Maps listing. Or sometimes click a link that will open in Swiggy/Zomato, and you just book from there.
At least in this city, reserving a table using the apps you already have on your phone is faster than using an LLM. Maybe these products are for people who know about the existence of Claude and ChatGPT but not Google Maps?
If I'm booking a restaurant I'll probably be discussing the date/time options with the people I'm going with, and I can't see how adding a chatbot to the mix is meant to help?
If it works reliably, sure? But like, it's not something I'd pay for.
I'm retired and write code as a hobby, not presentations or docs, but it fits how I use exe.dev to make a code change. I will have a conversation with the AI, then say "Please go ahead" and close my laptop to do something else. Since I use web browsers exclusively for this, I can check on its progress from my phone, tablet, or laptop.
(They have iOS and Android apps too, but I've never bothered to install them.)
I use Claude/opencode and Orca and Hermes across my vpses, Macbook, Linux, and iOS devices. This is exactly how I want to keep working.
Would be interested to hear more about your setup!
VPS is production. Linux dev box is my local big box. Macbook just lets me move about and do iOS development. Orca lets me run agents either on linux box or locally on the macbook depending on how I'm feeling, and Orca handles all the worktrees & sub-agents for me (via Opencode). Everything runs in a tailscale network, ssh connects all the things.
My github repos all have CONSTITUTION.md files that keep multi-agent dev grounded. Production runs trigger github actions that automatically check logs daily, file bug reports, etc. I launch Orca and literally just type "checking in" and the CONSTITUTION.md file governs the scanning of github issues and prioritization of which issues need to be addressed. Another process prioritizes and bundles them. When I type "checking in" the Orca worktree handles the launching of sub-agents that fix things. Occasionally I weigh in with an opinion or pick a recommendation. If I'm in the mood to pay attention I'll ask it to find another round of bugs and let's keep going. Often I just let it close after the first round. When I tell it "done?" it does a full regression and a production box review. Code auto-deploys to production twice/day. I don't regularly push to production manually.
On Fridays the CONSTITUTION.md kicks off a strategy and roadmap review when I type "checking in". As I have ideas I add them to the roadmap in one of the worktrees. Generally nothing happens until they're reviewed on Fridays together, unless I push one manually.. which happens.
Hermes runs on the production box. I have a few skills and quick commands that lets me check in on production runs and the status of things. It's read-only by design, basically my version of a dashboard. If I have a thought, idea, research link, or question it'll write to a github issue and we'll deal with it in the next check-in.
Orca lets me do all this remotely from an iPhone, and I talk to Hermes via Telegram. I generally don't INITIATE new work using either of these. I report issues/ideas and let the process do the things. But I can, and have.. I just tend not to. The whole thing churns, and so my inclination to "jump in and do a thing" is less than before. My inclination now is to toss things into the machine and let the machine work the schedule. I'm at a point where I could easily automate 80% of this and do my manual things 1x/2x week and I think progress would stay steady. I'll get there at some point, but I like the level of engagement I'm participating in now.
Occasionally I use Claude Code as a watchtower review of things, or do a wholesale code review, or do a review of logs. But deepseek is much better at building this machine -- I built something manually over time and through a ship of theseus process it got ugly. Deepseek reviewed the situation and wrote the CONSTITUTION.md and designed the processes independently. It favors deterministic scripts for process but launches them via LLM and monitors for exceptions; also, no memory system which also disqualifies Claude from being at the center of this. Memory plugins didn't work well because instructions start accumulating all through the chain of instruction files (AGENTS.md, memory, etc) and that gets very undeterministic very quickly. Instructions are written once, reviewed/audited intentionally.. it's best not to let LLMs learn and write their own dynamically (yet).
I've since used it as a template for other projects I've started -- they work the same. Currently building the GTM agent that will handle all SEO, marketing, keywording, etc for my projects -- it'll be a significantly autonomous Hermes agent. Don't really need Claude for any of this and my Deepseek bill is $50-100/mo.
You are not the typical office worker.
A lot of dev tools now support remote execution, even those used in corporate environments. And for orgs that need locked down dev instances under regulatory control, these tools fit that bill too. I don't think I'm operating outside of norms that are used across all kinds of industries. It may not be the most common config yet, but these tools are building for customer bases that exist.
weirdly enough, that aligns exactly with my life.
Two years ago, I had no commute, and presentations were tedious ( i was NOT a good google slides user ), I kinda prefer this world for now
It aligns with me pretty well. Not exactly presentations, but it’s great to be able to do stuff on the go when an idea or motivation strikes.
Seriously a trillion dollar company with zero product management expertise.
I do this pretty regularly. It's great.
Just like every demo is some pathetic one shot trash game that nobody would play, or something equally inane.
Hi, this is my team! Happy to answer any questions.
There's a lot in this launch, but the core idea is to simplify the product while giving users access to more capabilities. You no longer need to know ahead of time how much work a conversation might involve. If you're at your computer, Claude can use your local files and apps. If you close your laptop, Claude can keep working on its own computer.
This launch also lets you use Claude Design, Claude Docs, and Claude Slides directly from conversations. That's possible because we made Artifacts much more powerful: whenever Claude makes you an app, website, design system, or anything else, it can deploy an artifact with multiplayer features and databases.
As many of you probably know from your own work, giving users more power while making the experience simpler is really, really hard. It took many iterations to get to this version. We're far from done, but I expect people will be able to do much more while having to think about it less.
From a user experience standpoint this sounds like a clear win. Congratulations.
With AI safety top of mind so much lately, I can't help but notice the announcement does not address this.
With "Chat" mode, there was a user expectation that session had only limited capability to produce unexpected side effects, read sensitive files, etc.
With "Cowork" mode, it seems like more powerful capabilities have been on by default, requiring deep settings and safety understanding to disable if desired.
Merging modes feels like it is removing a simple and easy to understand risk management tool. How does the combined mode help users understand, manage and feel confident about what risks they are accepting?
It's not a bad idea, however if Cowork, Designs, Docs and Slides are still different products, it'd be great for the user to know they are being routed to that particular product automatically because of what they were asking for in their prompt.
> Claude Design, Claude Docs, and Claude Slides
Why/when would I ever need any of these? Can't y'all focus on useful features instead?
There are a lot of people using this outside of the coding space, most of my friends use these to create decks and things like that to present. Mind you they never double check information and numbers.
Claude Code has almost always been able to do a passable job of this. I wonder if this new version will need less cleanup. Claude Code would write python scripts to generate the deck. It at least formatted cleanly with templates.
> create decks and things like that to present
Not really proving their usefulness there :P
I guess they didn't realize you're their only user!
How are you determining how much oomph Claude gives a particular question? When I ask Claude Chat to e.g. summarize a document, it does this and returns a summary inline, whereas cowork does a bunch of tool calls and intermediate steps, eventually writing a document to disk. How does new claude know how deep to go?
Why is Claude Code still proprietary, while Codex is Open Source? What would it take to fix that? This is a major reason to pick OAI models over Anthropic (because while you can use an open agent with any model, the AI labs charge much more for that interaction style).
I don't know how proprietary Claude Code really is.
Just wait for the next leak!
That's exactly the point: it's not as if anything is being gained by it being proprietary, because nothing is actually secret.
I liked the ability to explicitly only chat, without the possibility of Cowork activating since I have never wanted that. It's unclear how I can still guarantee that now.
This! My use of the web chat/app was already basically 0, now it will be 0.
Will there be a Claude Sheets too? I find myself using Claude to work with spreadsheets more than almost any other document type. It does okay now, but feel like it could be even better, especially with visualizations and formulas.
What sort of issues do you see with spreadsheets? I almost never use Claude or Cowork. Just Claude Code. When I've generated or modified spreadsheets it tends to build python scripts for it which makes regeneration trivial. I have to fight a bit on formatting sometime, but I find if I format it the way I want I can get CC to read the file follow that in the future.
Oh sheet son.
I have a question about project memory. Last time I tried using Claude to organise my chats into Projects, it turned them into isolated silos. I asked a different chat to refer to what we'd discussed in the other chat, and it said it wasn't allowed to access it.
This was pretty frustrating. By trying to organise my chats, I actively made them worse. ChatGPT at least gives you the option to have either open or closed memory. Is this being considered?
I rarely use Claude or Claude Desktop. Almost 100% of my usage is Claude Code because it ties into all of the private infrastructure I've built around it. I've had decent success using CC for generating Office documents of various formats. It tends to reach for a python script and library to generate them and the output is typically clean and easy to apply a template to. I have had issues where I need to manually modify the files to demonstrate format and style, but that's also something that can built into context after with references to templates.
Where do you see Claude doing better than what I'm seeing with Claude Code in these circumstances? Are there any reasons for me to leave my pretty heavily customized CC environment for the other versions?
As part of all the work being done I really hope the sidebar gets more user control. There is a bunch of stuff that I don't intend to use myself but take up first class space (designer is always there, I never need to bring up artifacts, etc). I'd like to reduce the sidebar to what I use most often.
Specifically around the work UX. Trying to bring up scheduled tasks is really painful and required a lot of clicking around instead of just seeing the task results under my project. I hope this is being looked at.
Lastly, and this is a real nit. Let me turn off the "tips" when stuff is being worked on. I'm already paying, you don't need to force a rotating feature advertisement into the interface.
Can you standardize with all of your tools on a common registry key and settings paradigm and same plugin/skills paradigm and common method for enterprise delivery? It’s a pain in the neck getting all of this set up and deployed throughout the org.
In the web interface, are there any plans to support attaching local directories through File System Access API and accessing internal (as in, not reachable over Internet) MCPs, assuming correct CORS setup, of course?
Thanks for participating in the discussion Felix. A big issue for me - not sure if others suffer from this - is that it’s next to impossible to track work in progress, especially with when there are artefacts involved or reports etc. Things that require multi-day work of back and forth and reflection.
Chat is a horrible interface for this, having to scroll up and down across lengthy conversations to try and pick up a thread.
I’m really sorry but I really fail to understand why this massive push to “chatify” everything. A task with its own context and multiple chats feel like a much better approach to work. Even if the work is spawned from a main chat window - eg “Claude we need to work on XYZ” and it creates a task to track this piece of work.
Further, recurring tasks again are really really hard to manage with a chat interface - which of the 10 chats has that question that the recurring task raised?!
And of course all this UX debt will likely stand in your way of building reactive items - ie spawning a task in response to some outside event. Think an inbound email being handled by a prompt that triggers a task and creates a draft response ready for my review and approval. Very hard to do with chat windows.
this is good change for ux. thanks
I'm operating Team accounts, are there changes on how org-wide skill/plugin get loaded and used?
[dead]
[dead]
imo Claude is ok on taste, just with many sloppy bugs. ChatGPT/Codex is worse on taste.
A bit meta, but this is one of the first posts in Anthropic's official blogs I've read that has distinctively Claude-style prose, thought less prominent than usual:
And the unnecessary school-project level verbosity:Most normal users have no idea whether to choose chat or cowork. This makes sense for 95% of customers.
Exactly, a change like this will always be abhorrent to the 5% of technical users (aka this forum), and a net gain for the other 95%
I am a technical user and I literally last week was trying to work out what the difference between ChatGPT Chat and Work was and couldn't figure it out (other than they bill you differently)
The 5% of technical users should not bear the brunt of the technical ineptitude of others.
99% of the 5% can use claude code.
Is there a mature centralized platform, self hosted, that sync all your “AI stuff” across multiple devices? So regardless of the harness, it sync the api keys, providers, skills, and seasons for starters with other authenticated machines? Say you used codex on macos_1 through this platform, and later you used opencode on linux_3, it syncs everything to the new machine, no ssh to centralized machine, no git, none of that duct tape solutions, which they work but cumbersome.
The mistake here is thinking or treating AI tools as special or different in any way. We’ve had tooling that does this for decades. Ansible, puppet, saltstack, NixOS/home-manager, and dozens of other solutions. Even better, writing the configuration is extra simple now, since you can just ask your agent to do it.
True but that’s not the same, think of it like a git but for AI related, I just open whatever harness, I authenticate to that centralized server, and it pulls everything and continue from where I was done last time, seamless and secure. Right now personally I have multiple machines, and I use key managers for the secret API, git for the skills, and sessions? Manually imported/exported. The alternative is using ssh (or herdr or others) and remotely accessing a specific machine where it has everything you need, but it’s not always reliable and better to avoid single point of failure too if anything goes wrong.
config is easier than the memory/state/history stuff.
I haven't found this yet but want to.
There are several projects that injest coding sessions from lots of agents, but for the chat side I haven't found much, except replacements for Claude Desktop. Jan, for example.
Nobody really knows how to product-ize any of these LLM interfaces beyond just chat.
They don't want a traditional UI with buttons and forms and labels because they want the interface to be "chat". The problem is that "chat" is tedious. And the turn-based, linear nature of the chat interaction model makes it even more tedious and unproductive.
I don't think it's that anyone wants the interface to be chat. I think it's that the underlying technology is inherently words in and words out. It's similar to how devices with capacitive screens are most naturally going to support tapping, dragging, and pinching.
It's funny to me that instead of moving existing tech to a word in word out paradigm, people are dropping millions getting the chat to work with the existing paradigm.
Probably inevitable, but seems like a lot of disruption could happen there.
Text based workflows are king
I mean, that's the first adoption phase where you don't start from scratch but keep using the old tech with the bolted on LLMs. It would take some time for devs to build the primitives around the new tech.
My personal hot take is that the product people (engineers too probably) at these companies are just straight up lazy. Yes, a new UI paradigm is hard, but it's been painfully obvious that chat just absolutely sucks. It's also obvious that some DSL-ish thing is possible, something that just does token juggling and the end-user sees some UI behavior.
I know this all sounds abstract. I've been mulling over it for the past year and it's very hard; and LLMs are super janky and inconsistent so it's 100% not trivial. So in some sense I understand why a lazy bottom-of-the-barrel "chat interface" has become the de facto standard.
Can you elaborate? How in the world would I interface with claude other than sending it messages?
For the abstract "you can do anything" product, you want an open ended interface, but if I'm, eg, renting a car, is a free form text box really the way to go? The rental company knows exactly what information they need from me before they'll rent me a car, so a free form text box so the computer can ask me questions one by one vs a traditional form, seems suboptimal.
But we're talking about Claude right? Not some bespoke app powered by AI? The interface cant only be wide enough to rent cars.
> How in the world would I interface with claude other than sending it messages?
The protocol will of course be `text -> text` for the typical LLM (though some new models are structured, as we saw yesterday with Jev). Even so, the user doesn't need to be exposed to the protocol (almost ever). It's not like I'm crafting a POST request to send this form.
There's a lot of room for window dressing, and I look at AI like I look at the touch-screen. It was a fun technical toy until people (mostly Apple) poured in a lot of serious effort into "ok, how do we make this pleasant to use, intuitive, and genuinely useful?"
So ... what?
Dictation? Neuralink thought-to-text?
[dead]
So what kind of interface would you prefer?
The great confusion for Anthropic and OAI is how to break out from pure chat without signalling you in fact intend to eat all of your customer's business too.
It’s not surprising.
Technology folks don’t really understand people and what they need.
This always happens. This is why woz needed Steve.
Steve Jobs is sorely missed tbh. For all the shit he got - he was a true visionary. He lived at the intersection of technology and the humanities… he kept preaching this. And now we are seeing why.
This is so true, it really struck me, thanks. For me, this is the most important post of the week. By far.
This has been my ongoing gripe. Someone put chat in front of a transformer model. It exploded like all prior versions of “AI chat” did. Now what?
The moment someone figures out a new modality for LLMs is when we’ll see the next hockey stick.
> The moment someone figures out a new modality for LLMs is when we’ll see the next hockey stick.
Honestly, that fills me with fear. LLMs exist to make money to their companies, and said companies are not gonna turn around and say, "you know what, go are going to make an android for each elderly person, that can not only help them with their medications, but that can actually make their medications, tailored to their biologies." Instead, they are going to go for the low-handing fruit of "you know Bob, the guy who makes jokes in meetings but who is grumpy about delivery timelines? Well, we are going to make an android to replace Bob. MetalBob will make even better jokes. The blue model will be able to explain in excruciating level of detail why timelines aren't reasonable. The red model will walk through the cubicles with a whip to ensure everybody keeps working all the time, and nobody goes to pee."
Perhaps we could see a Yahoo Pipes style interface and feature set, handled by LLM (when simple transforms aren't an option).
One thing I would like is a tree-like chat structure like reddit / HN. Many times I abandon the direction things have gone but would like to resume at some ancestor or sibling response.
I solved this with task management and got work trees.
The downside of course is branching in got, but I usually don’t go of course more then a handful of tasks.
I had to engineer my own ticket management to keep opus on target. It’s been great for managing work, history, audit trails and commits are tagged with the task id.
I implemented subtasks to deal with the way Claude likes to stage its own objectives.
Then I made enforcement logic in the task manager so tasks can’t be closed out until reviewers have consensus on the same sha.
The adhd that is Anthropic demanded I build it and now I’m knocking out issues faster than Batman.
I'm not affiliated, and did not actually try it, but your comment reminded me of the juggler coding agent, discussed here: https://news.ycombinator.com/item?id=48883305
>Each conversation is a Yjs CRDT tree. It can branch into sub-threads (recursively), and you can drill down, backtrack, edit, undo/redo, and inspect everything: tool calls, approvals, and the raw context JSON going to the model, etc.
Gemini has something kind of like this. In a chat, click on the "..." icon next to one of Gemini's responses and select "Branch in new chat".
It creates a new chat which has the conversation up to that point and a link to the conversation you branched it from.
The UI isn't tree-like, though. It's more like Git where each branch name is just mixed in with the rest of the soup of references.
GitHub Copilot in VS Code has a similar feature to "Fork" a conversation at any point in its history.
Yes this would be very useful. Or at least being able to selectively compact a linear part and continue from before.
Often times I just sort of wish to undo.
You can do that right now in ChatGPT? Every answer has a "Branch in new chat" button. You can go back to any point and branch again.
Before OpenAI dropped ChatGPT, nobody knew chat would be such a hit. Several labs had versions of these things, Google engineers were getting finessed by a rudimentary model just like in the movie Ex Machina that came out 10 years prior which was satirizing Google
But I don't agree that they aren't product-ized. There are many applications doing calls to LLMs behind the scenes and are hits, leveraging structured data very heavily and not having conversations with users at all. I would say that there is a predictable scope creep from executives to surface a conversational aspect though. We need to bring representation to that so we can point to some other best practice to push back
Can you give examples of wrapped LLM logic?
API calls are expensive and slow, so I’ve shied away from integrating them into my personal projects. And of course there’s always the risk of hallucination.
For converting unstructured data into (more) structured data, they're phenomenal.
Yes, they are expensive, but flexibility has a lot of value.
The keyboard in iOS since 18
IMO no one knows what LLMs are capable of yet and the goal post keeping moving every few months that building any specific UIs right now risks rendering them obsolete or too slow.
e.g. We went from somewhat smarter code autocomplete, to asking chatgpt copy paste, to cli agent running inside your project, managing session, to GUI to manage that, to projects where you talk to a "Chief of Staff" agent that manages other sessions, to who knows what's next.
I think the right interfaces for LLMs right now need to be very simple and easy to change/evolve. And chat still seems to be the best default solution.
We have all that? I'm still copying and pasting off the WebUI of ChatGPT.
Traditional and boring works for me.
"Traditional and boring works for me."
I wouldn't call copy & paste code from a webui of chatgpt either traditional or boring. I'd call it tedious, error prone and guaranteed to get poor results. There is much better tooling and harnesses to leverage now.
We’re working on this problem at my company. All of our competitors have churned out a bunch of generic chat based agents (basic harness with a few tools connected and different system prompts) to do basic stuff you’d used to do from a wizard or even just a couple of buttons in a form. We’re steering away from this by building full ‘surface’ agents with A2UI and an initial surface message. The users just see a normal looking web application but under the hood it’s all driven by a harness. UI updates are effectively just new events received by the client. Events are either generated by an LLM directly, or deterministically - it’s up to the dev.
Why would I want that for a form with buttons?
It sounds painful to debug
Do you know that you can have multiple chats at the same time?
Doubling the tediousness does not solve the interface problem.
Chat is only tedious because it’s slow. If models responded in 300ms it would be the ideal interface.
"Chat" makes it feel more like you're talking to a human. They need people to believe AGI is imminent when using the product.
[dead]
To Anthropic: I hope you don't merge Claude Code and chat, I like keeping their memory separate.
I very intentionally have all memory turned off for chat. The amount of times I want to discuss an approach for it to pull memory out and have that steer the decision making is so obnoxious
Chat, create a new landing page for the company!
"Sure! And since you were asking me about lobsters yesterday, I'll make it red and seafood themed!"
Reminds me few months ago I asked it about a TV model specs, then later I was doing a UI for something, response was “I modified xyz and now should works perfectly with the HDR in your TV model!!”
> I like keeping their memory separate.
When casually a friend asks you to ask your Claude something about topic you chatted about earlier and then Claude brings back a secret you didn't want anyone to know yet.
Yes you can set up a project and then ask a question, but this is tedious.
I'm starting to explore alternative options because Claude has become an awful value proposition. Any suggestions?
I turned off Anthropic properly and switched all of that over to OpenAI yesterday. (I use other models for other things too, especially DeepSeek in Pi).
Honestly aside from the voice it uses you wouldn't notice a difference. Switching costs are low, vote with your wallet.
> I turned off Anthropic properly and switched all of that over to OpenAI yesterday.
Same, just a while longer ago.
I much prefer how OpenAI models write to Anthropic's, will probably revisit Anthropic in a generation or two. Context size is more limited, but no critical forgetfulness due to compaction so far, though I also like having plan files around both for future reference and improving chances of success at long form work.
Also tried out Kimi K3, was nice but slow (and apparently routed some requests to Claude anyways), GLM 5.3 was faster and still pretty good but the token allowances were kinda limited.
OpenAI is a better value proposition too because they give you effectively infinite image generation + chat usage, which is separate from work/codex.
So you voted for Kodos instead of Kang [1]. Local models are the way out of the rug pulling
[1] https://m.youtube.com/watch?v=BUAnyVAanac&ra=m
I don't actually disagree, and advocate that for now the thing to do is use both.
You need to use frontier models to understand where the puck is going, but also to use local ones for anything remotely sensitive.
I would like to have alternative to chat - model and provider agnostic (BYOK), but with history, project, maybe memory, with good search and quality tools. I know openwebui and i dont want to host it.
codex (their app) is pretty good and lots of banked resets, luna is cost effective and Astra seems better than Fable for many tasks. Most important one is less refusals and I can use it the way I want without the fear of getting banned. I do like Claude Code a lot when it comes to pure coding use cases but the work often touches outside code and I don't want to keep switching
> I do like Claude Code a lot when it comes to pure coding use cases
The harness is great, but opus is such an arrogant little prick that spews out unintelligible word salad. Opus 5 is so bad at communication it amazes me that somebody green-lit it. It's absolutely awful.
The fact that this isn't an acknowledged regression (and Fable 5.1 isn't much better) leads me to believe that people at Anthropic actually like Opus 5's output.
Of the major providers, Codex/Astra. By far the highest quality model, especially for cross-discipline work.
Opencode Go.
[dead]
Basically DOA in healthcare, even with a BAA I don't want Claude Cowork enabled in our workstations, ever.
why tho? genuinely curious
Good to see AI providers finally starts understanding how bad was the user experience. This is not just a Claude issue but same issue exists in ChatGPT. While it is in the right direction, there is more to be done. I personally think that they should separate Chat and Coding into two separate apps. Both are completely different use-cases. As a developer, I have need for both but I don't like them in the same app. For example, I consider my chats as knowledge base like Obsidian that I want to keep it organized and browse in the future. On the other hand, once I done a coding task, I don't really need that chat anymore. Having coding tasks in the app feel like polluting my regular chats. This frustration led me build my own Chat-only App https://www.vinaa.ai/ where I keep my regular chats in one place organised into projects and folders incl. the ability to auto-organise chats. App is still in early days and I am experimenting a lot to make it a real AI knowledgebase.
The problem is that knowledge workers just need the simplest way possible to automate their work. Currently I feel like Anthro is overcomplicating this. If I'm an accountant, I want to open Claude App, describe my workflow for balancing some books, then have Claude design a reusable WORKFLOW that it runs whenever I ask. The workflow only needs 2 primitive skills to do anything: /work, and /review. Compose these together with guidance docs into workflows, then workflows and call other workflows. It's all just callable skills.
/balance-books hey claude here's the books. Go balance them
Claude then uses the guidance docs and various subagents to review the books, send off parallel workers with cheap models, then review with a more expensive model. Done in the repeatable, correct order with independent review every time. As new models come out, your workflow structure stays the same. You just delete some prescription from the guidance docs.
https://github.com/DanMcInerney/orchflows
Really excited to try the gsuite replacements, which seems like the bigger announcement than the title suggest
Why? I only use gsuite for interop between Claude and ChatGPT and other services...so a moat like that is totally worthless to me
I still use Google Workspace for personal and work stuff, but for solo agent workflows, I think I have some use cases that would help me personally. I can let agents spin up working documents that I can review, comment on, and iterate over asynchronously.
For example, I’ve spent the last few weeks building an internal ticketing setup for my agents so they stop trying to fix every tangential issue they encounter. I wrote a skill where subagents bubble out-of-scope issues up to a parent orchestrator, which decides whether to log a ticket.
Currently, that setup mirrors between a local SQLite database and an Obsidian vault: agents write to SQLite, which syncs to Obsidian; when I edit notes in Obsidian, it syncs back. Native Claude docs mean I could ditch the Obsidian middleman for remote review—editing tickets from anywhere while keeping the core SQLite state local to where the agents execute.
That makes sense since you treat it as your central productivity manager, but for me all of Claude/Codex/etc are not productivity tools, just heterogeneous compute /for/ the productivity tool :P
1) If you want to make Claude usable how about addressing the fact that OpenAI 20x lasts all day on Astra, yet Claude 20x pips out in 20 mins of agentic with Fable 5.1, which is totally useless.
And unlike Sol: Opus like yoda, anything for unusable he is.
2) The first AI company to reduce their token pricing such that Astra/Fable can run for less than $150/day from credits would see a 5x usage increase and 20x spend increase from me. It’s currently around $500 for Astra and $5000 for Claude.
3) They seem to recently have addressed the fact that their safeguards were OTT so kudos to them for realising and now Fable is usable for wire and memory level programming.
Just yesterday I experienced this product weirdness -- oh you're in a chat in this project and I can't do that, you need to open a task.
The product management at Anthropic is pretty... nonexistent?
I love this
Claude's mobile app wouldn't show any Cowork initiated conversations, even if those conversations didn't leverage cowork specific features
I hope this doesn’t get confusing like ChatGPT made it.
I feel like Claude has some of the best UX, and hope this doesn’t dilute the experience.
Claude has a jagged UX if you support non-devs. The surfaces before today in the desktop app were: Chat, cowork, Claude code.
Inside Claude code you’ve got cloud environments and local.
Let’s say as a normal desktop user you wanted to just automate something normal in your work, going to a website (maybe some internal app at your company) getting some data. Depending on what surface you used this will either not work, not work well (cowork), or work quite well but possibly be blocked because of bot controls (local Claude code) or again not work (cloud Claude code)
It was nuts. You can clearly tell these were different teams and products mashed into the same app. Super confusing for non technical users. Hopefully it’s a little better today
Sounds like they need a "Claude for Normal People" and a "Claude for people who know how computers work" lol
I would argue they need 1 product owner for the app surface so I can’t see their org structure so clearly. Those jagged edges and users hitting blocks depending on what section of the app they were in are design failures, I know how a computer works and when using the Claude app it wasn’t always clear to me what the right combo of things to press to get my desired outcomes.
Yeah I don't disagree, ChatGPT/Claude both could really stand a UI/UX pass (and a perf pass for Firefox's sake lol). I was more expressing the perennial fear of power user features being jettisoned in favor of the lowest common denominator.
Do you think ChatGPT UX is better (after they merged everything recently)?
Genuinely curious.
For me it’s still jagged, but chatgpt never had the weird “cowork” abstraction with the Linux machine.
I don’t talk to non-technical users that use the chatGPT app for work so I’m not sure if that’s been a positive change for them, but I would imagine if that was put in place for Claude it would be better for them. Chat is instant answers, different model, faster; work is more detailed, thinks more, has access to all the tools. This is what I see the users doing, but there is a third mode they use (deep research, exhaustively find me the right answer to a crucial question) that is underserved right now.
All of this is moot to me as a dev, I use terminal, herdr, pi… so I don’t see the jagged edges myself. But regular users do
I really appreciated that the users that I administrate had an option where Claude was not going to get carried away and orchestrate workslop, I will have to completely re-write my usage guides after testing how prompts are routed now.
Feels like the same "let us do _all_ of the thinking for you" messaging that Microsoft has with Copilot.
It's funny how they keep trying to add more interfaces and everything keeps coming back to chat. And of course it would, natural language is the simplest interface there is. Anything they add beyond chat, that isn't placed within chat, that isn't some artifact of chat, is friction.
I really hope ChatGPT doesn't follow suit. Chat and Work serve 2 distinct tasks, and unifying Chat and Codex limits will cause the users to avoid using Chat in order to save the tokens for Codex.
That's the natural outcome. AI will eventually collapse into a single, universal entry point, much like Google Search became the default gateway for the web.
Why would anyone install software that the owner himself said is unsafe to use?
I found out the claude compro crestes a large vm in my fisk which i did not want.
They could have figured that out before launching Cowork to everyone.
[dead]
not once did I ever understand what the difference was between Claude Chat and Claude Cowork. I use both and I mostly just guessed when to use Cowork.
I guess this is a feature worth announcing
One Claude one damn.
Less is more. Thanks.
[flagged]
"Don't be snarky."
"Please don't post shallow dismissals, especially of other people's work. A good critical comment teaches us something."
https://news.ycombinator.com/newsguidelines.html
My point with above is to criticize their strategy, I have been losing faith on Anthropic and feeling like they are throwing everything at the wall to see what sticks in a more desperate fashion rather than careful one. This sort of decision to me, alongside all their other initiatives recently with shipping agentic PaaS rather than enabling others to build with their tools has been a miss on thoughtful leadership. Sorry for the snark take, thank you for moderating.
Thanks and yes, there's often a sensible HN comment behind posts like the one I was replying to.
[dead]
[flagged]
[flagged]
A change for the better. The distinction made it unnecessarily complex for users. The ChatGPT desktop app inherited the same separation. Having to change between ChatGPT for day to day questions, to Codex for everything else always seemed too cumbersome.