I partly identify with the article. While I don't use OpenClaw itself, I hacked together my own small Claude-in-a-loop/cronjob, and it seems we're all getting our morning briefings and personalized morning podcasts now.
The other common use case seems to be kicking off an automated Claude session from an email / voicetext / text / Telegram, and getting replies back. I'm emailing Claude throughout the day now, and sometimes it's useful to just forward an email to Claude and ask it to handle the task within it for me.
But I think many people criticizing the various Claws are missing out on the cronjob aspect. There's value in having your AI do work automatically while you're asleep. You don't even need OpenClaw for that, just a cronjob that runs claude -p in the early morning. If you give your AI enough context about yourself, you get to a point where it just independently works on things for you, and comes to you with suggestions. It doesn't need to be specifically prompted. The environment of data it can access is its own context, its own prompt. With that, it can sometimes be surprising and spooky what you wake up to, without being directly prompted.
Give it enough context, long term memory, and ability to explore all of that, and useful stuff emerges.
Just a heads up if you're using claude -p with the sub your account might be in trouble, they're cracking down on this kind of stuff lately.
Enforcement seems to be a combination of string matching for 3rd party sysprompts, heavy usage, and some random factor.
Not clear if there are any hard rules you can stay on the good side of, the only way to be safe seems to be to pay per token. (There goes the ~90% discount!)
--
Also yeah you get ~80% of Claw by shoving Claude Code in a Telegram Bot ;) It's already a general purpose computer use thing, people forget! (And it's a lot better at extending itself than the actual claws, lol)
I think the least illegal and also least bad option is to just use ngrok and tmux tho
> I could tell it (via Telegram) to add something to my TODO list at home while I'm in the office. It would call a custom API I had set up that adds items to my TODO list. How can I replicate this without the hassle of setting up OpenClaw?
The general idea is make a simple deterministic program that runs on your PC at home in a never ending loop. Every minute or so, check Telegram for a new message. If a message is received, then the program runs "claude -p" with a prompt, whatever MCP tools or CLI permissions it needs, and the contents of your Telegram message. Just leave the program running on your home computer while you're out, and you're done.
I don't use Telegram, so coding the part to check Telegram would be the hard part. I use email instead, and have the program check every minute for new mail (I leave my email program running and check the local inbox file). I'd already coded up a local MCP server to manage my ToDo list (Toodledo) so Claude just calls the MCP tools to add the task.
For things like TODO, I guess an email will suffice.
However, it was really nice being able to use Telegram and get quick validation. I also had a flow set up where I could send a voice memo. It would take the audio file (ogg), run Whisper, and then pass through an LLM for cleanup, and follow the instructions in my message. Really handy to use while I'm walking around.
I guess I want to create my own OpenClaw like agent, but not with its crazy broad access: Just limited to the functionality I allow, and with the convenience of using Telegram. I don't care about memory, soul, etc.
I did the exact same voice memo thing too, except I had Claude make an Android app to record the file and send it to Whisper. In the end I had the app just email the transcription & trigger Claude that way (ie receiving the email triggers my PC to wake up Claude), rather than sending Claude the audio file directly.
My reverse audio reply loop is convoluted - I have Claude generate its TTS file from Whisper/Mistral, and upload them to a server with an RSS file it updates, so I can play them in my podcast app (AntennaPod), then send me a notification via Pushover that the reply is waiting. I ended up building out an MCP tool for that workflow, so Claude really just calls the MCP tool with the text of what it wants to say, everything else is a deterministic program doing the work.
Memory is really useful to have - it can just be a bucket of searchable Markdown files. It's also useful to have a "reminders to self" Markdown file that Claude reads each time, and that Claude can update. I don't continue the same context window, and that "reminders to self" plus the ability to read previous emails in the conversation seems to be enough to keep the context going for me.
You'll feel better if you know exactly how your Claw is locked down. Mine doesn't have the open email access others are granting, not at all. Claude gets a bit grumpy about that and keeps begging for more access :)
I felt pretty clever until (1) I found a repo where they used this trick to create a full OpenAI compatible API endpoint[0] (lmao, the VC money distortion field spawning truly comical Rube Goldberg machines), and (2) they started banning "unauthorized" usage of the Claude sub, which trend unfortunately seems to be accelerating recently as their lower value consumers have grown in both number and usage.
I think shoving claude -p in your bash script / cronjob / messaging app bot of choice counts as "unathorized 3rd party harness", but your guess is as good as mine...
(claude -p with per-token billing (i.e. paying 7x more) is allowed though, of course)
-- There's also an Agents SDK (formerly Claude Code SDK?) which is basically just claude -p but with more typing, as far as I could tell.
Qwant is working on that. Together with Ecosia they're building their own index called the European Search Perspective:
"Today, Europe receives 99% of the answers to search queries from external infrastructures. We believe, however, that a higher level of digital sovereignty is essential for a functioning democracy and economy. With our new web index, we are creating a European perspective on politics, culture and values. This is a long overdue step towards more plurality in the digital world, which is also being called for by our society."
> a European perspective on politics, culture and values
To be honest this does not sound much better. 40 years ago maybe I would have preferred EU values over the US' puritan values. Nowadays I'd just expect a different flavor of poison.
It's up to the company, but since many companies don't want to keep card numbers around (and some processors don't let you see the card number anyway), they're probably more likely to block on identity. Maybe flag the IP address of the transaction for "additional screening" on all future transactions, etc.
IPs are notoriously unreliable for identity pinning, particularly in this age of CGNAT.
If they can’t or don’t want cc numbers (makes sense considering how painful PCI guidelines are anyway) does that mean they need to rely on more tools from the processors or user accounts maintained by the merchant themselves?
CC numbers are also bound to get recycled eventually as cards expire and/or get replaced... even if you block a card, it might have a new owner 6 months or so later.
The number space between the first 6 digits (BIN) and the Luhn check digit is 9 digits — that's 1 billion numbers that issuers can give out before a collision happens.
That doesn't seem to be more than an order of magnitude off between available numbers and issued cards - a cursory search says there are over a billion credit cards in circulation in the US alone.
I think you're confusing the available number space per BIN (often used for a single card product) with the number of available numbers per network.
Visa and Mastercard each have 14 digits worth of permutations to play with, excluding the first and last digits. That's one hundred trillion numbers.
Assuming 8 billion people in the world, each person can hold 12,500 of either Visa or Mastercard before a collision happens. (As above, the number space is smaller because of how BINs are issued, but that's still plenty.)
Not that anyone is going to listen to Australia, but Australia believes it is part of the ceasefire terms that were agreed to:
Asked on 7.30 if the ceasefire should apply to Israel's action in Lebanon, [Australian Foreign Minister] Senator Wong was adamant it should. "Yes," she said. "Our position is that the world expects that the ceasefire should apply to the region."
"Our position is that it should" is very different from "the text of their statement says". This is Senator Wong's (or Australia's) idea of what would happen in an ideal world, not anything anybody involved recognizes as relevant. (I mean, they may not recognize the text as being relevant, either, but this is a step below even that.)
Genuine question - if you don't think the models are improved or that the code is any good, why do you still have a subscription?
You must see some value, or are you in a situation where you're required to test / use it, eg to report on it or required by employer?
(I would disagree about the code, the benefits seem obvious to me. But I'm still curious why others would disagree, especially after actively using them for years.)
The assumption that the other person made was that I would only use it for coding. If you look through my other comments today, I suggest that they are useful for performing repetitive tasks i.e. checking lint on PR, etc. Also, can be used for throwaway code, very useful.
I don't think the issue is with the model, it is with the implication that AGI is just around the corner and that is what is required for AI to be useful...which is not accurate. The more grey area is with agentic coding but my opinion (one that I didn't always hold) is that these workflows are a complete waste of time. The problem is: if all this is true then how does the CTO justify spending $1m/month on Anthropic (I work somewhere where this has happened, OpenAI got the earlier contract then Cursor Teams was added, now they are adding Anthropic...within 72 hours of the rollout, it was pulled back from non-engineering teams). I think companies will ask why they need to pay Anthropic to do a job they were doing without Anthropic six months ago.
Also, the code is bad. This is something that is non-obvious to 95% of people who talk about AI online because they don't work in a team environment or manage legacy applications. If I interview somewhere and they are using agentic workflow, the codebase will be shit and the company will be unable to deliver. At most companies, the average developer is an idiot, giving them AI is like giving a monkey an AK-47 (I also say this as someone of middling competence, I have been the monkey with AK many times). You increase the ability to produce output without improving the ability to produce good output. That is the reality of coding in most jobs.
AI isn't good enough to replace a competent human, it is fast enough to make an incompetent human dangerous.
That sounds intriguing - are there examples of desktop apps (eg productivity apps) that are made with JUCE?
I'm already very familiar with JUCE in VST plugins, I have hundreds of VSTs made with JUCE and I love their UIs. But I don't think I've ever seen it for a standalone business application?
The JUCE pricing is certainly far more reasonable than Qt.
You can configure your JUCE project to build to any of the standard audio plugin formats, or for it to build a standalone app for the target platform (i.e. plugins, desktop, mobile, embedded) or indeed to build all of these targets, at once, in one build.
However, during app (i.e. non-plugin) development I often switch between linux and macos environments as part of my workflow, and during testing after I've pushed to the proper branch, the build server plops out the .exe/.pkg/.app/.tar.gz bundles as needed for the test group to crack at it.
JUCE, being at heart a C++ framework intended to be the engine of a very diverse swath of different OS, plugin, and packaging standards, does all the glue to get you there - how you use it, is up to you. (All of this can happen in github actions, btw, really easy to set up..)
Yes, there are 'business'/'productivity' UI elements in JUCE app targets, and yes they are consistent across all platforms. And yes, you can for example build a UI from an .xml form, with cross-platform datastore and so on, easily enough.
Here's a nice place to start, if you wanna understand JUCE capabilities from a 'productivity'-app perspective:
David Rowland, a core JUCE developer, explaining the guts of things.
Basically, the scope is high performance applications, and there are no really good reasons for why a high performance application cannot also be a productivity app - the distinction is arbitrary - except of course you ask, for the GUI!
But: JUCE' GUI is pretty darn good, I have to say, in face of the onslaught of platform vendor fuckery. You can embed a WebView if you really need it, and wire it up to the rest of the cross-platform event handing system, etc. But I think its not really needed, given the plethora of 'normal' UI controls, out of the box.
If you're serious about looking at JUCE for a variety of application types, then a lot of the questions you're going to have about JUCE' GUI suitability for standalone business applications can be answered by running the DemoRunner application that's built-in to the JUCE codebase.
Clone the repo, build the DemoRunner project for your platform (or all of them), and you'll see - there are plenty of business-like cross-platform UI elements in the kitty. A huge collection of business-/productivity- like UI elements, right out of the box. (High-performance plugin UI's are there with the business UI stuff, too.)
And .. once you've marvelled at the glory of DemoRunner(.exe,.app,.apk)[etc.] .. please do yourself a favour and spend an extra hour parsing the awesome-juce list:
You might have sold me on this, because on top of the productivity-ish stuff I do, I'd really love to get into VST development. If I hadn't fallen into Photoshop plugins first, I would always have been an audio plugin dev. I loved PSP VintageWarmer in the late 90s.
The UI in Rowland's demo at 32min is feeling a little DearImGui / Reaper-ish, kinda bland while also being non-native. But I've seen all the UIs that Arturia & everyone else makes with JUCE, even what Valhalla does with it with their NASA inspired interfaces, so it can clearly make amazing stuff.
I can't believe we're now looking to PACE of all companies to help indie devs. I hated iLok for so long even though I have two iLok dongles now ;)
Well, I hope you will take a serious look at it as a technology stack worth learning, because when you get things rigged up so that you can really just push your changes into a repo and get the builds for all platforms, its some kind of magic, and you will - most likely - be really inspired to create something great.
Which, I hope, you do.
(PS; - the default UI skin is, imho, intentionally a tad bland in order to promote developer uptake of JUCE' pretty amazing skinning/UIBehaviour system...)
I think I prefer that JUCE is a one-off licence at $800 (for small business) and that they offer a 30% discount if you want to upgrade when they release a new major framework version. I'd rather not get locked into Qt's annual $618/subscription every year just to continue distribution & development.
But Qt has the bonus of also running on HaikuOS, and I've definitely seen more business apps in Qt (pretty sure 010 Editor is Qt).
> If your customer is legitimately having a money issues or something else, and they are actually a customer, they will contact you and attempt to work something out.
Depending on customer personality, they may or may not contact you. It's also possible that if they're having money issues that they are freaking out, frozen, and scared to reach out. Though I'm thinking more of very small businesses here where it's more of a personal thing than a corporate thing.
On your payment reminder notes, when it really is late, you can consider putting in some wording that hints at "we know sometimes small businesses can have financial issues" and genuinely suggest there are ways you can help. If it's a small friendly business in a scary time, they might be genuinely relieved that you show that you care... while simultaneously, if it's a business that isn't in financial trouble at all, and was just trying to stretch out their Net 30 to earn extra interest, they may be so outraged by the implication that they are "poor" they they will pay up quickly just to show they are not poor.
It won't help every case, and this only happened very very rarely to me (so I'd defer to the judgment of others). But depending on your customers, this could be a useful approach to have in your toolkit.
The other common use case seems to be kicking off an automated Claude session from an email / voicetext / text / Telegram, and getting replies back. I'm emailing Claude throughout the day now, and sometimes it's useful to just forward an email to Claude and ask it to handle the task within it for me.
But I think many people criticizing the various Claws are missing out on the cronjob aspect. There's value in having your AI do work automatically while you're asleep. You don't even need OpenClaw for that, just a cronjob that runs claude -p in the early morning. If you give your AI enough context about yourself, you get to a point where it just independently works on things for you, and comes to you with suggestions. It doesn't need to be specifically prompted. The environment of data it can access is its own context, its own prompt. With that, it can sometimes be surprising and spooky what you wake up to, without being directly prompted.
Give it enough context, long term memory, and ability to explore all of that, and useful stuff emerges.
reply