Posted by brdd 1 day ago
- Why do you need price trackers for airbnb? It is not a superliquid market with daily price swings.
- Cataloguing your fridge requires taking pictures of everything you add and remove which seems... tedious. Just remember what you have?
- Can you not prepare for the next day by opening your calendar?
- If you have reminders for everything (responding to texts, buying gloves, whatever else is not important to you), don't you just push the problem of notification overload to reminder overload? Maybe you can get clawdbot to remind you to check your reminders. Better yet, summarize them.
I agree that removing items and taking pictures takes more effort than it saves, but I would use a simpler solution if one existed because it turns out I cannot remember what we have. When my partner goes to the store I get periodic text messages from them asking how much X we have and to check I look in the fridge or pantry in the kitchen and then go downstairs to the fridge or pantry in the basement.
> Can you not prepare for the next day by opening your calendar?
In the morning I typically check my work calendar, my personal calendar, the shared family calendar, and the kids' various school calendars. It would be convenient to have these aggregated. (Copying events or sending new events to all of the calendars works well until I forget and one slips through the cracks...)
> If you have reminders for everything (responding to texts, buying gloves, whatever else is not important to you), don't you just push the problem of notification overload to reminder overload?
Yes, this is the problem I have. This doesn't look like a suitable solution for me, but I understand the need.
Am I missing this in the article? Do you mean the shoes he's holding? He explains it immediately.
>when i visited REI this weekend to find running shoes for my partner, i took a picture of the shoe and sent it to clawdbot to remind myself to buy them later in a different color not available in store. the todo item clawdbot created was exceptionally detailed—pulling out the brand, model, and size—and even adding the product listing URL it found on the REI website.
- Why do you need a reminder to buy gloves when you are holding them?
I think AI is about to do the same thing to pair programming that full self-driving has done for driving. It will be a long time before it's perfect but it's already useful. I also think someone is going to make a Blockbuster quality movie with AI within a couple years and there will be much fretting of the brows rather than seeing the opportunity to improve the tooling here.
But I'll make a more precise prediction for 2026. Through continual learning and other tricks that emerge throughout the year, LLMs will become more personalized with longer memories, continuing to make them even more of a killer consumer product than they already are. I just see too many people conversing with them right now to believe otherwise.
These people have taken over the industry in the past 10 years.
They don't care anything about the tech or product quality. They talk smooth, loud, and fast so the leaders overlook their incompetence while creating a burden for the rest of the team.
I had a spectacular burnout a few years ago because of these brogrammers and now I have to compete with them in what feels like a red queen's race where social skills are becoming far more important than technical skills to land a job.
I'm tired.
>kill dragon
With what? Your bare hands?
>yes
Congratulations! You have just vanquished a dragon with your bare
hands! (Unbelievable, isn't it?)Approximately nothing?
Yeah, the sane solution here is much simpler. Put a magnet whiteboard. When you put something into the fridge, add it to the whiteboard. When you take something out, you erase that item from the whiteboard.
Note that the tendency to feel overwhelmed is rather widespread, particularly among those who need to believe that what they do is of great import, even when it isn't.
It's the equivalent of me having to press a button on the steering wheel of my Tesla and say "Open Glovebox" and wait 1-2 seconds for the glove box to open (the wonders of technology!) instead of just reaching over and pressing a button to open the glovebox instantly (a button that Tesla removed because "voice-operated controls are cool!"). Or worse, when my wife wants to open the glovebox and I'm driving she has to ask me to press the button, say the voice activated command (which doesn't work well with her voice) and then it opens. Needless to say, we never use the glovebox.
People saying 'Claude is now managing my life!11' are like gearheads messing with their carburetor or (closer to this analogy) people who live out of Evernote or Roam
All that said I've been thinking for a while that tool use and discrete data storage like documents/lists etc will unlock a lot of potential in AI over just having a chatbot manipulating tokens limited to a particular context window. But personal productivity is just one slice of such use cases
More importantly, can Clawdbot even reliably access these sites? The last time I tried to build a hotel price scraper, the scraping was easy. Getting the page to load (and get around bot detection) was hard.
This is one of the stupidest things I have read on this site
Billions of people don't use todo list apps so they're useless; just remember what to do.
Billions of people don't use post-its apps so they're useless; just remember what you're going to write down.
Billions of people don't have cars; just walk.
You can dismiss any invention since industrial revolution with this logic.
You can justify the value of any ridiculous invention by comparing it to a world-changing invention.
Common internet tropes include both "look at this forgotten jar that's been in the back of my fridge since 1987" and "doesn't it suck how much food we waste in the modern world?"
Nearly every modern invention could be dismissed with this attitude. "Why do you need a typewriter? Just write on paper like the rest of the world does."
"Why do you need a notebook? Just remember everything like the rest of us do."
But suggesting "Why not just try remembering lol" isn't really a valid criticism of the process. What you said here is a real criticism that actually adds to the conversation.
But this is already built-in with gmail/gcalendar. Clawdbot does take it one step further by scraping his texts and WhatsApp messages. Hmmm... I would just configure whatever is sending notifications to send to gmail so I don't need Clawdbot.
One of the differences in risk here would be that I think you got some legal protection if your human assistant misuse it, or it gets stolen. But, with the OpenClaw bot, I am unsure if any insurance or bank will side with you if the bot drained your account.
These disincentives are built upon the fact that humans have physical necessities they need to cover for survival, and they enjoy having those well fulfilled and not worrying about them. Humans also very much like to be free, dislike pain, and want to have a good reputation with the people around them.
It is exceedingly hard to pose similar threats to a being that doesn’t care about any of that.
Although, to be fair, we also have other soft but strong means to make it unlikely that an AI will behave badly in practice. These methods are fragile but are getting better quickly.
In either case it is really hard to eliminate the possibility of harm, but you can make it unlikely and predictable enough to establish trust.
In fact, if I wanted to implement a large-scale identity theft operation targeting rich people, I would set up an 'offshore' personal-assistant-as-a-service company. I would then use a tool like OpenClaw to do the actual work, while pretending to be a human, meanwhile harvesting personal information at scale.
And OpenClaw could probably help :)
> an electronic fund transfer from a consumer's account initiated by a person other than the consumer without actual authority to initiate the transfer and from which the consumer receives no benefit
OpenClaw is not legally a person, it's a program. A program which is being operated by the consumer or a person authorized by said consumer to act on their behalf. Further, any access to funds it has would have to be granted by the consumer (or a human agent thereof). Therefore, baring something like a prompt injection attack, it doesn't seem that transfers initiated by OpenClaw would be considered unauthorized.
[0]: https://www.consumerfinance.gov/rules-policy/regulations/100...
Additionally:
- As has been pointed out elsewhere in the thread, it can be difficult to separate out "prompt injection" from "marketing" in some cases.
- Depending on what the vector for the prompt injection is, what model your OpenClaw instance uses, etc. it might not be easy or even possible to determine whether a given transfer was the result of prompt injection or just the bot making a stupid mistake. If the burden of proof is on the consumer to prove that it as prompt injection, this would leave many victims with no way to recover their funds. On the other hand, if banks are required to assume prompt injection unless there's evidence against it, I strongly suspect banks would respond by just banning the use of OpenClaw and similar software with their systems as part of their agreements with their customers. They might well end up doing that regardless.
- Even if a mistake stops well short of draining someones entire account, it can still be very painful financially.
chef's kiss
In the plugin docs is a config UI builder. Plugin is OSS, boards aren’t.
An additional benefit of isolating the account is it would help to limit damage if it gets frozen and cancelled. There's a non-zero chance your bot-controlled account gets flagged for "unusual activity".
I can appreciate there's also very high risk in giving your bot access to services like email, but I can at least see the high upside to thrillseeking Claw users. Creating a separate, dedicated, mail account would ruin many automation use cases. It matters when a contact receives an email from an account they've never seen before. In contrast, Amazon will happily accept money from a new bank account as long as it can go through the verification process. Bank accounts are basically fungible commodities, can easily be switched as long as you have a mechanism to keep working capital available.
you end up on the fraudster list and it will follow you for the rest of your life
(CIFAS in the UK)
and then if you tell them it's not you doing the transactions: you will be immediately banned
"oh it's my agent" will not go down well
Also, at best, you can only add to the system prompt to require confirmation for every purchase. This leaves the door wide open for prompt injection attacks that are everywhere and cannot be complete defended against. The only option is to update the system prompt based on the latest injection techniques. I go back to the case where known, supposedly solved, injection techniques were re-opened by just posing the same attack as a poem.
The courts have an answer for this one: intent. How do courts know if your intent meets the definition of fraud or theft or whatever crime is relevant? They throw a bunch of evidence in front of a jury and ask them.
From the point of view of a marketer, that means you need be well behaved enough that it is crystal clear to any prosecutor that you are not trying to scam someone, or you risk prosecution and possible conviction. (Of course, many people choose to take that risk).
From the point of view of a victim, it's somewhat reassuring to know that it's a crime to get ripped off, but in practice law enforcement catches few criminals and even if they do restitution isn't guaranteed and can take a long time. You need actual security in your tools, not to rely on the law.
It probably also violates local laws (including simple theft in my jurisdiction).
I just don't see a reason to allow OpenClaw to make purchases for you, it doesn't feel like something that a LLM should have access to. What happens if you accidentally end up adding a new compromised skill?
Or it purchases you running shoes, but due to a prompt injection sends it through a fake website?
Everything else can be limited, but the buying process is currently quite streamlined, doesn't take me more than 2 minutes to go through a shopify checkout.
Are you really buying things so frequently that taking the risk to have a bot purchase things for you is worth it?
I think that's what turns this post from a sane bullish case to an incredibly risky sentiment.
I'd probably use openclaw in some of the ways you're doing, safe read-only message writing, compiling notes etc & looking at grocery shopping, but i'd personally add more strict limits if I were you.
But don't you want the agents to book vacations and do the shopping for you!!?!
Though it would be nice if "deep research" could do the hard work of separating signal from the noise in terms of finding good quality products. But unfortunately that requires being extremely skeptical of everything written on the web and actively trying to suss out the ownership and supply chain involved, which isn't something agents can do unguided at the moment.
I've noticed this too, and I think it's a good thing: much better to start using the simplest forms and understand AI from first principles rather than purchase the most complete package possible without understanding what is going on. The cranky ones on HN are loud, but many of the smart-but-careful ones end up going on to be the best power users.
I feel lucky to have experienced early Facebook and Twitter. My friends and I figured out how to avoid stupidity when the stakes were low. Oversharing, getting "hacked", recognizing engagement-bait. And we saw the potential back when the goal was social networking, not making money. Our parents were late. Lambs for the slaughter by the time the technology got so popular and the algorithms got so good and users were conditioned to accept all the ads and privacy invasiveness as table stakes.
I think AI is similar. Lower the stakes, then make mistakes faster than everyone else so you learn quickly.
Another thing about early users is they are also longer-term users (assuming they are still on the platform) and have seen the platform evolve, which gives them a richer understanding of how everything fits together and what role certain features are meant to serve.
I was initially overly optimistic about AI and embraced it fully. I tried using it on multiple projects - and while the initial results were impressive, I quickly burned my fingers as I got it more and more integrated with my workflow. I tried all the things, last year. This year, I'm being a lot more conservative about it.
Now .. I don't pay for it - I only use the bare bones versions that are available, and if I have to install something, I decline. Web-only ... for now.
I simply don't trust it well enough, and I already have a disdain for remotely-operated software - so until it gets really, really reliable, predictable and .. just downright good .. I will continue to use it merely as an advanced search engine.
This might be myopic, but I've been burned too many times and my projects suffered as a result of over-zealous use of AI.
It sure is fun watching what other folks are daring to accomplish with it, though ..
Although that feels a bit exaggerated, I feel it's not far from the truth. If there were, say, 3 closed source animation software that could do professional animation in total, and they just all decided to just kill the product one day, it would actually kill the entire industry. Animators would have no software to actually create animation with. They would have to wait until someone makes one, which would take years for feature parity, and why would anyone make one when the existing software thought such product wasn't a good idea to begin with?
I feel this isn't much different with AI. It's a rush to make people depend on a software that literally can't run on a personal computer. Adobe probably loves it because the user can't pirate the AI. If people forget how to use image editing software and start depending entirely on AI to do the job, that means they will forever be slaves to developers who can host and setup the AI on the cloud.
Imagine if people forgot how to format a document in Word and they depended on Copilot to do this.
Imagine if people forgot how to code.
This is not about big increases of productivity, this is whole thing about selling dependence on privately controlled, closed source tools. To concentrate even more power in the hands of a very few, morally questionable people.
- Declare victory the moment their initial testing works
- Didn’t do the time intensive work of verifying things work
- Author will personally benefit from AI living up to the hype they’re writing about
In a lot of the authors examples (especially with booking), a single failure would be extremely painful. I’d still want to pay knowing this is not likely to happen, and if it does, I’ll be compensated accordingly.
> Tech people are always talking about dinner reservations . . . We're worried about the price of lunch, meanwhile tech people are building things that tell you the price of lunch. This is why real problems don't get solved.
What's puzzling to me is that there's little consideration of what one is trading away for this purported "value". Doing menial tasks is a respite for your brain to process things in the background. Its an opportunity to generate new thoughts. It reminds you of your own agency in life. It allows you to recognise small patterns and relate to other people.
I don't want AI to summarise chats. It robs me the opportunity to know about something from someone's own words, therefore giving a small glimpse in their personality. This paints a picture over time, adding (or not) to the desire to interact with that person in the future. If I'm not going to see a chat anyway, then that creates the possibility of me finding something new in the future. A small moment of wonder for me and satisfaction for the person who brought me that new information.
etc etc.
Its like they're trying to outsource living.
Maybe the story is that, outsourcing this will free them up to do more meaningful things. I've yet to see any evidence of this. What are these people even talking about on the coffee chats scheduled by the helpful assistant?
https://www.youtube.com/watch?v=eBSLUbpJvwA
"Do tape recorders ring a bell?"
There are so many things I don't want to do. I don't want to read the internet and social media anymore - I'd rather just have a digest of high signal with a little bit of serendipity.
Instead of bookmarking a fun physics concept to come back to later, I could have an agent find more and build a nice reading list for me.
It's kind of how I think of self-driving cars. When I can buy a car with Waymo (or whatever), jump in overnight with the wife and the dogs, and wake up on the beach to breakfast, it will have arrived in a big way. I'll work remotely, traveling around the US. Visit the Grand Canyon, take a work call, then off to Sedona. No driving, traffic, just work or leisure the whole time.
True AI agents will be like this and even better.
Ads, for sure, are fucked. If my pane of glass comes with a baked in model for content scrubbing, all sorts of shit gets wiped immediately: ads, rage bait, engagement bait, low effort content.
AdBlock was child's play. We're going to have kernel-level condoms for every pixel on screen. Thinking agents and fast models that vaporize anything we don't like.
The only thing that matters is that we have thin clients we control. And I think we stand a chance of that.
The ads model worked because of disproportionate distribution, platform power, and verticalization. Nobody could build competing infra to deal with it. That won't be the case in the future.
How does Facebook know the person calling their API is human? How do they know the feed being scrolled is flesh fingers?
Everything will filter though a final layer of fast and performance "filter" models.
Social media algorithms will be replaced by personal recommender agents acting as content butlers.
We just need a good pane of glass to house this.
Yeah this sounds totally sane!
i don't think we need ClawdBot, but we do need a way to easily interact with the model such that it can create long term memories (likely as files).