Basically: If you run OpenClaw connected to any meaningful system you are not fit to design, program or run any kind of software. That disregard for security and quality should leave a black mark on you for many years.
@tante Question: what if you name it?
@art_codesmith You should be forced into getting therapy
@tante So um... funny thing I read on LinkedIn today... #DHH

@art_codesmith @tante lol he’s so proud of his agent, and then his story amounts to “I told the AI to do something. It couldn’t and didn’t have any idea how to proceed. So I told it to do a different thing first. And it did that, all on its own, no prompting from me! Then it did the first thing I told it to do. With no direction at all!”

Like awww it’s almost sweet that you don’t know what prompting and direction are, David! Truly groundbreaking.

@quietewe

Hi Erin! Laughing at a screenshot of David’s 'aha!' moment feels cheap. We’ve all been 'David' at some point. Let that guy learn.

@art_codesmith @tante Loneliness is a sad thing.
@art_codesmith @tante I had picked up DHH was sitting on the fence, so DHH is loosing his AI-virginity to openclaw?! in other news...
@tante The similarities to the concept of "unprotected intercourse" are more than striking here.

@tante If the people doing that would understand security they would be very upset.

I'm not worried about the competent people, I'm worried about the majority that doesn't even consider why giving access is a bad idea.

@tante I agree BUT even if you run it on a Mac Mini or VPS or whatever without giving it your own data or accounts, it can still go and bother other internet users, for example by emailing them, sending spurious Github pull requests, making phone calls (!), the recent case of one making a hostile blog post about a named person, and so on.

Even if you make it not a danger to you, it remains a danger to others.

@tomw @tante The thing becomes then, why would you run it without giving it access to your accounts? The whole point is that it can Do Stuff For You.
@tomw @tante We desperately need a few incidents where their operators are identified and legally held responsible.
@larsmb @tomw they are identified: By OpenAI's and Meta's and Anthropic's HR department to offer them jobs
@tante @tomw I didn't mean the one or two people who had them vibed, but the masses who run and operate them even more recklessly.
@tomw @tante we've gotta put some kind of regulation in place for this. What happens if an agent does something illegal? Presumably someone needs to be held accountable (I would hope it would be the operator). So there's got to be some way to link agent to operator. Otherwise, where's the accountability?
Openclaw: KI-Agent leert unerwartet ganzen E-Mail-Posteingang - Golem.de

Eine KI-Sicherheitsforscherin ließ Openclaw nach wochenlangen Tests ihre echten E-Mails verarbeiten. Doch das ging gehörig schief.

Golem.de

@lennybacon

"Das wird nicht nochmal passieren."

😂😂😂

@tante

@tante

“Meaningful system” here, includes a network.

@tante Here's the corporate vice president of Microsoft Word running #OpenClaw to organize his family tasks and comms: https://www.omarknows.ai/p/meet-lobster-my-personal-ai-assistant

Don't worry, I'm sure no malicious actors will ever target his setup to compromise #Microsoft network. I mean, they'll only have access to all info and location of his loved ones. Surely there's zero monetary value in that...

Meet Lobster 🦞: My Personal AI Assistant

And How You Can Build Your Own

OmarKnows AI

@tante OMFG. He has now connected #OpenClaw to his smart home AND published a full playbook of his setup:
https://lobster.shahine.com/

A senior leader at Microsoft basically gave the keys to his front door to autonomous GenAI agents.

Lobster

Omar Shahine's personal and family AI agent, running on OpenClaw

Lobster | OpenClaw Playbook

@jukkan @tante

I am so glad that the executives at Microslop take security so seriously. Lol. Lmao.

@jukkan @tante fucking hell, some people REALLY crave having a personal assistant a la JARVIS. i particularly find the bit at the end creepy, where the AI compliments him on how "well written" his post is. validation/praise sycophant machine...
@patrick_h_lauke @jukkan @tante I mean - I’d want one too. I would also like to have a wand from any magic world you can imagine so magic could do my dishes. Getting either of those before the end of this decade is … unrealistic.
@hiiaminfi @jukkan @tante "aren't you worried about privacy? no, since i pay for the service" oh sweet sweet summer child

@jkmcnk @patrick_h_lauke @hiiaminfi @tante "Physical security matters. The Mac is in my office, which is in my house. If someone steals it, I have bigger problems."

2 weeks later: "Behold! I have connected #OpenClaw with the smart home system that can open the door to my house!!!"

https://www.omarknows.ai/p/lobster-gets-a-playbook-and-a-superpower

Lobster Gets a Playbook (and a Superpower) 🦞🏠

I really really cannot express how useful Lobster is to me

OmarKnows AI
@jukkan @jkmcnk @patrick_h_lauke @tante this also boggles my mind. I own Smart home stuff. I get the Appeal. I really do. But Smart Locks creep me out.
@jukkan @jkmcnk @patrick_h_lauke @hiiaminfi @tante
Not exactly security-related, but I can't help wondering... why don't these family members just *talk* to each other about their movie plans or travel memories? Why doesn't this guy just pop upstairs and peek into his kid's room to see if he's asleep, like a normal person? (Tangentially, why does a Microsoft VP use an exclusively Apple-based setup for his personal computing?) This seems to me like an overly technologically fetishized way to conduct one's private life. But then I'm not a tech VP, what would I know 🤷‍♀️
@ssilvonen @jukkan @jkmcnk @patrick_h_lauke @hiiaminfi Because dudes in tech rather talk to a chatbot than their family or a therapist.
@tante @jukkan @jkmcnk @patrick_h_lauke @hiiaminfi
Yes, this case certainly makes that painfully obvious.
@ssilvonen @tante @jukkan @jkmcnk @hiiaminfi in fairness, as a father of a stroppy teenager myself, i admit that i do often end up just communicating with him upstairs via whatsapp. however...it's still me doing it. what this agentic thing does is to make "speak with the actual person to check/coordinate things" into a step for the machine to do, as part of a larger process. i can understand why, but it doesn't make it any less dehumanising and odd...
@patrick_h_lauke @tante @jukkan @jkmcnk @hiiaminfi Well yeah, any way you can get your teenager to communicate counts as a plus! What strikes me as mechanized here is using "Lobster" to relay messages between family members when they could just as easily chat directly with each other. The use of an agentic AI assistant to organize your calendar and travel plans etc. seems to me a more natural and helpful use case (security issues aside), where the "communication" is just between one person and the AI.

@patrick_h_lauke @ssilvonen @tante @jukkan @jkmcnk @hiiaminfi There's nothing wrong with communicating with someone over their preferred channel. It being WhatsApp (ick Facebook but whatever, that's not the point here) doesn't make it any less communicating than using a normal phone call or text or shouting across the house.

What is gross is putting an extractive capitalist business in the middle as the mediator of that communication so you never actually get to communicate with each other.

Practically the point of technology for them, as far as I can tell.

@tante @ssilvonen @jukkan @jkmcnk @patrick_h_lauke @hiiaminfi

@ssilvonen I have no shade to throw at families using a family calendar; when my kids were in high school and my wife and I still traveled for work occasionally, etc. it was the only way to have a source of truth that people could rely on. It was especially helpful to tell the teenagers "if it's not on the calendar, I don't know about it" to avoid those "oh, but you were supposed to drive me to $thing, didn't I tell you? Oh, well, I really need a ride now." And so on.

I don't advocate for putting an LLM in the middle of all that, of course, but there is much to be said for putting family plans on a shared calendar and making it an expectation that every member checks said calendar. It doesn't replace conversation and such, but it sure can cut down on a lot of problems.

@jukkan @jkmcnk @patrick_h_lauke @hiiaminfi @tante

@jzb Oh yes, family calendars can be very useful, and printed ones have been around long before computers (we actually have one hanging in our kitchen right now). It's just the use of AI to mediate between family members using the same calendar that feels unnecessary and contrived to me. But of course this is a new development and this guy is clearly very excited about his current setup - it would be interesting to know if he can actually make it stick and how things are going a year from now.

@patrick_h_lauke @jukkan @tante come on at least he has multiple agents to isolate environments or something. That’s almost professional.

(I am kidding of course)

I think the most diplomatic take I can get behind is: „If you are reading a tutorial about how to set OpenClaw up: This is out of your league, keep your hands off.“

@patrick_h_lauke @hiiaminfi @jukkan @tante if i contributed to a w3c group and acted this flippant about widespread corporate exploitation i somehow don't think i personally would choose to do so by describing users who expect a very basic social contract as fundamentally misguided and naive. unless you're about to describe how you protect users from that and consider it to be a basic fundamental right. i believe that's what most people who are invited to contribute to committees that maintain a surface level appearance of democratic input usually do
@patrick_h_lauke @hiiaminfi @jukkan @tante i just can't see myself ever being motivated to find any mirth in someone thinking they're safe and being deceived. sorry

@patrick_h_lauke @jukkan @tante You mean: Daddy/Mommy issues are the driver of this AI bubble?

If they had had someone in their childhood telling them they did a good job we could have saved the world from climate change and wild capitalism?

@jukkan @tante

> From this point forward, this is only going to get better and more amazing

There's *no possible way* these words will ever come back to haunt him

@jukkan @tante @david_chisnall
To be fair, how much more damage could a malicious actor do to Microsoft compared to their current management?
@jukkan @tante Let me get this straight... He has full access to the AI while his wife and kids are relegated to "read dad's schedule and get reminded when he needs something" access? Sounds like a wonderfully narcissistic "dad." The kind that would run a modern day generative AI company. 🙄
@jukkan Tim Apple will join the gang one day. It is bound to happen.
@jukkan @tante is he really claiming that if you run openclaw on a laptop connected to Claude that this counts as 'local AI'?
@catch56 @tante It's coming from the makers of "sovereign EU cloud" that is still subject to US CLOUD Act.

@jukkan " → Catch-all binding
→ Route to: lobster agent
→ Apply: No sandbox, full access "

HAHA!

FFW 6 months - "Lobster, tell Omar I want a divorce"
Yeah fun fact: the people running companies and taking business decisions, are not usually fit to do the design or programming and those are the ones most excited about this kind of garbage, so it is not a tech problem, is a politics problem