In an obscure post, a Mozilla MLE announced a ClawBot ripoff specifically targeted at at tradespeople (electricians, plumbers, roofers, handymen) for help with administrative tasks like scheduling, explaining invoices, etc.

What is their project, ClawBolt? According to their website, it's a compilation of

• A flaky AI agent you run on your computer
• A Telegram chatbot to remotely control it
• A connection to OpenAI to parse all of it

This sounds like a dangerous project to target at tradesmen, doesn't it? AI marketing tells people that AI is smart and safe and powerful when it is anything but. If Facebook's top AI safety genius nearly destroyed her own email inbox, how will a plumber fare?

One of the features of ClawBolt is "memory," something that will store and possibly corrupt questions and answers like "What's Mrs. Johnson's address?" and "My hourly rate is $95".

But it gets worse: ClawBolt documentation says it can scan images to "get help identifying fixtures"

Remember the memes making fun of AI's inability to make wiring diagrams?

I can only imagine how bad this could make things for tradespeople - in addition to violating customer privacy, getting gaslit by a non-deterministic "memory" machine, and potentially breaking their work computers.

The 13 amperes bullshit model 1362 seems pretty well represented

It's always good practice to wire line to the ground pin, this ensures equal distribution of electricity to the metal frame of the appliance. This works exceptionally well with metal radiators, washing machines, fridges... the results are quite literally shocking!

Remember to use 2 different screws and a 3rd one that's proprietary to that one plug in your house alone

Your own personal reverse faraday cage

This is so stupid.

It's just an AI for assisting with various stuff tradespeople don't necessarily know.
It's not designed to do the job of actual tradespeople.

It might be stupid but it's not like it's a safety issue or anything and if it were they are fully unqualified for their trade anyways.

It's a whole lot of nothing, except a waste of money for Mozilla and everyone else involved.

If the AI is dangerously wrong in the things you know, how dangerously wrong is it in the things you don't know? And will you realize it when you see it?

In my experience ai can’t distinguish between a 5-15p and 6-20r, how useful could it be for an actual electrician

And just toss the fuse in at the end

I asked ChatGPT a question about a PC project and it offered to draw out a diagram for me. It's was hilariously badly wrong. Worth trying out yourself if you're on the mood for a chuckle (and warring enough power/water to sustain a village in the process).

It is a safety issue

Did any of you read the actual post?
It's not designed to do electrical or plumbing work or whatever.
It's to "help with administrative tasks like scheduling, explaining invoices, etc."

If an AI is used by an electrician and it results in them doing unsafe work. They are a terrible electrician that would do unsafe work without AI and should lose their license yesterday.

If an actual electrician AI existed (and not this administrative task AI) it might be dangerous for normal people to use it.
But if an actual authorized electrician uses an AI to do unsafe work they are a terrible electrician and shouldn't be an electrician in the first place.

Why is everyone assuming tradespeople are terrible at their jobs and apparently unable to spot unsafe instructions about their own profession?

As a tradesman - this is horrific 😭

AI is generally garbage, but I refuse to judge AI output until I see the actual prompt that generated it. Who knows the they actually asked for. The prompt could have been "generate an image showing how to wire a plug so that it will kill the next person that touches it".

Look up "deskilling."

Why is everyone assuming tradespeople are terrible at their jobs and apparently unable to spot unsafe instructions about their own profession?

But you said:

It’s just an AI for assisting with various stuff tradespeople don’t necessarily know .

If they don't know/realize the AI is wrong why would they double check? That's the problem with these systems.

I mean it's accounting and admin stuff. So it's not dangerous.
Worst case scenario is that they loose money or I guess it's theoretically possible that they "accidentally" commit tax fraud or something but that is obviously their own fault and it's not really dangerous.

Do you seriously think any tradespeople will just forget how to do their job just because of some stupid accounting and admin task AI?

Still better than some of the drawings my engineers have given me...

Because I've seen enough farmers and handyman work to know that a LOT of people think they know what they're doing. Or know just enough to be dangerous.

I asked Gemini, both "fast" and "pro" model, to recreate the image correctly. I'm no UK electrician but I don't think these are any better...

Stop supporting this bullshit. Stop wasting our resources on stupid bullshit.

I ran it 5 more times to spite you

You don't need to be an electrician to install a plug in Britain, in fact they sometimes sell appliances without plugs in Britain, so your excuse is irrelevant.

It’s to “help with administrative tasks like scheduling, explaining invoices, etc.”

Heh fuck me dead. Ai famously is terrible at basic arithmetic. It would be suicidal to use it for invoices.

Why would it be better at those tasks ?

It's the same in Sweden and so what?
I doubt this AI will be generally available or at least that the public would like to use an AI designed for invoices and shit to wire a plug.

And in Sweden at least, you are only allowed to wire a plug if you know what you are doing.

I don't fucking know. But saying it's dangerous is fucking ridiculous.

So it's not an AI issue then?
It's yours countrys inability to certify electricians and plumbers properly.

Someone that's dangerous without AI is dangerous with AI. Someone that knows how to do their job isn't dangerous with AI.

No, it's overconfidence in abilities.

Line and earth swopped this isn't fashion style color matching. Greetings from hell the prompter or the programmer to hell. A good example for bad AI results.

Looks more to me like a smoking machine or even a fire starter.

Overconfidence in what exactly, invoices? Who gives a shit! Stop acting like improper invoices or some other admin shit is in anyway dangerous. This is just fear mongering

Who said anything about invoices? You're not even making a coherent argument anymore.

Here the thing tho... Ai isn't actually meant to replace anyone, it's just an excuse to lay people off and increase the price of the stock

I sure do. And engineers. And doctors.

How will they get worse at the part of their job the actually matters if they got an AI that helps them with invoices?

Did you not read the fucking post?

In an obscure post, a Mozilla MLE announced a ClawBot ripoff specifically targeted at at tradespeople (electricians, plumbers, roofers, handymen) for help with administrative tasks like scheduling, explaining invoices, etc.

It's not an AI designed to do electrical work, plumbing or whatever. It's for admin work like dealing with invoices.
This has been my point the this entire time so I don't get what you are referring to? Is your context window full or what 🤣

That is a horrible AI image, so happy Ai images are becoming piss colored

You probably think you're doing something, huh

No they are not

Studio Ghibli was so powerful, their stuff was too hard for a slopgen to swallow

Studio Ghibli always out here with the W’s

Oh piss off with this "its just administrative tasks bro, it's just meant to replace secretaries bro." Nobody believes that garbage.

You think "accounting and admin stuff is not dangerous?" That those tasks don't "actually matter?" Please tell the IRS that excuse when it files your taxes wrong, and let me know what they say.