How might AI augment human TECH service-people?

[MROSS]

How might AI augment human TECH service-people during service call to fix Internet, phone, and TV service problems?

Might AI help curb those "head scratcher" issues - that is uncertainies in how to specifically troubleshoot issues? The goals being human TECH service-people would mainly replace defective TECH hardware, adn non-hardware issues would pretty-much be handled off-site remotely.

To be more proactive, how might AI assist humans in resolving the periodic issues of audio-visual TV reception being disrupted? This is where the visual reception yields a scrambled pattern, and the audio-reception goes silent.

 

[Crossbreed]

I am skeptical that AI will be able to offer novel solutions that human reps do not already have access to (if slower). They would be like FAQ pages on steroids. [FAQ pages] can be helpful for simple, repetitive matters, but I have often exhausted them as a resource.

Me: I would like to report that your website is not working [so I can pay a bill, etc.].
ChatBot Rep: There is no human rep available to help you right now. For faster service, please use our convenient website,...

 

[Misery]

Depends on the nature of the problem. And on the user.

I've used AI for exactly this sort of thing, to assist in fixing some tech issue or whatever. It worked out well. But then, I'm used to dealing with AI.

A lot of people have a tendency though to expect it to be like this 100% perfect magic thing that'll instantly fix everything all at once, and then get super agitated when met with the fact that, ya know, nothing works that way.

And then in the case of this scenario, you've now got a super angry customer on your hands.

Though that only says so much, since in a lot of cases you've got an angry customer on your hands because the sky is blue, or because water is wet. And of course it's absolutely not the customer's fault that they poured salt directly into some opening on the TV to get the demons out of it, it should have had a warning against salt printed on it, in small words with pictures, how dare that not be the case.

I dont know how tech support people manage to do their jobs without going mad, considering some of the stories that come out of it.

 

[Crossbreed]

I dont know how tech support people manage to do their jobs without going mad, considering some of the stories that come out of it.

When I worked at a helpdesk, I took it all with a grain of salt...

 

[FilterFreq]

Might AI help curb those "head scratcher" issues - that is uncertainies in how to specifically troubleshoot issues? The goals being human TECH service-people would mainly replace defective TECH hardware, adn non-hardware issues would pretty-much be handled off-site remotely.

The main issue is that AI needs to be specifically trained and finetuned in order to produce accurate results. So if you ever ask it something just mildly esoteric or off the beaten path (usually the best questions, or the most important ones), it gives you bat-turd crazy results because it hasn't confronted what you're going through yet.

So as much as it seems like it'll eventually understand obscure or complex problems that humans haven't directly fed it, that's just not going to be the case in our paradigm. Well-understood material that it can aggregate is where it shines the most -- actual thinking and complex problem-solving is better suited for humans and always will be. There's a reason why we're the ones still feeding the machines -- day in and day out.

So for now, it's basically a copy-pasta machine that can lift some sizable weight. Great tool (or is it parlor trick?) in a pinch (especially when everyone 'conveniently' cripples their search engines to make way for it), but very little substance or problem-solving underneath. In short, $$$

 

[Outsider]

How might AI augment human TECH service-people during service call to fix Internet, phone, and TV service problems?

Might AI help curb those "head scratcher" issues - that is uncertainies in how to specifically troubleshoot issues? The goals being human TECH service-people would mainly replace defective TECH hardware, adn non-hardware issues would pretty-much be handled off-site remotely.

To be more proactive, how might AI assist humans in resolving the periodic issues of audio-visual TV reception being disrupted? This is where the visual reception yields a scrambled pattern, and the audio-reception goes silent.

To fix audio-visual reception an AI system has to guess what a signal looked like when it was in a pristine state. In general, this can be done but there is a limit -- a signal might be so corrupted that its recovery is not possible. Shannon's information theory sets the limit beyond which recovery is not possible. Currently the AI systems are nowhere near this limit, so there is plenty of room for improvement.

 

[Yeshuasdaughter]

I think most white collar, and many upper blue collar jobs will be handed over to computer programs.

My advice is take up the trades.

 

[Outsider]

I think most white collar, and many upper blue collar jobs will be handed over to computer programs.

My advice is take up the trades.

Currently I work as a data analyst. Analyzing data is easy, even a primitive AI system can do that. But project design and subsequent collection of data are very challenging in most cases and in some cases require public skills. I'm not worrying about loosing my job to even most sophisticated AIs.
In the past I used to work as a system controls engineer (I have MS in Mechanical Engineering). This job requires a lot of experimentation which makes it impossible for an AI.

 

[Crossbreed]

In the past I used to work as a system controls engineer (I have MS in Mechanical Engineering). This job requires a lot of experimentation which makes it impossible for an AI.

PLCs? I trained to program PLCs, but never got to do so professionally.

 

[Outsider]

PLCs? I trained to program PLCs, but never got to do so professionally.

No. We were not using PLCs. We were using regular PCs. I took a course in PLC programming, but never used the devices in practice.

 

[MROSS]

With AI-TECH growing, any further comments?