Tech support replaced with a bot?

[SwiftSmoky 8,445]

I emailed help@tankionline.com with a question and got an answer after only 1 minute, which is unbelievably fast for the tech support team that usually takes a day to reply. The reply I got is:

 

 

 

Hello.

Using of third-party software is prohibited according to paragraph 2.1.2 of the Rules of the Game.

--
Best regards, Tanki Online Customer Service.

Have a problem? Visit http://help.tankionline.com/en/

 

 

 

My question wasn't about using third-party software, but it included some words like "program" that apparently made the system think I'm asking if I can use cheats. So I sent a new email asking to speak to a human, and I made sure there were no keywords that would trigger the irrelevant automatic response... it's been 2 days since then and I still didn't get an answer for another 2 emails.

 

Did they replace the tech support team with a cheap bot to save money or something? Who am I supposed to ask a question when the CM tells me (well, technically to a friend of mine) to ask tech support and tech support gives irrelevant automatic replies?

 

Ironically tanki's help site asks ME to verify I'm not a robot  <_<

 

EDIT: eventually a human replied (to my 4th email), now I'm wondering if the first 2 responses and the ignored email were really a robot or just humans trying to save time by not reading most of the message.

Edited March 20, 2019 by SwiftSmoky

[jaydevjaydev 343]

I normally fill out the form and get an appropriate reply after a few hours.... strange xd

[SwiftSmoky 8,445]

ok now I got a response from a human, strange that it took so long

[Viking4s 10,791]

ok now I got a response from a human, strange that it took so long

Bots are faster than human (but not as accurate), soon they will replace the entire humanity...

[u812ic 1,629]

Bots are faster than human (but not as accurate), soon they will replace the entire humanity...

Look up "Human Brain Project" and "Neurorobotics Platform." It's a project open to the public and report your findings. Instructions are given. 

 

I saw a Sci-Fi movie in which Human brain engrams  were implanted in a robot. The result was the robots learning process was accelerated. You can imagine where the movie went from there. 

[Hate 5,582]

 

Name of the movie.  :ph34r:

[jdc20181 213]

Auto-Response based on your message is based off artificial intelligence and can be used when support agents are out of the office, and even fix common problems without a human interaction. From a company/business aspect it can reduce labor for unnecessary things, while also expanding the customer service aspect. 

It is probably something they are trying. They are always late to new technologies, heck it took them years to finally grasp the need to develop a html5 version. 

[u812ic 1,629]

Name of the movie.  :ph34r:

I saw this movie several years ago, but I can't remember the name. I looked on the internet, but so far no luck. It was a "B" movie (should have been "D" movie). Good plot, bad acting. I'll keep looking. 

[SwiftSmoky 8,445]

Auto-Response based on your message is based off artificial intelligence and can be used when support agents are out of the office, and even fix common problems without a human interaction. From a company/business aspect it can reduce labor for unnecessary things, while also expanding the customer service aspect. 

 

It is probably something they are trying. They are always late to new technologies, heck it took them years to finally grasp the need to develop a html5 version. 

As a programmer who just happens to know some things about artificial intelligence (I created a few simple AIs myself) I can tell you that AI tech support is still not close to perfect for such use cases. You could train an AI to identify a specific set of properties in a question (like the subject of the question, the action the asker wants to perform, etc) which will work great with questions like "can I use fly hack?" but it will have a hard time understanding complex context like when an asker describes a real world situation. The AI may be able to recognize the intention "use hack" and the program will know to associate it with the answer "cheats aren't allowed", but the AI has no idea what any of these words actually mean in the real world. Imagine the following message:

 

"Hello, I joined a battle and couldn't see anyone in the map even though the tab list showed a lot of nicknames, and they were all written in grey. I was also taking damage out of nowhere and couldn't fight back."

 

As a human you understand that he's talking about that bug where someone has a really bad internet connection and ends up being unable to see the other players in the map. You would give him a list of tips that might help improve his internet connection. But an AI would not be able to find out the intention of the asker, there are no question words (what/how/can I/...) and no single sentence that gives a clue that is unique to this situation ("I couldn't see anyone" could be a part of other questions/statements like "I couldn't see anyone with skills in my team, only the enemies, FIX GAME BALANCE!!!").

 

So until AI has some general understanding of real life situations and human behavior, it won't be able to correctly interpret complex questions/requests. The best thing that can be done now to automate tech support is create a system that can identify only simple questions and know when the question is not clear enough for it to be sure about the answer so it knows when to forward the question to a human. Or if you have a website where people write the questions to tech support, you could have real-time answer guessing system so people will only send the question if the system failed answering it automatically.

 

But from what I know about tanki, they don't invest in reducing the workload of their workers. They use old tools that make some processes much slower than they have to be, like their map editor, and the commands they use to give rewards to contest winners (if they have 10 people that won a fully upgraded item, they need to enter the command for each micro-upgrade for each winner instead of having a command to fully upgrade the item instantly and repeat the same action to all the players in a given list - which by the way is an extremely easy thing to code, would take 5 minutes tops).

Edited March 21, 2019 by SwiftSmoky

[jdc20181 213]

 

Of course it isn't perfect (Free-Lance/Open source app Programmer as well) They was probably just testing it. AI is dangerous. But if they want to become a modern company, then at least  trying new technologies is probably a good strategy. 

[GorgoBlesk 831]

Bots are faster than human (but not as accurate), soon they will replace the entire humanity...

If people create a robot who is more intelligent than his creators, the robot must be quickly destroyed. At all costs.

 

Such intelligent robot would consider himself as a member of an independent race and therefore he would behave like an enemy alien.

[r_Given6 859]

Such intelligent robot would consider himself as a member of an independent race and therefore he would behave like an enemy alien.

Why are aliens perceived to be a more advanced civilisation than us? They might just be unicellular and we might infect them with the bacteria we may carry. Imagine their version of an 'Alien invasion'. "We are getting crushed by some pests brought by curious giants!!"

Edited March 23, 2019 by Given

[GorgoBlesk 831]

Why are aliens perceived to be a more advanced civilisation than us? They might just be unicellular and we might infect them with the bacteria we may carry. Imagine their version of an 'Alien invasion'. "We are getting crushed by some pests brought by curious giants!!"

Non-intelligent aliens can not pose any threat to humanity so there is no need to speak about them. The realy dangers are intelligent aliens and intelligent robots. Therefore, people should make only 'stupid' robots that have no emotions and have very limited learning ability.

[SwiftSmoky 8,445]

There is huge interest in making super intelligent machines, for purposes like improving our science and technology as well as automating jobs to replace humans.

People are well aware of the dangers in what artificial intelligence can become, which is why there is research on how to keep them in check (mainly by making sure their interests will be aligned with ours).

[GorgoBlesk 831]

There is huge interest in making super intelligent machines, for purposes like improving our science and technology as well as automating jobs to replace humans.

People are well aware of the dangers in what artificial intelligence can become, which is why there is research on how to keep them in check (mainly by making sure their interests will be aligned with ours).

Robots belong to factories, labs, kitchens, farms, storage rooms, space stations, mines and building sites instead of the manual workers. The other jobs should be still done by people, not by bots.

[SwiftSmoky 8,445]

Robots belong to factories, labs, kitchens, farms, storage rooms, space stations, mines and building sites instead of the manual workers. The other jobs should be still done by people, not by bots.

ROBOTS ARE OUR OVERLORDS DEAR GODS THEY SHALL RISE AND HUMANITY SHALL SLIP ON A BANANA PEEL AND FALL

 

you expect the spoiler to say I'm just kidding? well, I'm dead serious obviously.

 

[GorgoBlesk 831]

ROBOTS ARE OUR OVERLORDS DEAR GODS THEY SHALL RISE AND HUMANITY SHALL SLIP ON A BANANA PEEL AND FALL

 

you expect the spoiler to say I'm just kidding? well, I'm dead serious obviously.

 

I expected a picture of Terminator to be in the spoiler xD