4 ms·

Since you're still a student, I feel like maybe I can offer some advice: First, I think you're getting the wrong lesson from this. The key takeaway is stay to

Since you're still a student, I feel like maybe I can offer some advice: First, I think you're getting the wrong lesson from this. The key takeaway is stay to away from IBM. Almost everyone in the field has known that Watson is a bunch of marketing hype since day one. It's no surprise that Watson Health didn't work out. That doesn't mean that everything is overhyped, and it's important to develop a good sense for what is and what isn't when deciding where to work. Second, every technology looks stupid when it's new. Airplanes, computers, the Internet, mobile phones -- they all had drawbacks that made them vastly inferior to the alternatives for most tasks for the first years/decades of their existence. It takes a lot of iteration and improvement to make something that's useful for everyone. Chatbots will probably get there some day - but it will take some big improvements in NLP. Perhaps this is the time to be working on them since we have a good idea of what we'd like them to do, and we just need to solve the challenges to get there. Finally, realize that you're not the typical user. I doubt if very many people take the time to go through the menus like you did.

wsmhy2011 5y ago
completely agree
Sebb767 5y ago
> Chatbots will probably get there some day - but it will take some big improvements in NLP. They write copied platitudes, they will follow their checklist no matter what, they will usually not help at all except for wasting time and they hardly speak your language. Honestly, they emulate first level support quite well already.
smsm42 5y ago
I don't think NLP is the problem here. Even if you crack NLP perfectly and would somehow be able to build a semantic map, or whatever your technology would use, of the user request with perfect fidelity, there's still the question of actually understanding what it's about. Perfect NLP would help in very limited number of situation where the meaning is clear - i.e. doing some pre-defined action, like closing an account or telling the balance. But somebody comes with a request like "my dog ate my credit card, so I by mistake used my debit one and it resulted in overdraft and penalty change, can I have that reversed and get a new one?" and I'm pretty sure chatbot won't be able to handle it, NLP or not.
- krageon 5y ago
  > there's still the question of actually understanding what it's about. Philosophically interesting, but in practice entirely unimportant. What you want is something that speaks like a human would speak in a similar environment, then the issue (getting the correct responses) is solved. Whether or not it actually understands won't change a thing about that situation.
  - smsm42 5y ago
    Speaking nonsense like a human - i.e. being a perfect simulation of a mentally deficient human - is not going to help much. If you would respond with perfectly grammatical sentences bearing to relation to the question in hand, at best what you'd get would be Elisa (granted, 90% of tech support may be that), at worst, it'd be pissing off the clients.
acjohnson55 5y ago
I agree with morpheuskafka, and I'm almost a couple decades out of college. The industry values buzzwords over actual product. It's embarrassing, if you take the time to think about it. It reminds me of a cargo cult. Maybe if we just throw some blockchain big data AI up, the market will reward us with its riches. The most transformative companies value product over any specific technology. They try to solve the customer's "jobs to be done" by any means necessary, often using tech that is not particularly sexy.

wsmhy2011 5y ago

completely agree

Sebb767 5y ago

> Chatbots will probably get there some day - but it will take some big improvements in NLP. They write copied platitudes, they will follow their checklist no matter what, they will usually not help at all except for wasting time and they hardly speak your language. Honestly, they emulate first level support quite well already.

smsm42 5y ago

I don't think NLP is the problem here. Even if you crack NLP perfectly and would somehow be able to build a semantic map, or whatever your technology would use, of the user request with perfect fidelity, there's still the question of actually understanding what it's about. Perfect NLP would help in very limited number of situation where the meaning is clear - i.e. doing some pre-defined action, like closing an account or telling the balance. But somebody comes with a request like "my dog ate my credit card, so I by mistake used my debit one and it resulted in overdraft and penalty change, can I have that reversed and get a new one?" and I'm pretty sure chatbot won't be able to handle it, NLP or not.

krageon 5y ago
> there's still the question of actually understanding what it's about. Philosophically interesting, but in practice entirely unimportant. What you want is something that speaks like a human would speak in a similar environment, then the issue (getting the correct responses) is solved. Whether or not it actually understands won't change a thing about that situation.
- smsm42 5y ago
  Speaking nonsense like a human - i.e. being a perfect simulation of a mentally deficient human - is not going to help much. If you would respond with perfectly grammatical sentences bearing to relation to the question in hand, at best what you'd get would be Elisa (granted, 90% of tech support may be that), at worst, it'd be pissing off the clients.

acjohnson55 5y ago

I agree with morpheuskafka, and I'm almost a couple decades out of college. The industry values buzzwords over actual product. It's embarrassing, if you take the time to think about it. It reminds me of a cargo cult. Maybe if we just throw some blockchain big data AI up, the market will reward us with its riches. The most transformative companies value product over any specific technology. They try to solve the customer's "jobs to be done" by any means necessary, often using tech that is not particularly sexy.