News for hackers. Stuff that matters.
Re:California law banning bots from pretending to be real pe... by dessant on Friday July 05, 2019 @02:04AM (#20360194)
I felt stupid, but I have fallen victim to this on dpd.com while tracking a package. A helpful chat popup has appeared where I could request assistance from a support agent, but it was not disclosed that the agent is a bot.

Needless to say I have spent a couple of minutes repeatedly asking a question, and even rephrasing it while being frustrated that this person does not seem to grasp my issue.

Re:California law banning bots from pretending to be real pe... by jonnydubowsky on Friday July 05, 2019 @03:05AM (#20360456)
My recent experience with a credit card customer support agent might as well have been me conversing with a bot. It was actually more frustrating for both me and the agent, that they were so strictly bound to a script, with no room for their creative human brain to actually adapt their response to the unique context of my issue. When customer service is nothing more then an agent choosing from various scripted responses, what is the point? Escalating my issue up through 2 senior agents, i finally reached a human who was allowed by their employer to actively apply context to their support tasks. The use of programmed constraints on a system and how that system is sold to customers, this is a wicked problem of our times.
Re:California law banning bots from pretending to be real pe... by amelius on Friday July 05, 2019 @05:23AM (#20361059)
I guess the point is that they still save on average. Most calls can probably be handled using a simple script.
Re:California law banning bots from pretending to be real pe... by jacquesm on Friday July 05, 2019 @06:38AM (#20361430)
They may save on average, but that's just shifting the expense to the customer in terms of wasted time. If there is anything I've learned over 30 years of running businesses that being disrespectful towards your customers' time it is that you will end up losing those customers.
Re:California law banning bots from pretending to be real pe... by headsoup on Friday July 05, 2019 @07:06AM (#20361560)
The saving grace for these online services seems to basically just be that the potential customer base is so ridiculously huge that it still pays vs effort required. Same for the big social giants etc.

No matter how many people know better, there are plenty more that don't.

Re:California law banning bots from pretending to be real pe... by AnIdiotOnTheNet on Friday July 05, 2019 @07:54AM (#20361797)
Another way they save is by discouraging their use. Since people who've called with real issues have had to waste a whole lot of their time, they're more reluctant to do so the next time they have a problem.
Re:California law banning bots from pretending to be real pe... by notafrog on Friday July 05, 2019 @05:45PM (#20366135)
A few months ago, I contacted Microsoft support regarding an old Hotmail account of mine. I couldn't get access due to my own fault in the end, as I had set it up as a kid and 10 years later I could barely remember any details. However, the conversation with the support agent was by far the best. It felt natural, without the "copy/paste" ready-to-use responses they usually have.
Re:California law banning bots from pretending to be real pe... by luckylion on Friday July 05, 2019 @02:30AM (#20360303)
AliExpress does this as well and in an annoying manner that I consider insulting. Support chat will start with a useless bot that just looks for keywords and then replies with FAQ links. "Escalate to a service agent" will make you wait for 30-45 seconds to then be greeted by "Anna" or some other human name. It's another bot, with typing-delay and will follow a simple script to keep you busy, low quality and no apparent option to actually get to a human support agent.
Re:California law banning bots from pretending to be real pe... by dorgo on Friday July 05, 2019 @04:34AM (#20360832)
I'm curious. Are you complaining about support not beeing able to solve your issue, or about talking to bots instead of humans? Assume a strong AI with enough authority would handle support. Would you still prefer humans?
Re:California law banning bots from pretending to be real pe... by tastroder on Friday July 05, 2019 @06:25AM (#20361370)
I'd have a problem with both in the support case. The alternative to a human contact is a natural language interface that just wastes my time and/or laziness in other parts of the business. If my problem is covered by your FAQ, improve search. If my problem can be solved by a bot, let me do it in your UI. If handling my problem through a human requires more information prepend a form. Provide a clear path to skip/escalate.

To be fair though, I'm annoyed by human first level support for similar reasons most of the time but at least escalation/problem solving with them is easier than a bot that explodes when you go off-script. I think, personally, that making me talk to your fancy answering machine is just devalueing. If a company wants my business they should have human support and pay them enough to work with the customer.

Re:California law banning bots from pretending to be real pe... by notahacker on Friday July 05, 2019 @07:27AM (#20361663)
> If my problem is covered by your FAQ, improve search. If my problem can be solved by a bot, let me do it in your UI.

This assumes that people who have a question they need answering will consider searching for an answer or getting to grips with the UI first, rather than heading straight to ask customer support. That's often not the case. Long before support chatbots became widespread, telephone trees made a point of telling every single user in the queue the help section of their website existed...

Agree bots should often make it easier to escalate and should default to escalation rather than blowing up[1], but if a large proportion of your queries are actually answered by telling people the FAQ page is a thing, it's a bit difficult to justify paying humans to do it.

[1]if escalation takes a couple of minutes for a response escalation and blowing up might be indistinguishable...

Re:California law banning bots from pretending to be real pe... by tastroder on Friday July 05, 2019 @08:02AM (#20361839)
Yeah, I can totally agree with that under the constraints you mention. I might conflate my annoyed experience with marketing type bots with the actual requirements for a support one, there's certainly dark patterns with how bots are implemented and integrated these days but if they become more wide-spread I guess we'll figure out how to tackle those as well.
Re:California law banning bots from pretending to be real pe... by luckylion on Friday July 05, 2019 @04:42AM (#20360870)
If it could, I wouldn't mind. It didn't, and then went into a loop allowing me to "escalate" it to a real human which turned out to be the same bot, only with a different name.
Re:California law banning bots from pretending to be real pe... by burfog on Friday July 05, 2019 @02:29AM (#20360300)
I've been assuming that it's a pre-determined initial message, but that a real human in a low-cost country would immediately be involved if I bothered to respond.

What portion of the chat pop up windows do you think are purely bot? Might any of them be purely human?

Re:California law banning bots from pretending to be real pe... by LostJourneyman on Friday July 05, 2019 @02:35PM (#20364699)
It depends build to build. The standard approach is one of 4:

1) The bot initiates the conversation and your initial message gets sent directly to an agent. This is the older model that's in most common use today.

2) The bot initiates contact and based on your responses does some simple keyword matching and delivers help article links where possible or asks for more information IVR style, then when it hits an "I don't know" point or if the agent option is selected, offloads to an agent.

3) This is my favorite style, honestly: The bot initiates the interaction, and does some machine learning backed AI chat, all the while the interaction is monitored by an agent who can take over at any time. Similar to #2, if the bot hits a sticking point, it'll just queue to an agent. This unfortunately is the least common of the implementations.

4) This is the most modern and is becoming the industry leader: Fully AI bot trained against a veritable Everest of chat conversations for that entity/industry, only offloads to a human when you shout "HUMAN" at it enough times or if it gets really stuck and confidence intervals start falling rapidly.

NOTE/DISCLAIMER: I design and implement these systems for a living, and we don't often get much say in the customer-side UX, so I'm sorry if you've gotten stuck with an arguably bad build!

(Edit: formatting)

Re:California law banning bots from pretending to be real pe... by HillaryBriss on Friday July 05, 2019 @05:26PM (#20366010)
#3 reminds me of self-serve retail checkout kiosks: when I get stuck, or the kiosk gets stuck, a nearby live human approaches and resolves the issue (except when there's no live human nearby in which case I just get to stand there feeling awkward and useless. it's pretty humbling.)
Re:California law banning bots from pretending to be real pe... by zamadatix on Friday July 05, 2019 @05:17PM (#20365954)
https://xkcd.com/806/

Nobody will notice, you know you want to add it ;).

Re:California law banning bots from pretending to be real pe... by LostJourneyman on Monday July 08, 2019 @12:39PM (#20384176)
You have NO IDEA how tempted I've been to add that! I've literally had printed this out and pinned to the wall of my cubicle for years!
Re:California law banning bots from pretending to be real pe... by DEADBEEFC0FFEE on Friday July 05, 2019 @04:01AM (#20360694)
Bot initiated. Human, assisted with menus of canned phrases. "Just one moment while I check that for you." Bot to wrap-up; survey, emails, etc

That's my assumption and if I'm not in a rush, pretty ok with it.

Re:California law banning bots from pretending to be real pe... by viraptor on Friday July 05, 2019 @03:35AM (#20360571)
It really depends on the site. Some go to bots, some to script-followers, some to good dedicated support (shout out to datadog here). Some vary by time of the day.

But yes, some of them are contact boxes for real people.

Re:California law banning bots from pretending to be real pe... by danieltillett on Friday July 05, 2019 @02:28AM (#20360292)
Apparently the solution is just to swear at the bot and you get punted to a human, the downside is many of the humans are no better than a bot.
Re:California law banning bots from pretending to be real pe... by kbenson on Friday July 05, 2019 @01:26PM (#20364183)
Or you get a human that assumes the customer they are helping is abusive or problematic based on the prior messages, in which case they may be very unmotivated to help you.
Re:California law banning bots from pretending to be real pe... by danieltillett on Friday July 05, 2019 @10:26PM (#20367360)
Just abuse the bot, not the human :)
Re:California law banning bots from pretending to be real pe... by WalterBright on Friday July 05, 2019 @02:11AM (#20360218)
Those are fairly easy to spot. One dead giveaway is their excessively chirpy voice.

Or, as you did, just say something that puts them off script.

Re:California law banning bots from pretending to be real pe... by wpietri on Friday July 05, 2019 @02:17AM (#20360244)
Currently they're fairly easy to spot -- if you work in tech, anyhow. A lot of money will be put into making them harder and harder to spot. I'm glad to see this law come at the beginning of that process, so there isn't an established industry that can put money into fighting it.
Re:California law banning bots from pretending to be real pe... by luckylion on Friday July 05, 2019 @02:27AM (#20360287)
> Or, as you did, just say something that puts them off script.

I don't know about that. Does Amazon's support use bots? Or are they just in a remote call center using auto translate to communicate with customers?

Because in general, they appear like bots to me: they will parse for key words and reply to those, ignoring context and subtle differences that make their replies sound weird.

Re:California law banning bots from pretending to be real pe... by egypturnash on Friday July 05, 2019 @10:57AM (#20362968)
My support experience with Amazon Fulfillment involved several months of going back and forth with what I think were humans whose native language was not English, bound by a script, until I learnt that the shortcut to getting an actual human who is allowed to deviate from a script is sending an email to Bezos.
Re:California law banning bots from pretending to be real pe... by WalterBright on Friday July 05, 2019 @03:12AM (#20360474)
I doubt it would be better than their email reply bot, which is transparently inept. I wouldn't mind an email bot if it got me what I needed, but it was useless.
Re:California law banning bots from pretending to be real pe... by viraptor on Friday July 05, 2019 @03:37AM (#20360583)
> Does Amazon's support use bots?

Depends on the section of Amazon. AWS enterprise support does not.

Re:California law banning bots from pretending to be real pe... by luckylion on Friday July 05, 2019 @03:51AM (#20360649)
I'm talking about retail / security.
Re:California law banning bots from pretending to be real pe... by lathiat on Friday July 05, 2019 @03:09AM (#20360468)
I guess you didn't watch the Google I/O presentation earlier this year (which is probably what this law is about I would guess): https://www.youtube.com/watch?v=D5VN56jQMWM
Re:California law banning bots from pretending to be real pe... by WalterBright on Friday July 05, 2019 @03:22AM (#20360514)
It was selected out of a large number of tests they made. So while it sounds good, I'll remain skeptical for now.
Re:California law banning bots from pretending to be real pe... by progval on Friday July 05, 2019 @03:33AM (#20360561)
I run an IRC bot using Markov chains to generate text. It's a very crude method, so usually it's pure nonsense. But sometimes it says a bunch of sentences that make sense in a row to a new user, who then thinks it's a human.