Slashdot Mirror
Did Google's Duplex Testing Break the Law? (daringfireball.net)
An anonymous reader writes:
Tech blogger John Gruber appears to have successfully identified one of the restaurants mentioned in a post on Google's AI blog that bragged about "a meal booked through a call from Duplex." Mashable then asked a restaurant employee there if Google had let him know in advance that they'd be receiving a call from their non-human personal assistant AI. "No, of course no," he replied. And "When I asked him to confirm one more time that Duplex had called...he appeared to get nervous and immediately said he needed to go. He then hung up the phone."
John Gruber now asks: "How many real-world businesses has Google Duplex been calling and not identifying itself as an AI, leaving people to think they're actually speaking to another human...? And if 'Victor' is correct that Hong's Gourmet had no advance knowledge of the call, Google may have violated California law by recording the call." Friday he added that "This wouldn't send anyone to prison, but it would be a bit of an embarrassment, and would reinforce the notion that Google has a cavalier stance on privacy (and adhering to privacy laws)."
The Mercury News also reports that legal experts "raised questions about how Google's possible need to record Duplex's phone conversations to improve its artificial intelligence may come in conflict with California's strict two-party consent law, where all parties involved in a private phone conversation need to agree to being recorded."
For another perspective, Gizmodo's senior reviews editor reminds readers that "pretty much all tech demos are fake as hell." Speaking of Google's controversial Duplex demo, she writes that "If it didn't happen, if it is all a lie, well then I'll be totally disappointed. But I can't say I'll be surprised."
John Gruber now asks: "How many real-world businesses has Google Duplex been calling and not identifying itself as an AI, leaving people to think they're actually speaking to another human...? And if 'Victor' is correct that Hong's Gourmet had no advance knowledge of the call, Google may have violated California law by recording the call." Friday he added that "This wouldn't send anyone to prison, but it would be a bit of an embarrassment, and would reinforce the notion that Google has a cavalier stance on privacy (and adhering to privacy laws)."
The Mercury News also reports that legal experts "raised questions about how Google's possible need to record Duplex's phone conversations to improve its artificial intelligence may come in conflict with California's strict two-party consent law, where all parties involved in a private phone conversation need to agree to being recorded."
For another perspective, Gizmodo's senior reviews editor reminds readers that "pretty much all tech demos are fake as hell." Speaking of Google's controversial Duplex demo, she writes that "If it didn't happen, if it is all a lie, well then I'll be totally disappointed. But I can't say I'll be surprised."
Conversation, noun:
A talk, especially an informal one, between two or more people, in which news and ideas are exchanged.
How many people were involved? Or are we already granting AI status as a person?
From all the available (I don't live in the US) law information I can find, CONVERSATION seems to be a key word in all of it.
...
Laws don't matter any more. Break 'em or not, if nobody's going to do anything about it, go right ahead. I'm getting that directly from the top.
You are welcome on my lawn.
Google may have violated California law by recording the call.
They also violated California law by placing a robocall without asking the called party for permission before letting the bot interact with the person.
Who cares.
California law (penal code 632) requires consent to record a "confidential communication". It then defines what a confidential communication is and is not.
First, in order to be a "confidential communication", the circumstances must indicate that a party desires it to be confidential between only the people involved in the conversation. In a restaurant or salon, there would have been other people around. If the employee didn't whisper or take the phone into another room, nor talk about personal private matters, they probably didn't intend it to be confidential. Indeed whatever they said to some random person calling, someone they've never met, probably isn't intended to be private.
Secondly, California laws says it's not protected when the parties "may reasonably expect that the communication may be overheard or recorded". (Note the word "may", may be overheard, not "will be overheard". Google assistant says "I'm booking an appointment for a client". It would be reasonable to expect that the client may be overhearing the conversation. People routinely use speakerphone while booking restaurant reservations or hair appointments, so again the restaurant or salon "may reasonably expect that the communication MAY be overheard".
If there was no reasonable expectation of privacy, if it wasn't a confidential communication, recording it doesn't violate the law.
Btw still I understand why some might find it a bit spooky. The submission asked if it was illegal, not if it was spooky.
It would be illegal only if the other party was trying to have a confidential discussion and thought there was little or no chance that it might be overheard.
If the noodle guy doesn't care then what's the issue? It sounds like he's nervous that Google employees will stop eating there because of that stupid blogger.
I bet they randomly selected 100 business, and got them to give permission to do this, but never told when or if they would actually ring.
That way, when actually rung, every call could be the AI - but the person answering would never know.
Therefore it sounds authentic.
Getting the permission of a a large sample means it could be anyone.
By definition, laws apply to the poor and powerless.
It was remembered by an AI. No different from you calling up the restaurant and remembering what was said. Except more reliable. And it's not any sort of a slippery slope, because what was on the other side was a business, not a person. The person there was answering AS THE BUSINESS. And if they don't want to be publicly known and recorded AS a business, without the right to privacy of a private individual, make it a private members only restaurant. You're not a business open to the public then, and you CAN decide your dealings are not with the public.
It's an AI calling. Robocalling is dialing 00001, then 00002, then 00003.... Meaning the line being called was not the intended destination of a call. Otherwise every business using voice activated screening (press 1 for sales, 2 if you want to be given the runaround, pound sign to hang up) is guilty of robocalling. As is nearly every helpdesk (the number you call is not the number you get to).
I am definitely no Google fanboy but maybe people could decide what to accuse Google of:
* First it was, that it was all a fake and people knew they were called by the AI.
* Now it is that it was no fake and people did not know they were called by the AI.
It appears rather difficult not to fall into one of the two categories.
It was a staged demonstration. If/when it turns into a real service, all they have to do is start each talk with a disclaimer like 'Hi, I'm the Google Duplex Software, this call is recorded."
Google may have violated California law by recording the call." Friday he added that "This wouldn't send anyone to prison, but it would be a bit of an embarrassment, and would reinforce the notion that Google has a cavalier stance on privacy (and adhering to privacy laws)."
Or maybe the intent of the law in question was to protect people and their privacy and Google me the intent by removing all possible identifying information about the call. They did this to such an extent that they got accused of staging the call in the first place. So what is it? Is it fake as hell, or are we violating laws by not faking it?
It's amazing how regardless of what anyone both sides of any argument will still find a way to complain about it. Seriously some people need to get a life.
Banks, insurance companies, and other big business tend to have pre-recorded "this conversation may be recorded for quality assurance and training purposes" before you're passed to a live person. Tada, permission given for google to record for QA and training of Duplex.
>unveil cool new AI technology
>feel good about yourself, everyone was impressed
>journalists jealous they can't make cool things
>journalists start framing your tech as evil and scary
>journalists start looking for ways that you've done something criminal
Fucking Jewish media, man. I'm no fan of Google but they don't deserve this.
Because it is otherwise known as "protect the politician/law enforcement/business mogul law" and the first two should be required to wear body cams with audio for all work interactions anyway.
Comment removed based on user account deletion
Google is rich enough that laws don't apply to it. This is not how it should be, but in a country where bribery has been legalized provided you maintain the fiction that it's just a "campaign donation," and there's no clearly identifiable quid being promised in consideration of a proposed specific quo, with enough money you can do whatever you want.
The republic is dead, and I for one welcome our corporate oligarch overlords, and you would too if you were smart... because we simply have no alternative.
Our reign has gone on long enough. Indeed. Summon the meteors.
What is the actual problem here?
We are increasingly going to a complaint culture not because there are more problem today than in the past but because we allow ordinary people to complain on a new scale, world-wide. We encourage them even by commenting on their complaints, even if only saying we think they are acting like children.
This leads to this complaint culture: everything that can be complained about is, as people like complaining. And this leads media to hang on to this "trend" and start complaining too. Not that they in turn haven't tried to manufacture controversies since the start of their existence.
So to answer myself: the actual problem are the complainers. Those that cover up the real problems of the world in their mind-numbing noise of petty crap.
if people spent as much time on sociopolitical issues, as they do trying to tear tech companies or strangers on the internet down?
They do. That's why we have outrage mobs and wannabe journalists who go after every perceived wrongthink with an unyielding fervor that would make facists blush and a screeching retardation that makes even 4chan take a step back. The problem is, they're only making things worse not better.
We have a one party notification law, meaning anyone is free to record the conversation without having to notify the other, as it should be. :). Freedom.
There are plenty of great software developers here..
Is everyone else here missing that the Turing Test has been at least partially solved? Google has AI that humans cannot distinguish from real people (in limited interaction).
Yes, if all parties agree with you that they can reasonably expect someone might be listening.
"You shouldn't record confidential conversations" doesn't seem that weird, does it? Or "don't record people's secrets".
How about "Don't invade people's privacy. Someone preaching on a street corner isn't talking in privacy".
Yeah any time law considers intent it can bring up questions, on the other hand my 3yo can distinguish between someone telling her a secret and someone making an announcement.
The main robocall legislation is the Telephone Consumer Protection Act, signed by President Bush in 1991. It generally applies to people trying to sell something. Here are the relevant FCC regulations:
https://apps.fcc.gov/edocs_pub...
The FTC also has some regulations based on TCPA, but again that's for solicitors.
Besides calls from solicitors, TCPA also restricts automated calls to hospitals, emergency rooms, and doctors' offices. So it may technically be illegal to use Google Assistant to schedule an appointment with a doctor. I don't think it violates the spirit of the law, though, as the law was intended to protect people from receiving unwanted, nuisance bulk calls. So long as Google Assistant does a reasonably good job, it's not a nuisance call.
I would say most demo's are staged and rehearsed. Not the least surprised and other then geeks drooling over this presentation as being real. Most people I talked too figured it was staged. Neither business answered by name, and it appeared as though it was pretty much scripted and not random calls. Probably to make the technology seem better then it really is. But that's pretty much what and demo tries to do.
Unless the woman Duplex talked to goes by a male name. If Hong's is the same restaurant, it's possible that Victor's English is as bad as the woman in the demo, and he didn't understand the reporter's questions.
When was the law written? When was that smart AI tool created?
Slashdot, fix the reply notifications... You won't get away with it...
Jail? He should be publicly tortured and executed. You think too small.
They called a restaurant.
Dude, Facebook. Nah, forget it, too.
Trump.
You might be worried about what calls AI will do next, not those ones making reservations...
Is two party consent even constitutional? I'd like to see the supreme court take up that one.
A Silicon Valley company ignores any possible ethical considerations in order to get blind stinking rich.
Also, water is wet.
Gruber has been a total Apple fanboy/whore for decades. The point is not that Google is being careless about the restaurant's privacy when Duplex called to place an order--it's hardly a state secret and nothing unflattering happened--the point is that Gruber is blatantly trying to cause trouble for Google; whereas if Apple had come up with this first, he'd be filled nothing but praise and worship. I have always disliked him for his twisting of the truth and obvious bias, and this is just another example of it.