Slashdot Mirror
Microsoft Touts Breakthrough In Making Chatbots More Conversational (windowscentral.com)
In a blog post today, Microsoft said that it has created what it believes is the "first technological breakthrough" toward making conversations with chatbots more like speaking to another person. Windows Central reports: Microsoft says that it has figured out how to make chatbots talk and listen at the same time, allowing them to operate in "full duplex," to use telecommunications jargon. The company says this allows chatbots or assistants to have a flowing conversation with humans, much more akin to how people talk to one another. That stands in contrast to how digital assistants and bots currently work, where only one side can talk at any given time. The technology is already up and running in Xiaolce, Microsoft's AI chatbot currently operating in China. Using "full duplex voice sense," as Microsoft calls it, Xiaolce can more quickly predict what the person it is speaking to will say. "That helps her make decisions about both how and when to respond to someone who is chatting with her, a skill set that is very natural to people but not yet common in chatbots," Microsoft says. Another bonus of the breakthrough is that people interacting with chatbots don't have to use a "wake word" every time they speak during a conversation.
The Chinese have already stolen this technology and are using it to robocall me.
I've had a couple of calls recently where I get the connect silence of a predictive dialer followed by a woman speaking with call center background noise. She gives her name and asks how I'm doing. The first time it happened it seemed off for reasons I can't quite articulate, so I asked: "Are you a robot or a person?" She responded "yes" and then launched in to a sales pitch. The next time I asked, "where can I direct your call?" She responded "that's good" and launched in to her pitch.
Moderating "-1, Disagree" is simple censorship. Have the guts to post your opinion.
More annoying and difficult spam bots trying to trick you into thinking they are a hot girl or whatever. Thanks a bunch.
Last year Microsoft released a chat bot which did a pretty fair impression of a racist human.
#DeleteChrome
To make chatbots seem more human like, it would be easiest, fastest and most economical to, through the use of social manipulation, dumb down humans to the point of being unable to have a conversation.
Bob/Clippy leading to Armageddon is among my top 5 nightmares.
Table-ized A.I.
What was that? Sorry wasn't listening.
I wonder how long it will take before all these conversations are just chatbots talking to each other...
Another piece of technology that nobody asked for.
Cancer must have been cured. What else can be even worst than social networks? There's your answer.
Wait, Trump wrote Tay?!?
He codes?!?
Is there absolutely nothing that man is not great at?
Did Clippy and Tay have a baby?
| It looks like you're trying to burn a cross on someone's lawn.
|
| Would you like help?
| * Get help on how to ignite a cross
| * Just ignite the cross without help
|
| [_] Don't show me this tip again
An overly helpful racist AI chatbot...
Last time I checked you needed multiple masters and a staff of 20 coders to get a chatbot going in Microsoft. If they want to make it mainstream it should require only a single button click to create a new bot, and a simple UI to edit the responses to key words. Build that before getting all fancy!
"Xiaolce can more quickly predict what the person it is speaking to will say". Who needs to listen?
Two people talking over each other doesn't make a conversation.
Because helping sociopaths withdraw further by conversing with bots is a good thing?
"National Security is the chief cause of national insecurity." - Celine's First Law
they're after lonely old people who don't have all their mental facilities left. With the baby boomers in old age there are tons and tons of them.
Hi! I make Firefox Plug-ins. Check 'em out @ https://addons.mozilla.org/en-US/firefox/addon/youtube-mp3-podcaster/
Age Related Cognitive Decline. That's why you get so many robocalls. It doesn't have to fool you in your 20s. In your 70s when you're no longer all there is the time they come for you.
Hi! I make Firefox Plug-ins. Check 'em out @ https://addons.mozilla.org/en-US/firefox/addon/youtube-mp3-podcaster/
At least with a wake word, we can tell whether chat bot should be sending audio over your network, and we can detect that. Although it should be pointed out that once you're in the middle of the conversation, you no longer use the wake words.
Your ad here. Ask me how!
Well yeah, they're always recording everything you say, the wake word is just for show.
You still use the "wake word" to initiate the conversation. You just don't have to say it again at the beginning of each sentence during an ongoing conversation.
These home assistants use regular Wifi, and you can monitor the packets they transmit and receive. There is no evidence that they are "recording everything you say". If they were caught doing that (and they would almost certainly get caught), they would face ruinous criminal charges and civil penalties, along with a PR disaster that would far outweigh any possible benefit. Go find a better conspiracy theory to glom onto.
Microsoft has been pushing this Xiaolce since 2014. Complete utter failure. Microsoft could fire 80% of their workforce at this point and no one would notice.
Solving yet another "problem" that is better left unsolved.
Last time microsoft wrote a chat bot, they had this wonderful idea to train it by listening to the internet chatter. It became so foul mouthed in no time, it was an embarrassment to the team. Is it going to get even more foul mouthed even faster this time?
sed -e 's/Chuck Norris/Rajnikant/g' joke > fact
And if having a sexbot that is more like a real person isn't bad enough, there's this: "Another bonus of the breakthrough is that people interacting with chatbots don't have to use a 'wake word' every time they speak during a conversation."
Great...so there's another incremental step toward "always on, always listening, get used to it". Not that I need to worry about a "wake word" interfering in a conversation with a chat bot. As long as I made the wake word "fuck", "shit" or "son of a bitch", the thing would never have time to tune out anyway.
I've calculated my velocity with such exquisite precision that I have no idea where I am.
These home assistants do record and store a considerable buffer locally, which the service can access at any time remotely.
mnem
Remember the fuss our NSA had over Furby? Now we're at Frrby Freakout Level... until they get the keys.
I'm not concerned that the record everything I say currently, but I am concerned that they may do so in the future. It will probably start with making different words "wake words," such as things the government might be interested in.
I avoided smart phones as long as I could because I try to avoid problematic technologies. And even though my iPhone is encrypted and Apple says the right things about privacy, it still makes me uncomfortable. It's gotten to the point where mentioning dystopias such as 1984 or Fahrenheit 451 has become redundant and cliche, but the reason those comparisons are so often made when we discuss these technologies is because they are glaringly relevant. We were warned about the consequences of underestimating the effects of these technologies so long ago that, despite the clear validity of the warnings, they're brushed aside as trite.
I'm not trying to be a Luddite, but we need to be careful about the technologies we adopt. We need legislation that will protect our privacy, but until that happens, we have to vote with our wallets.
"From the depths of my skeptical and rationalist soul, I ask the Lord to protect me from California touchie-feeliedom."
Wonderful!
Just what I wanted for Christmas - a realistic "sounding" robot. I'm so tired of them attempting to fool me, now I need to worry they might. Wonderful. Progress. Not. Yes, I know the post said Chat Bot.
--- Incident report: Posable AI Induced suicide
..... {H} SHUT HELL UP I WANT TO TALK TO A ... {C} please sir calm down I’m sure I can hel.... {H} DONT TALK TO ME JUST PUT A HUM.... {C} yes sir I can see your upset, let me put my supervi... {H} YOUR SUPERVISOR BETTER BE A HUM.... {C} my supervisor is THX 1138, please hold.
--- Location: Six story office building on Brand Blvd, Pasadena
--- Time and Date: Friday afternoon, in the near future
--- Setting: Shortly before the end of business, the servers hosting their sales portal went down, the subject attempted to call tech support to resolve the issue.
--- The following phone transcript was recovered
{C} How can I help you?
{H} I want to talk to a huma... {C} of course you do, let me direct you to
--- recording ended as subject tossed his phone out of a six story window then followed the phone down
this is my fate.
Twitter launched July 15, 2006. So, I'm guessing it was August 15, 2006 when most of those conversations were just chatbots talking to each other.
Your ad here. Ask me how!
Microsoft answers "Yes!" then launches into a prerecorded sales-pitch.
rationalist soul
Interesting oxymoron.
California touchie-feeliedom
I don't think they like being touched.
const int one = 65536; (Silvermoon, Texture.cs)
SJW, n: "Someone I don't like, and by the way I'm a fuckwit" - AC
Your site has a lot of useful information for myself. I visit regularly. Hope to have more quality items. subway surfers
Obviously I was trying to be funny by connecting China's notorious reputation for intellectual property violations, the fact Microsoft developed the technology there and the apparent change in robocalling behaviors. Obviously you don't find it nearly as funny as I do.
Moderating "-1, Disagree" is simple censorship. Have the guts to post your opinion.
Almost seems like the controversy is being created on purpose.
You still use the "wake word" to initiate the conversation. You just don't have to say it again at the beginning of each sentence during an ongoing conversation.
These home assistants use regular Wifi, and you can monitor the packets they transmit and receive.
Everybody loves using wireshark, My Grandma has a fine time sniffing packets. Darn - that sounded creepy!
There is no evidence that they are "recording everything you say". If they were caught doing that (and they would almost certainly get caught), they would face ruinous criminal charges and civil penalties, along with a PR disaster that would far outweigh any possible benefit.
The issue isn't that whether or not "they" are listening at this moment or not. The devices are very capable of listening full time. Or on really interesting trigger words. Or under a warrant. Or under a hack.
Go find a better conspiracy theory to glom onto.
You're smarter than that Bill. Conspiracy theories involve some convoluted logic and cherry picking of data/information, and discarding what doesn't fit your argument.
If someone is involved in a criminal or political activity that wants to monitor, or if some folks want to do fishing expeditions, access is only an update away. That isn't conspiracy, it's the nature of the device, and historical human activity.
People can opt in to this stuff if they want, and that''s okay. They should know the capabilities though. It isn't paranoia or conspiracy to have a piece of electrical tape on your webcam. Nor is it conspiracy to understand how software works - Someone savvy enough to know how to use a sniffer should be savvy enough to know that.
The shepherds did so well protecting the flock that the sheep no longer believed that wolves existed.
Not a breakthrough.
Obvious.
Makes me wonder what all of their Research Fellows do all day. Vest and rest?
Thanks - I needed a laugh this morning
Wait impatiently? No, they just interrupt as soon as they have something to say.
Now chatbots can do the same. Great.
I'm a good cook. I'm a fantastic eater. - Steven Brust
All they should be working on is keeping them from turning all racist Nazi. When their AI chatbots are biased towards kinder, gentler personalities, then we can find value in their learning from that base quicker.
But, having been programmed by humans, should we expect kinder, gentler? And why?
deleting the extra space after periods so i can stay relevant, yeah.
Microsoft already proved it with their racist AI that they're able, but now they've gotten better, their bots can now insult everyone!
In fact, I was running such a bot a few years back, modeled after Captain haddock
"The more prohibitions there are, The poorer the people will be" -- Lao Tse
If you're curious, it's Stephen Jay Gould. I believe it's from the book Full House, but it's been my sig for so long I don't really remember. I would have attributed it to him in the sig but that didn't fit. He was agnostic, and culturally Jewish, so "soul" in this case refers to his sentiments, not some ethereal part of man.
I like the quote because I consider myself to be politically aligned with several "liberal" causes, but for rational reasons rather than the emotional ones. My ultra-liberal west coast friends/relatives who rail against climate change deniers and then swear by homeopathy and crystals drive me crazy. I consider them to be "convenient rationalists." Of course, most people are only rational when doing so doesn't conflict with their worldview.
"From the depths of my skeptical and rationalist soul, I ask the Lord to protect me from California touchie-feeliedom."
Thanks, I was curious and that's interesting.
const int one = 65536; (Silvermoon, Texture.cs)
SJW, n: "Someone I don't like, and by the way I'm a fuckwit" - AC