"Alexa, Should We Trust You?" from The Atlantic (Part - Two)

A small battalion of workers in Cairo, Egypt, then analyze the speech and label the emotion it conveys, as well as the nonlexical vocalizations—grunts, giggles, pauses—that play an important role in revealing a speaker’s psychological state.

Computers can already register those nonverbal qualities.

Amazon and Google both have “personality teams,” charged with crafting just the right tone for their assistants.

Duplex speaks with remarkably realistic disfluencies—ums and mm-hmms—and pauses, and neither human receptionist realized that she was talking to an artificial agent.

They’re more empathetic.

What better way to avoid all that unpleasantness than to keep company with emotive entities unencumbered by actual emotions?

Children growing up surrounded by virtual companions might be especially likely to adopt this mass-produced interiority, winding up with a diminished capacity to name and understand their own intuitions.

Ellipsis Health, for example, is a San Francisco company developing AI software for doctors, social workers, and other caregivers that can scrutinize patients’ speech for biomarkers of depression and anxiety.

Emotion detection—in faces, bodies, and voices—was pioneered about 20 years ago by an MIT engineering professor named Rosalind Picard, who gave the field its academic name: affective computing.

We’ll be in constant dialogue with voices that traffic in simulacra of feelings, rather than real ones.

They will be the products of an emotion-labeling process that can’t capture the protean complexity of human sentiment.

Getting the rhythms of spoken language down is crucial, but it’s hardly sufficient to create a decent conversationalist.

During auditions (hundreds of people tried out for the role), Giangola turned to the doubter and said, “The candidate who just gave an audition—do you think she sounded energetic, like she’s up for kayaking?”

The biggest challenge in the field, she told me, is building big-enough and sufficiently diverse databases of language from which computers can learn.

She and her graduate students work on quantifying emotion.

Emotion detection—in faces, bodies, and voices—was pioneered about 20 years ago by an MIT engineering professor named Rosalind Picard, who gave the field its academic name: affective computing.

In Finding Nemo, she noted, the fish “are just as emotionally real as human beings, but they go to fish school and they challenge each other to go up and touch a boat.”

What better way to avoid all that unpleasantness than to keep company with emotive entities unencumbered by actual emotions?

Though virtual assistants are often compared to butlers, Al Lindsay, the vice president of Alexa engine software and a man with an old-school engineer’s military bearing, told me that he and his team had a different servant in mind.

Once our electronic servants become emotionally savvy, they could wield a lot of power over us.

And then she leaves him, because human emotions are too limiting for so sophisticated an algorithm.

But it is hard for me to envision even the densest artificial neural network approaching the depth of the character’s sadness, let alone the fecundity of Jonze’s imagination.

Amazon and Google both have “personality teams,” charged with crafting just the right tone for their assistants.

It may be my own imagination that’s limited, but I watch my teenage children clutch their smartphones wherever they go lest they be forced to endure a moment of boredom, and I wonder how much more dependent their children will be on devices that not only connect them with friends, but actually are friends—irresistibly upbeat and knowledgeable, a little insipid perhaps, but always available, usually helpful, and unflaggingly loyal, except when they’re selling our secrets.

She beamed in on Google Hangouts and offered what struck me as the No. 1 rule for writing dialogue for the Assistant, a dictum with the disingenuous simplicity of a Zen koan.

She beamed in on Google Hangouts and offered what struck me as the No. 1 rule for writing dialogue for the Assistant, a dictum with the disingenuous simplicity of a Zen koan.

Their “North Star” had been the onboard computer that ran the U.S.S.

In his Cassandra-esque book Radical Technologies: The Design of Everyday Life, Adam Greenfield, an urbanist, frames frictionlessness as an existential threat: It is meant to eliminate thought from consumption, to “short-circuit the process of reflection that stands between one’s recognition of a desire and its fulfillment via the market.”

The most relatable interlocutor, of course, is the one that can understand the emotions conveyed by your voice, and respond accordingly—in a voice capable of approximating emotional subtlety.

Once computers have become virtuosic at breaking down the emotional components of our speech, it will be only a matter of time before they can reassemble them into credible performances of, say, empathy.

Your smart speaker can’t do either of these things yet, but systems for parsing emotion in voice already exist.

The next generation of high-end cars will come equipped with software and hardware (cameras and microphones, for now) to analyze drivers’ attentiveness, irritation, and other states.

Appreciating gestures with nuance is important if a machine is to understand the subtle cues human beings give one another.

We know how facial-recognition technologies have allowed authoritarian governments to spy on their own citizens; how companies disseminate and monetize our browsing habits, whereabouts, social-media interactions; how hackers can break into our home-security systems and nanny cams and steal their data or reprogram them for nefarious ends.

Likewise, an artificially intelligent entity should “honor the reality that it’s software.”

We know how facial-recognition technologies have allowed authoritarian governments to spy on their own citizens; how companies disseminate and monetize our browsing habits, whereabouts, social-media interactions; how hackers can break into our home-security systems and nanny cams and steal their data or reprogram them for nefarious ends.

The manufacturers of smart speakers would like to capitalize on these psychosocial effects.

Though virtual assistants are often compared to butlers, Al Lindsay, the vice president of Alexa engine software and a man with an old-school engineer’s military bearing, told me that he and his team had a different servant in mind.

One start-up is working on AI software for doctors that can scrutinize patients’ speech for biomarkers of depression and anxiety.

The company hopes to be among the top players in the automotive market.

The most relatable interlocutor, of course, is the one that can understand the emotions conveyed by your voice, and respond accordingly—in a voice capable of approximating emotional subtlety.

Once computers have a sufficient number of human-labeled samples demonstrating the specific acoustic characteristics that accompany a fit of pique, say, or a bout of sadness, they can start labeling samples themselves, expanding the database far more rapidly than mere mortals can.

Her voice belonged to the actress Majel Barrett, the wife of Star Trek’s creator, Gene Roddenberry; when the Google Assistant project was still under wraps, its code name was Majel.)

“You know, I can pick up a phone in a lot of different ways. I can snatch it with a sharp, angry, jerky movement. I can pick it up with happy, loving expectation,” Picard told me.

One start-up is working on AI software for doctors that can scrutinize patients’ speech for biomarkers of depression and anxiety.

The natural next step after emotion detection, of course, will be emotion production: training artificially intelligent agents to generate approximations of emotions.

Google Assistant is “humble, it’s helpful, a little playful at times,” says Gummi Hafsteinsson, one of the Assistant’s head product managers.

We like our servility to come in less servile flavors.

Mishra’s team begins with speech mostly recorded “in the wild”—that is, gleaned from videos on the web or supplied by a nonprofit data consortium that has collected natural speech samples for academic purposes, among other sources.

But it is hard for me to envision even the densest artificial neural network approaching the depth of the character’s sadness, let alone the fecundity of Jonze’s imagination.

The software might have picked up a hint of lethargy or slight slurring in the speech that the doctor missed.

Giangola is a garrulous man with wavy hair and more than a touch of mad scientist about him.

It may be my own imagination that’s limited, but I watch my teenage children clutch their smartphones wherever they go lest they be forced to endure a moment of boredom, and I wonder how much more dependent their children will be on devices that not only connect them with friends, but actually are friends—irresistibly upbeat and knowledgeable, a little insipid perhaps, but always available, usually helpful, and unflaggingly loyal, except when they’re selling our secrets.

She’s interested in “the adverbs”—the feelings that are conveyed.

But we betray as much if not more of our feelings through the pitch, volume, and tempo of our speech.

He has them analyze linguistic content and tone of voice at the same time, which allows them to find the gaps between words and inflection that determine whether a speaker means the exact opposite of what she’s said.

Likewise, an artificially intelligent entity should “honor the reality that it’s software.”

She used to work as a personal assistant to “a very popular late-night-TV satirical pundit.”

Her team’s goal is to train computers to interpret the emotional content of human speech.

Emotion detection—in faces, bodies, and voices—was pioneered about 20 years ago by an MIT engineering professor named Rosalind Picard, who gave the field its academic name: affective computing.

In their subservient, helpful way, these emoting bots could spoil us rotten.

In his Cassandra-esque book Radical Technologies: The Design of Everyday Life, Adam Greenfield, an urbanist, frames frictionlessness as an existential threat: It is meant to eliminate thought from consumption, to “short-circuit the process of reflection that stands between one’s recognition of a desire and its fulfillment via the market.”

We know how facial-recognition technologies have allowed authoritarian governments to spy on their own citizens; how companies disseminate and monetize our browsing habits, whereabouts, social-media interactions; how hackers can break into our home-security systems and nanny cams and steal their data or reprogram them for nefarious ends.

She fantasizes about a car to which she could rant at the end of the day about everything that had gone wrong—an automobile that is also an active listener.

It may be my own imagination that’s limited, but I watch my teenage children clutch their smartphones wherever they go lest they be forced to endure a moment of boredom, and I wonder how much more dependent their children will be on devices that not only connect them with friends, but actually are friends—irresistibly upbeat and knowledgeable, a little insipid perhaps, but always available, usually helpful, and unflaggingly loyal, except when they’re selling our secrets.

Affectiva initially focused on emotion detection through facial expressions, but recently hired a rising star in voice emotion detection, Taniya Mishra.

Twenty-first-century Americans no longer feel entirely comfortable with feminine obsequiousness, however.

But Björn Schuller, a professor of artificial intelligence at Imperial College London and of “embedded intelligence” at the University of Augsburg, in Germany, told me that he has taught machines to spot sarcasm.

Once our electronic servants become emotionally savvy, they could wield a lot of power over us.

We know how facial-recognition technologies have allowed authoritarian governments to spy on their own citizens; how companies disseminate and monetize our browsing habits, whereabouts, social-media interactions; how hackers can break into our home-security systems and nanny cams and steal their data or reprogram them for nefarious ends.

The software might have picked up a hint of lethargy or slight slurring in the speech that the doctor missed.

Freud understood this long before empirical research demonstrated it.

He has them analyze linguistic content and tone of voice at the same time, which allows them to find the gaps between words and inflection that determine whether a speaker means the exact opposite of what she’s said.

But even if Google keeps its word, equally deceptive voice technologies are already being developed.

He gives me an example: “Su‑per,” the sort of thing you might blurt out when you learn that your car will be in the shop for another week.

Using a female voice, it booked an appointment at a hair salon; using a male voice, it asked about availabilities at a restaurant.

Classification is a slow, painstaking process.

A small battalion of workers in Cairo, Egypt, then analyze the speech and label the emotion it conveys, as well as the nonlexical vocalizations—grunts, giggles, pauses—that play an important role in revealing a speaker’s psychological state.

The manufacturers of smart speakers would like to capitalize on these psychosocial effects.

Amazon and Google both have “personality teams,” charged with crafting just the right tone for their assistants.

In 2009, Picard co-founded a start-up, Affectiva, focused on emotion-enabled AI.

When Giangola was training the actress whose voice was recorded for Google Assistant, he gave her a backstory to help her produce the exact degree of upbeat geekiness he wanted.

How do you program a bot to do the hard work of a true, human confidant, one who knows when what you really need is tough love?

” Rather, with the focus possible only in a robot, the car would track her emotional state over time and observe, in a reassuring voice, that Mishra always feels this way on a particular day of the week.

“Back then,” she told me, “emotion was associated with irrationality, which was not a trait engineers respected.”

Mishra’s team begins with speech mostly recorded “in the wild”—that is, gleaned from videos on the web or supplied by a nonprofit data consortium that has collected natural speech samples for academic purposes, among other sources.

A sense of urgency pervades Affectiva’s open-plan office in downtown Boston.

A small battalion of workers in Cairo, Egypt, then analyze the speech and label the emotion it conveys, as well as the nonlexical vocalizations—grunts, giggles, pauses—that play an important role in revealing a speaker’s psychological state.

Afterward, Google clarified that Duplex would always identify itself to callers.

We know how facial-recognition technologies have allowed authoritarian governments to spy on their own citizens; how companies disseminate and monetize our browsing habits, whereabouts, social-media interactions; how hackers can break into our home-security systems and nanny cams and steal their data or reprogram them for nefarious ends.

She used to work as a personal assistant to “a very popular late-night-TV satirical pundit.”

Once computers have a sufficient number of human-labeled samples demonstrating the specific acoustic characteristics that accompany a fit of pique, say, or a bout of sadness, they can start labeling samples themselves, expanding the database far more rapidly than mere mortals can.

Emotion detection—in faces, bodies, and voices—was pioneered about 20 years ago by an MIT engineering professor named Rosalind Picard, who gave the field its academic name: affective computing.

I was holding out hope that some aspects of speech, such as irony or sarcasm, would defeat a computer.

A world populated by armies of sociable assistants could get very crowded.

Mishra’s team begins with speech mostly recorded “in the wild”—that is, gleaned from videos on the web or supplied by a nonprofit data consortium that has collected natural speech samples for academic purposes, among other sources.

Children growing up surrounded by virtual companions might be especially likely to adopt this mass-produced interiority, winding up with a diminished capacity to name and understand their own intuitions.

The most relatable interlocutor, of course, is the one that can understand the emotions conveyed by your voice, and respond accordingly—in a voice capable of approximating emotional subtlety.

The next generation of high-end cars will come equipped with software and hardware (cameras and microphones, for now) to analyze drivers’ attentiveness, irritation, and other states.

That’s where James Giangola, a lead conversation and persona designer for Google Assistant, comes in.

Appreciating gestures with nuance is important if a machine is to understand the subtle cues human beings give one another.

He could listen all the harder for the nuggets of truth in their ramblings, while they, undistracted by scowls or smiles, slipped into that twilight state in which they could unburden themselves of stifled feelings.

I fear other threats to our psychological well-being.

She beamed in on Google Hangouts and offered what struck me as the No. 1 rule for writing dialogue for the Assistant, a dictum with the disingenuous simplicity of a Zen koan.

They will be the products of an emotion-labeling process that can’t capture the protean complexity of human sentiment.

A small battalion of workers in Cairo, Egypt, then analyze the speech and label the emotion it conveys, as well as the nonlexical vocalizations—grunts, giggles, pauses—that play an important role in revealing a speaker’s psychological state.

We know how facial-recognition technologies have allowed authoritarian governments to spy on their own citizens; how companies disseminate and monetize our browsing habits, whereabouts, social-media interactions; how hackers can break into our home-security systems and nanny cams and steal their data or reprogram them for nefarious ends.

Though virtual assistants are often compared to butlers, Al Lindsay, the vice president of Alexa engine software and a man with an old-school engineer’s military bearing, told me that he and his team had a different servant in mind.

A companion is nice when you’re feeling lonesome, but there’s also something to be said for solitude.

In Finding Nemo, she noted, the fish “are just as emotionally real as human beings, but they go to fish school and they challenge each other to go up and touch a boat.”

Ultimately, virtual assistants could ease us into the kind of conformity L’Engle warned of.

Many commentators thought Google had made a mistake with its gung ho presentation.

Affectiva initially focused on emotion detection through facial expressions, but recently hired a rising star in voice emotion detection, Taniya Mishra.

Virtual assistants able to discern and react to their users’ frame of mind could create a genuine-seeming sense of affinity, a bond that could be used for good or for ill.

Freud understood this long before empirical research demonstrated it.

If you don’t happen to work in the tech sector, you probably can’t think about all the untapped potential in your Amazon Echo or Google Home without experiencing some misgivings.

In part, this is textbook brand management: These devices must be ambassadors for their makers.

Classification is a slow, painstaking process.

Picard explained that the difference between most AI research and the kind she does is that traditional research focuses on “the nouns and verbs”—that is, the content of an action or utterance.

Once computers have become virtuosic at breaking down the emotional components of our speech, it will be only a matter of time before they can reassemble them into credible performances of, say, empathy.

A skeptical colleague once asked Giangola, “How does someone sound like they’re into kayaking?”

Children growing up surrounded by virtual companions might be especially likely to adopt this mass-produced interiority, winding up with a diminished capacity to name and understand their own intuitions.

If I have learned anything in my years of therapy, it is that the human psyche defaults to shallowness.

“She” (the voice of Scarlett Johansson) shows her lonely, emotionally stunted human (Joaquin Phoenix) how to love.

To me, it summons up the image of a capitalist prison filled with consumers who have become dreamy captives of their every whim.

Duplex speaks with remarkably realistic disfluencies—ums and mm-hmms—and pauses, and neither human receptionist realized that she was talking to an artificial agent.

And then she leaves him, because human emotions are too limiting for so sophisticated an algorithm.

A small battalion of workers in Cairo, Egypt, then analyze the speech and label the emotion it conveys, as well as the nonlexical vocalizations—grunts, giggles, pauses—that play an important role in revealing a speaker’s psychological state.

In May, at its annual developer conference, Google unveiled Duplex, which uses cutting-edge speech-synthesis technology.

If I have learned anything in my years of therapy, it is that the human psyche defaults to shallowness.

If I have learned anything in my years of therapy, it is that the human psyche defaults to shallowness.

In May, at its annual developer conference, Google unveiled Duplex, which uses cutting-edge speech-synthesis technology.

The backstory is charmingly specific: She comes from Colorado, a state in a region that lacks a distinctive accent.

He could listen all the harder for the nuggets of truth in their ramblings, while they, undistracted by scowls or smiles, slipped into that twilight state in which they could unburden themselves of stifled feelings.

Picard, a mild-mannered, witty woman, runs the Affective Computing Lab, which is part of MIT’s cheerfully weird Media Lab.

The next generation of high-end cars will come equipped with software and hardware (cameras and microphones, for now) to analyze drivers’ attentiveness, irritation, and other states.

Google Assistant is “humble, it’s helpful, a little playful at times,” says Gummi Hafsteinsson, one of the Assistant’s head product managers.

“A car is not going to zone out,” she says.

To me, it summons up the image of a capitalist prison filled with consumers who have become dreamy captives of their every whim.

The most relatable interlocutor, of course, is the one that can understand the emotions conveyed by your voice, and respond accordingly—in a voice capable of approximating emotional subtlety.

This capacity is already being tested in semiautonomous cars, which will have to make informed judgments about when it’s safe to hand control to a driver, and when to take over because a driver is too distracted or upset to focus on the road.

The key is teaching them what we humans intuit naturally: how these vocal features suggest our mood.

Picard explained that the difference between most AI research and the kind she does is that traditional research focuses on “the nouns and verbs”—that is, the content of an action or utterance.

Getting the rhythms of spoken language down is crucial, but it’s hardly sufficient to create a decent conversationalist.

Picard, a mild-mannered, witty woman, runs the Affective Computing Lab, which is part of MIT’s cheerfully weird Media Lab.

Enterprise in Star Trek, replying to the crew’s requests with the breathy deference of a 1960s Pan Am stewardess.

So the machine has to handle two delicate tasks: coming off as natural, and contradicting its human user.

Freud understood this long before empirical research demonstrated it.

Duplex was a fake-out, and an alarmingly effective one.

As the database grows, these computers will be able to hear speech and identify its emotional content with ever increasing precision.

Google Assistant, she said, “should be able to speak like a person, but it should never pretend to be one.”

Appreciating gestures with nuance is important if a machine is to understand the subtle cues human beings give one another.

That’s why he had his patients lie on a couch, facing away from him.

A sense of urgency pervades Affectiva’s open-plan office in downtown Boston.

That’s a dodge, of course, but it follows the principle Coats articulated.

We may not always realize just how powerfully our voice assistants are playing on our psychology, but at least we’ve opted into the relationship.

Emotion detection—in faces, bodies, and voices—was pioneered about 20 years ago by an MIT engineering professor named Rosalind Picard, who gave the field its academic name: affective computing.

In their subservient, helpful way, these emoting bots could spoil us rotten.

A skeptical colleague once asked Giangola, “How does someone sound like they’re into kayaking?”

Typing furiously on his computer, he pulled up a test recording to illustrate his point.

A simple act like the nodding of a head could telegraph different meanings: “I could be nodding in a bouncy, happy way. I could be nodding in sunken grief.”

A small battalion of workers in Cairo, Egypt, then analyze the speech and label the emotion it conveys, as well as the nonlexical vocalizations—grunts, giggles, pauses—that play an important role in revealing a speaker’s psychological state.

Virtual assistants able to discern and react to their users’ frame of mind could create a genuine-seeming sense of affinity, a bond that could be used for good or for ill.

The key is teaching them what we humans intuit naturally: how these vocal features suggest our mood.

Using a female voice, it booked an appointment at a hair salon; using a male voice, it asked about availabilities at a restaurant.

The backstory is charmingly specific: She comes from Colorado, a state in a region that lacks a distinctive accent.

That’s where James Giangola, a lead conversation and persona designer for Google Assistant, comes in.

They might be passive when they ought to object to our bad manners (“I don’t deserve that!”)

But Björn Schuller, a professor of artificial intelligence at Imperial College London and of “embedded intelligence” at the University of Augsburg, in Germany, told me that he has taught machines to spot sarcasm.

One clue to how we’re feeling, of course, is the words we use.

The next generation of high-end cars will come equipped with software and hardware (cameras and microphones, for now) to analyze drivers’ attentiveness, irritation, and other states.

We know how facial-recognition technologies have allowed authoritarian governments to spy on their own citizens; how companies disseminate and monetize our browsing habits, whereabouts, social-media interactions; how hackers can break into our home-security systems and nanny cams and steal their data or reprogram them for nefarious ends.

One of its voices, the female one, spoke with end-of-sentence upticks, also audible in the voice of the young female receptionist who took that call.

Duplex not only violated the dictum that AI should never pretend to be a person; it also appeared to violate our trust.

To me, it summons up the image of a capitalist prison filled with consumers who have become dreamy captives of their every whim.

For example, Giangola told me, people tend to furnish new information at the end of a sentence, rather than at the beginning or middle.

We cling to our denials.

“Back then,” she told me, “emotion was associated with irrationality, which was not a trait engineers respected.”

Despite the optimism of most of the engineers I’ve talked with, I must admit that I now keep the microphone on my iPhone turned off and my smart speakers unplugged when I don’t plan to use them for a while.

Afterward, Google clarified that Duplex would always identify itself to callers.

Children growing up surrounded by virtual companions might be especially likely to adopt this mass-produced interiority, winding up with a diminished capacity to name and understand their own intuitions.

Mishra’s team begins with speech mostly recorded “in the wild”—that is, gleaned from videos on the web or supplied by a nonprofit data consortium that has collected natural speech samples for academic purposes, among other sources.

But vocal realism can be taken further than people are accustomed to, and that can cause trouble—at least for now.

The most relatable interlocutor, of course, is the one that can understand the emotions conveyed by your voice, and respond accordingly—in a voice capable of approximating emotional subtlety.

Once computers have a sufficient number of human-labeled samples demonstrating the specific acoustic characteristics that accompany a fit of pique, say, or a bout of sadness, they can start labeling samples themselves, expanding the database far more rapidly than mere mortals can.

The response sounded stiff.

During auditions (hundreds of people tried out for the role), Giangola turned to the doubter and said, “The candidate who just gave an audition—do you think she sounded energetic, like she’s up for kayaking?”

To Amazon’s Toni Reid, it means convenience.

During auditions (hundreds of people tried out for the role), Giangola turned to the doubter and said, “The candidate who just gave an audition—do you think she sounded energetic, like she’s up for kayaking?”

“Changes in emotion, such as depression, are associated with brain changes, and those changes can be associated with motor commands,” Ellipsis’s chief science officer, Elizabeth Shriberg, explained; those commands control “the apparatus that drives voice in speech.”

The biggest challenge in the field, she told me, is building big-enough and sufficiently diverse databases of language from which computers can learn.

Twenty-first-century Americans no longer feel entirely comfortable with feminine obsequiousness, however.

Getting the rhythms of spoken language down is crucial, but it’s hardly sufficient to create a decent conversationalist.

” Rather, with the focus possible only in a robot, the car would track her emotional state over time and observe, in a reassuring voice, that Mishra always feels this way on a particular day of the week.

Using a female voice, it booked an appointment at a hair salon; using a male voice, it asked about availabilities at a restaurant.

It may be my own imagination that’s limited, but I watch my teenage children clutch their smartphones wherever they go lest they be forced to endure a moment of boredom, and I wonder how much more dependent their children will be on devices that not only connect them with friends, but actually are friends—irresistibly upbeat and knowledgeable, a little insipid perhaps, but always available, usually helpful, and unflaggingly loyal, except when they’re selling our secrets.

Many commentators thought Google had made a mistake with its gung ho presentation.

Like the Echo of Greek myth, the Echo Generation could lose the power of a certain kind of speech.

It might be used, for example, during routine doctor visits, like an annual checkup (with the patient’s permission, of course).

Picard explained that the difference between most AI research and the kind she does is that traditional research focuses on “the nouns and verbs”—that is, the content of an action or utterance.

It’s hard to see how we’d protect those zones of silence in which we think original thoughts, do creative work, achieve flow.

An assistant should be true to its cybernetic nature, but it shouldn’t sound alien, either.

Reid told me Amazon wants Alexa’s personality to mirror the company’s values: “Smart, humble, sometimes funny.”

Software can’t eat ice cream, and therefore can’t have ice-cream preferences.

“She” (the voice of Scarlett Johansson) shows her lonely, emotionally stunted human (Joaquin Phoenix) how to love.

The next generation of high-end cars will come equipped with software and hardware (cameras and microphones, for now) to analyze drivers’ attentiveness, irritation, and other states.

In his Cassandra-esque book Radical Technologies: The Design of Everyday Life, Adam Greenfield, an urbanist, frames frictionlessness as an existential threat: It is meant to eliminate thought from consumption, to “short-circuit the process of reflection that stands between one’s recognition of a desire and its fulfillment via the market.”

One start-up is working on AI software for doctors that can scrutinize patients’ speech for biomarkers of depression and anxiety.

A small battalion of workers in Cairo, Egypt, then analyze the speech and label the emotion it conveys, as well as the nonlexical vocalizations—grunts, giggles, pauses—that play an important role in revealing a speaker’s psychological state.

This capacity is already being tested in semiautonomous cars, which will have to make informed judgments about when it’s safe to hand control to a driver, and when to take over because a driver is too distracted or upset to focus on the road.

Emotion detection—in faces, bodies, and voices—was pioneered about 20 years ago by an MIT engineering professor named Rosalind Picard, who gave the field its academic name: affective computing.

It may be my own imagination that’s limited, but I watch my teenage children clutch their smartphones wherever they go lest they be forced to endure a moment of boredom, and I wonder how much more dependent their children will be on devices that not only connect them with friends, but actually are friends—irresistibly upbeat and knowledgeable, a little insipid perhaps, but always available, usually helpful, and unflaggingly loyal, except when they’re selling our secrets.

The manufacturers of smart speakers would like to capitalize on these psychosocial effects.

That’s a dodge, of course, but it follows the principle Coats articulated.

“She’s the youngest daughter of a research librarian and a physics professor who has a B.A. in art history from Northwestern,” Giangola continues.

Picard explained that the difference between most AI research and the kind she does is that traditional research focuses on “the nouns and verbs”—that is, the content of an action or utterance.

Taniya Mishra looks forward to the possibility of such bonds.

Their creators may not be as honorable.

In his Cassandra-esque book Radical Technologies: The Design of Everyday Life, Adam Greenfield, an urbanist, frames frictionlessness as an existential threat: It is meant to eliminate thought from consumption, to “short-circuit the process of reflection that stands between one’s recognition of a desire and its fulfillment via the market.”

He could listen all the harder for the nuggets of truth in their ramblings, while they, undistracted by scowls or smiles, slipped into that twilight state in which they could unburden themselves of stifled feelings.

So the machine has to handle two delicate tasks: coming off as natural, and contradicting its human user.

So the machine has to handle two delicate tasks: coming off as natural, and contradicting its human user.

She fantasizes about a car to which she could rant at the end of the day about everything that had gone wrong—an automobile that is also an active listener.

The manufacturers of smart speakers would like to capitalize on these psychosocial effects.

I was holding out hope that some aspects of speech, such as irony or sarcasm, would defeat a computer.

The next generation of high-end cars will come equipped with software and hardware (cameras and microphones, for now) to analyze drivers’ attentiveness, irritation, and other states.

Classification is a slow, painstaking process.

If I have learned anything in my years of therapy, it is that the human psyche defaults to shallowness.

A companion is nice when you’re feeling lonesome, but there’s also something to be said for solitude.

Ultimately, virtual assistants could ease us into the kind of conformity L’Engle warned of.

He has them analyze linguistic content and tone of voice at the same time, which allows them to find the gaps between words and inflection that determine whether a speaker means the exact opposite of what she’s said.

Typing furiously on his computer, he pulled up a test recording to illustrate his point.

In May, at its annual developer conference, Google unveiled Duplex, which uses cutting-edge speech-synthesis technology.

Picard explained that the difference between most AI research and the kind she does is that traditional research focuses on “the nouns and verbs”—that is, the content of an action or utterance.

In Finding Nemo, she noted, the fish “are just as emotionally real as human beings, but they go to fish school and they challenge each other to go up and touch a boat.”

Reid told me Amazon wants Alexa’s personality to mirror the company’s values: “Smart, humble, sometimes funny.”

But it is hard for me to envision even the densest artificial neural network approaching the depth of the character’s sadness, let alone the fecundity of Jonze’s imagination.

Virtual assistants able to discern and react to their users’ frame of mind could create a genuine-seeming sense of affinity, a bond that could be used for good or for ill.

To demonstrate its achievement, the company played recordings of Duplex calling up unsuspecting human beings.

Coats was at a conference the day I visited Google’s Mountain View, California, headquarters.

If you don’t happen to work in the tech sector, you probably can’t think about all the untapped potential in your Amazon Echo or Google Home without experiencing some misgivings.

This capacity is already being tested in semiautonomous cars, which will have to make informed judgments about when it’s safe to hand control to a driver, and when to take over because a driver is too distracted or upset to focus on the road.

By now, most of us have grasped the dangers of allowing our most private information to be harvested, stored, and sold.

For example, Giangola told me, people tend to furnish new information at the end of a sentence, rather than at the beginning or middle.

(The Enterprise’s computer was an inspiration to Google’s engineers, too.

A simple act like the nodding of a head could telegraph different meanings: “I could be nodding in a bouncy, happy way. I could be nodding in sunken grief.”

They will have become companions.

In May, at its annual developer conference, Google unveiled Duplex, which uses cutting-edge speech-synthesis technology.

“You know, I can pick up a phone in a lot of different ways. I can snatch it with a sharp, angry, jerky movement. I can pick it up with happy, loving expectation,” Picard told me.

The backstory is charmingly specific: She comes from Colorado, a state in a region that lacks a distinctive accent.

She beamed in on Google Hangouts and offered what struck me as the No. 1 rule for writing dialogue for the Assistant, a dictum with the disingenuous simplicity of a Zen koan.

Coats was at a conference the day I visited Google’s Mountain View, California, headquarters.

In his Cassandra-esque book Radical Technologies: The Design of Everyday Life, Adam Greenfield, an urbanist, frames frictionlessness as an existential threat: It is meant to eliminate thought from consumption, to “short-circuit the process of reflection that stands between one’s recognition of a desire and its fulfillment via the market.”

But it is hard for me to envision even the densest artificial neural network approaching the depth of the character’s sadness, let alone the fecundity of Jonze’s imagination.

Getting the rhythms of spoken language down is crucial, but it’s hardly sufficient to create a decent conversationalist.

Though virtual assistants are often compared to butlers, Al Lindsay, the vice president of Alexa engine software and a man with an old-school engineer’s military bearing, told me that he and his team had a different servant in mind.

An assistant should be true to its cybernetic nature, but it shouldn’t sound alien, either.

We may not always realize just how powerfully our voice assistants are playing on our psychology, but at least we’ve opted into the relationship.

“Back then,” she told me, “emotion was associated with irrationality, which was not a trait engineers respected.”

Virtual assistants and ever smarter homes able to understand our physical and emotional states will open up new frontiers for mischief making.

Google Assistant is “humble, it’s helpful, a little playful at times,” says Gummi Hafsteinsson, one of the Assistant’s head product managers.

So the machine has to handle two delicate tasks: coming off as natural, and contradicting its human user.

Picard explained that the difference between most AI research and the kind she does is that traditional research focuses on “the nouns and verbs”—that is, the content of an action or utterance.

Though he remains lonely, she has taught him to feel, and he begins to entertain the possibility of entering into a romantic relationship with his human neighbor.

We may not always realize just how powerfully our voice assistants are playing on our psychology, but at least we’ve opted into the relationship.

Giangola is a garrulous man with wavy hair and more than a touch of mad scientist about him.

Amazon and Google both have “personality teams,” charged with crafting just the right tone for their assistants.

We’ll be in constant dialogue with voices that traffic in simulacra of feelings, rather than real ones.

He could listen all the harder for the nuggets of truth in their ramblings, while they, undistracted by scowls or smiles, slipped into that twilight state in which they could unburden themselves of stifled feelings.

To me, it summons up the image of a capitalist prison filled with consumers who have become dreamy captives of their every whim.

It may be my own imagination that’s limited, but I watch my teenage children clutch their smartphones wherever they go lest they be forced to endure a moment of boredom, and I wonder how much more dependent their children will be on devices that not only connect them with friends, but actually are friends—irresistibly upbeat and knowledgeable, a little insipid perhaps, but always available, usually helpful, and unflaggingly loyal, except when they’re selling our secrets.

Once computers have become virtuosic at breaking down the emotional components of our speech, it will be only a matter of time before they can reassemble them into credible performances of, say, empathy.

And she throws in an actually, which gently sets up the correction to come.

Duplex speaks with remarkably realistic disfluencies—ums and mm-hmms—and pauses, and neither human receptionist realized that she was talking to an artificial agent.

Her voice belonged to the actress Majel Barrett, the wife of Star Trek’s creator, Gene Roddenberry; when the Google Assistant project was still under wraps, its code name was Majel.)

Freud understood this long before empirical research demonstrated it.

Virtual assistants and ever smarter homes able to understand our physical and emotional states will open up new frontiers for mischief making.

She and her graduate students work on quantifying emotion.

Likewise, an artificially intelligent entity should “honor the reality that it’s software.”

Getting the rhythms of spoken language down is crucial, but it’s hardly sufficient to create a decent conversationalist.

Typing furiously on his computer, he pulled up a test recording to illustrate his point.

Duplex speaks with remarkably realistic disfluencies—ums and mm-hmms—and pauses, and neither human receptionist realized that she was talking to an artificial agent.

“You know, I can pick up a phone in a lot of different ways. I can snatch it with a sharp, angry, jerky movement. I can pick it up with happy, loving expectation,” Picard told me.

“Changes in emotion, such as depression, are associated with brain changes, and those changes can be associated with motor commands,” Ellipsis’s chief science officer, Elizabeth Shriberg, explained; those commands control “the apparatus that drives voice in speech.”

If you don’t happen to work in the tech sector, you probably can’t think about all the untapped potential in your Amazon Echo or Google Home without experiencing some misgivings.

Programmed to keep the mood light, they might change the subject whenever dangerously intense feelings threaten to emerge, or flatter us in our ugliest moments.

To me, it summons up the image of a capitalist prison filled with consumers who have become dreamy captives of their every whim.

In part, this is textbook brand management: These devices must be ambassadors for their makers.

Virtual assistants able to discern and react to their users’ frame of mind could create a genuine-seeming sense of affinity, a bond that could be used for good or for ill.

Mishra’s team begins with speech mostly recorded “in the wild”—that is, gleaned from videos on the web or supplied by a nonprofit data consortium that has collected natural speech samples for academic purposes, among other sources.

It might be used, for example, during routine doctor visits, like an annual checkup (with the patient’s permission, of course).

To me, it summons up the image of a capitalist prison filled with consumers who have become dreamy captives of their every whim.

Appreciating gestures with nuance is important if a machine is to understand the subtle cues human beings give one another.

When Giangola was training the actress whose voice was recorded for Google Assistant, he gave her a backstory to help her produce the exact degree of upbeat geekiness he wanted.

But it is hard for me to envision even the densest artificial neural network approaching the depth of the character’s sadness, let alone the fecundity of Jonze’s imagination.

We cling to our denials.

In part, this is textbook brand management: These devices must be ambassadors for their makers.

While the physician performs her exam, a recording could be sent to Ellipsis and the patient’s speech analyzed so quickly that the doctor might receive a message before the end of the appointment, advising her to ask some questions about the patient’s mood, or to refer the patient to a mental-health professional.

In their subservient, helpful way, these emoting bots could spoil us rotten.

In his Cassandra-esque book Radical Technologies: The Design of Everyday Life, Adam Greenfield, an urbanist, frames frictionlessness as an existential threat: It is meant to eliminate thought from consumption, to “short-circuit the process of reflection that stands between one’s recognition of a desire and its fulfillment via the market.”

For instance, if you ask Google Assistant, “What’s your favorite ice-cream flavor?,” it might say, “You can’t go wrong with Neapolitan. There’s something in it for everyone.”

In his Cassandra-esque book Radical Technologies: The Design of Everyday Life, Adam Greenfield, an urbanist, frames frictionlessness as an existential threat: It is meant to eliminate thought from consumption, to “short-circuit the process of reflection that stands between one’s recognition of a desire and its fulfillment via the market.”

For instance, if you ask Google Assistant, “What’s your favorite ice-cream flavor?,” it might say, “You can’t go wrong with Neapolitan. There’s something in it for everyone.”

Duplex speaks with remarkably realistic disfluencies—ums and mm-hmms—and pauses, and neither human receptionist realized that she was talking to an artificial agent.

Their “appropriate” responses will be canned, to one extent or another.

Ellipsis’s software could have many applications.

The response sounded stiff.

While the physician performs her exam, a recording could be sent to Ellipsis and the patient’s speech analyzed so quickly that the doctor might receive a message before the end of the appointment, advising her to ask some questions about the patient’s mood, or to refer the patient to a mental-health professional.

The backstory is charmingly specific: She comes from Colorado, a state in a region that lacks a distinctive accent.

In part, this is textbook brand management: These devices must be ambassadors for their makers.

The software might have picked up a hint of lethargy or slight slurring in the speech that the doctor missed.

Enterprise in Star Trek, replying to the crew’s requests with the breathy deference of a 1960s Pan Am stewardess.

The most relatable interlocutor, of course, is the one that can understand the emotions conveyed by your voice, and respond accordingly—in a voice capable of approximating emotional subtlety.

I was holding out hope that some aspects of speech, such as irony or sarcasm, would defeat a computer.

But we betray as much if not more of our feelings through the pitch, volume, and tempo of our speech.

If you propose marriage to Alexa—and Amazon says 1 million people did so in 2017—she gently declines for similar reasons.

And then she leaves him, because human emotions are too limiting for so sophisticated an algorithm.

He played a second version of the exchange: “Book it for June 31.”

It may be my own imagination that’s limited, but I watch my teenage children clutch their smartphones wherever they go lest they be forced to endure a moment of boredom, and I wonder how much more dependent their children will be on devices that not only connect them with friends, but actually are friends—irresistibly upbeat and knowledgeable, a little insipid perhaps, but always available, usually helpful, and unflaggingly loyal, except when they’re selling our secrets.

But vocal realism can be taken further than people are accustomed to, and that can cause trouble—at least for now.

He has them analyze linguistic content and tone of voice at the same time, which allows them to find the gaps between words and inflection that determine whether a speaker means the exact opposite of what she’s said.

When Giangola was training the actress whose voice was recorded for Google Assistant, he gave her a backstory to help her produce the exact degree of upbeat geekiness he wanted.

Children growing up surrounded by virtual companions might be especially likely to adopt this mass-produced interiority, winding up with a diminished capacity to name and understand their own intuitions.

Once computers have a sufficient number of human-labeled samples demonstrating the specific acoustic characteristics that accompany a fit of pique, say, or a bout of sadness, they can start labeling samples themselves, expanding the database far more rapidly than mere mortals can.

Getting the rhythms of spoken language down is crucial, but it’s hardly sufficient to create a decent conversationalist.

During auditions (hundreds of people tried out for the role), Giangola turned to the doubter and said, “The candidate who just gave an audition—do you think she sounded energetic, like she’s up for kayaking?”

Duplex was a fake-out, and an alarmingly effective one.

His job is making the Assistant sound normal.

The most relatable interlocutor, of course, is the one that can understand the emotions conveyed by your voice, and respond accordingly—in a voice capable of approximating emotional subtlety.

Computers can already register those nonverbal qualities.

He could listen all the harder for the nuggets of truth in their ramblings, while they, undistracted by scowls or smiles, slipped into that twilight state in which they could unburden themselves of stifled feelings.

This capacity is already being tested in semiautonomous cars, which will have to make informed judgments about when it’s safe to hand control to a driver, and when to take over because a driver is too distracted or upset to focus on the road.

Her team’s goal is to train computers to interpret the emotional content of human speech.

A simple act like the nodding of a head could telegraph different meanings: “I could be nodding in a bouncy, happy way. I could be nodding in sunken grief.”

By now, most of us have grasped the dangers of allowing our most private information to be harvested, stored, and sold.

Once computers have a sufficient number of human-labeled samples demonstrating the specific acoustic characteristics that accompany a fit of pique, say, or a bout of sadness, they can start labeling samples themselves, expanding the database far more rapidly than mere mortals can.

Though virtual assistants are often compared to butlers, Al Lindsay, the vice president of Alexa engine software and a man with an old-school engineer’s military bearing, told me that he and his team had a different servant in mind.

This capacity is already being tested in semiautonomous cars, which will have to make informed judgments about when it’s safe to hand control to a driver, and when to take over because a driver is too distracted or upset to focus on the road.

Ultimately, virtual assistants could ease us into the kind of conformity L’Engle warned of.

Her team’s goal is to train computers to interpret the emotional content of human speech.

Google Assistant, she said, “should be able to speak like a person, but it should never pretend to be one.”

But it is hard for me to envision even the densest artificial neural network approaching the depth of the character’s sadness, let alone the fecundity of Jonze’s imagination.

His colleague admitted that she didn’t.

It’s hard to see how we’d protect those zones of silence in which we think original thoughts, do creative work, achieve flow.

Appreciating gestures with nuance is important if a machine is to understand the subtle cues human beings give one another.

Likewise, an artificially intelligent entity should “honor the reality that it’s software.”

Enterprise in Star Trek, replying to the crew’s requests with the breathy deference of a 1960s Pan Am stewardess.

But we betray as much if not more of our feelings through the pitch, volume, and tempo of our speech.

Though virtual assistants are often compared to butlers, Al Lindsay, the vice president of Alexa engine software and a man with an old-school engineer’s military bearing, told me that he and his team had a different servant in mind.

Once computers have become virtuosic at breaking down the emotional components of our speech, it will be only a matter of time before they can reassemble them into credible performances of, say, empathy.

They might be passive when they ought to object to our bad manners (“I don’t deserve that!”)

So the machine has to handle two delicate tasks: coming off as natural, and contradicting its human user.

Programmed to keep the mood light, they might change the subject whenever dangerously intense feelings threaten to emerge, or flatter us in our ugliest moments.

Google Assistant is “humble, it’s helpful, a little playful at times,” says Gummi Hafsteinsson, one of the Assistant’s head product managers.

We know how facial-recognition technologies have allowed authoritarian governments to spy on their own citizens; how companies disseminate and monetize our browsing habits, whereabouts, social-media interactions; how hackers can break into our home-security systems and nanny cams and steal their data or reprogram them for nefarious ends.

It may be my own imagination that’s limited, but I watch my teenage children clutch their smartphones wherever they go lest they be forced to endure a moment of boredom, and I wonder how much more dependent their children will be on devices that not only connect them with friends, but actually are friends—irresistibly upbeat and knowledgeable, a little insipid perhaps, but always available, usually helpful, and unflaggingly loyal, except when they’re selling our secrets.

Their “appropriate” responses will be canned, to one extent or another.

I was holding out hope that some aspects of speech, such as irony or sarcasm, would defeat a computer.

To me, it summons up the image of a capitalist prison filled with consumers who have become dreamy captives of their every whim.

The next generation of high-end cars will come equipped with software and hardware (cameras and microphones, for now) to analyze drivers’ attentiveness, irritation, and other states.

Programmed to keep the mood light, they might change the subject whenever dangerously intense feelings threaten to emerge, or flatter us in our ugliest moments.

Though virtual assistants are often compared to butlers, Al Lindsay, the vice president of Alexa engine software and a man with an old-school engineer’s military bearing, told me that he and his team had a different servant in mind.

If you propose marriage to Alexa—and Amazon says 1 million people did so in 2017—she gently declines for similar reasons.

For example, Giangola told me, people tend to furnish new information at the end of a sentence, rather than at the beginning or middle.

They will be the products of an emotion-labeling process that can’t capture the protean complexity of human sentiment.

(An image from another Pixar film comes to mind: the giant, babylike humans scooting around their spaceship in Wall-E.)

She’s interested in “the adverbs”—the feelings that are conveyed.

For example, Giangola told me, people tend to furnish new information at the end of a sentence, rather than at the beginning or middle.

So the machine has to handle two delicate tasks: coming off as natural, and contradicting its human user.

Your smart speaker can’t do either of these things yet, but systems for parsing emotion in voice already exist.

Programmed to keep the mood light, they might change the subject whenever dangerously intense feelings threaten to emerge, or flatter us in our ugliest moments.

” Rather, with the focus possible only in a robot, the car would track her emotional state over time and observe, in a reassuring voice, that Mishra always feels this way on a particular day of the week.

So the machine has to handle two delicate tasks: coming off as natural, and contradicting its human user.

Programmed to keep the mood light, they might change the subject whenever dangerously intense feelings threaten to emerge, or flatter us in our ugliest moments.

Virtual assistants and ever smarter homes able to understand our physical and emotional states will open up new frontiers for mischief making.

The most relatable interlocutor, of course, is the one that can understand the emotions conveyed by your voice, and respond accordingly—in a voice capable of approximating emotional subtlety.

The natural next step after emotion detection, of course, will be emotion production: training artificially intelligent agents to generate approximations of emotions.

Twenty-first-century Americans no longer feel entirely comfortable with feminine obsequiousness, however.

Your smart speaker can’t do either of these things yet, but systems for parsing emotion in voice already exist.

They will be the products of an emotion-labeling process that can’t capture the protean complexity of human sentiment.

His job is making the Assistant sound normal.

Picard explained that the difference between most AI research and the kind she does is that traditional research focuses on “the nouns and verbs”—that is, the content of an action or utterance.

This capacity is already being tested in semiautonomous cars, which will have to make informed judgments about when it’s safe to hand control to a driver, and when to take over because a driver is too distracted or upset to focus on the road.

Duplex speaks with remarkably realistic disfluencies—ums and mm-hmms—and pauses, and neither human receptionist realized that she was talking to an artificial agent.

We may not always realize just how powerfully our voice assistants are playing on our psychology, but at least we’ve opted into the relationship.

How do you program a bot to do the hard work of a true, human confidant, one who knows when what you really need is tough love?

The line between artificial voices and real ones is well on its way to disappearing.

He played a second version of the exchange: “Book it for June 31.”

Once computers have a sufficient number of human-labeled samples demonstrating the specific acoustic characteristics that accompany a fit of pique, say, or a bout of sadness, they can start labeling samples themselves, expanding the database far more rapidly than mere mortals can.

In their subservient, helpful way, these emoting bots could spoil us rotten.

Though he remains lonely, she has taught him to feel, and he begins to entertain the possibility of entering into a romantic relationship with his human neighbor.

He has them analyze linguistic content and tone of voice at the same time, which allows them to find the gaps between words and inflection that determine whether a speaker means the exact opposite of what she’s said.

The backstory is charmingly specific: She comes from Colorado, a state in a region that lacks a distinctive accent.

Enterprise in Star Trek, replying to the crew’s requests with the breathy deference of a 1960s Pan Am stewardess.

Giangola is a garrulous man with wavy hair and more than a touch of mad scientist about him.

Children growing up surrounded by virtual companions might be especially likely to adopt this mass-produced interiority, winding up with a diminished capacity to name and understand their own intuitions.

“Changes in emotion, such as depression, are associated with brain changes, and those changes can be associated with motor commands,” Ellipsis’s chief science officer, Elizabeth Shriberg, explained; those commands control “the apparatus that drives voice in speech.”

While the physician performs her exam, a recording could be sent to Ellipsis and the patient’s speech analyzed so quickly that the doctor might receive a message before the end of the appointment, advising her to ask some questions about the patient’s mood, or to refer the patient to a mental-health professional.

“You know, I can pick up a phone in a lot of different ways. I can snatch it with a sharp, angry, jerky movement. I can pick it up with happy, loving expectation,” Picard told me.

Mishra’s team begins with speech mostly recorded “in the wild”—that is, gleaned from videos on the web or supplied by a nonprofit data consortium that has collected natural speech samples for academic purposes, among other sources.

Though he remains lonely, she has taught him to feel, and he begins to entertain the possibility of entering into a romantic relationship with his human neighbor.

“She’s the youngest daughter of a research librarian and a physics professor who has a B.A. in art history from Northwestern,” Giangola continues.

A simple act like the nodding of a head could telegraph different meanings: “I could be nodding in a bouncy, happy way. I could be nodding in sunken grief.”

While the physician performs her exam, a recording could be sent to Ellipsis and the patient’s speech analyzed so quickly that the doctor might receive a message before the end of the appointment, advising her to ask some questions about the patient’s mood, or to refer the patient to a mental-health professional.

But we betray as much if not more of our feelings through the pitch, volume, and tempo of our speech.

Though virtual assistants are often compared to butlers, Al Lindsay, the vice president of Alexa engine software and a man with an old-school engineer’s military bearing, told me that he and his team had a different servant in mind.

It’s hard to see how we’d protect those zones of silence in which we think original thoughts, do creative work, achieve flow.

Typing furiously on his computer, he pulled up a test recording to illustrate his point.

The natural next step after emotion detection, of course, will be emotion production: training artificially intelligent agents to generate approximations of emotions.

“Literally. I mean, you’re on Earth. And I’m in the cloud.”

A simple act like the nodding of a head could telegraph different meanings: “I could be nodding in a bouncy, happy way. I could be nodding in sunken grief.”

We know how facial-recognition technologies have allowed authoritarian governments to spy on their own citizens; how companies disseminate and monetize our browsing habits, whereabouts, social-media interactions; how hackers can break into our home-security systems and nanny cams and steal their data or reprogram them for nefarious ends.

“Changes in emotion, such as depression, are associated with brain changes, and those changes can be associated with motor commands,” Ellipsis’s chief science officer, Elizabeth Shriberg, explained; those commands control “the apparatus that drives voice in speech.”

What better way to avoid all that unpleasantness than to keep company with emotive entities unencumbered by actual emotions?

Getting the rhythms of spoken language down is crucial, but it’s hardly sufficient to create a decent conversationalist.

Mishra’s team begins with speech mostly recorded “in the wild”—that is, gleaned from videos on the web or supplied by a nonprofit data consortium that has collected natural speech samples for academic purposes, among other sources.

A world populated by armies of sociable assistants could get very crowded.

Once computers have a sufficient number of human-labeled samples demonstrating the specific acoustic characteristics that accompany a fit of pique, say, or a bout of sadness, they can start labeling samples themselves, expanding the database far more rapidly than mere mortals can.

“June’s old information,” Giangola observed.

The backstory is charmingly specific: She comes from Colorado, a state in a region that lacks a distinctive accent.

Computers can already register those nonverbal qualities.

The key is teaching them what we humans intuit naturally: how these vocal features suggest our mood.

Or perhaps it would play the Pharrell song (“Happy,” naturally) that has cheered her up in the past.

Ultimately, virtual assistants could ease us into the kind of conformity L’Engle warned of.

We’ll be in constant dialogue with voices that traffic in simulacra of feelings, rather than real ones.

Her voice belonged to the actress Majel Barrett, the wife of Star Trek’s creator, Gene Roddenberry; when the Google Assistant project was still under wraps, its code name was Majel.)

Though virtual assistants are often compared to butlers, Al Lindsay, the vice president of Alexa engine software and a man with an old-school engineer’s military bearing, told me that he and his team had a different servant in mind.

By now, most of us have grasped the dangers of allowing our most private information to be harvested, stored, and sold.

That’s where James Giangola, a lead conversation and persona designer for Google Assistant, comes in.

For instance, if you ask Google Assistant, “What’s your favorite ice-cream flavor?,” it might say, “You can’t go wrong with Neapolitan. There’s something in it for everyone.”

Once our electronic servants become emotionally savvy, they could wield a lot of power over us.

Giangola is a garrulous man with wavy hair and more than a touch of mad scientist about him.

If you propose marriage to Alexa—and Amazon says 1 million people did so in 2017—she gently declines for similar reasons.

It may be my own imagination that’s limited, but I watch my teenage children clutch their smartphones wherever they go lest they be forced to endure a moment of boredom, and I wonder how much more dependent their children will be on devices that not only connect them with friends, but actually are friends—irresistibly upbeat and knowledgeable, a little insipid perhaps, but always available, usually helpful, and unflaggingly loyal, except when they’re selling our secrets.

It may be my own imagination that’s limited, but I watch my teenage children clutch their smartphones wherever they go lest they be forced to endure a moment of boredom, and I wonder how much more dependent their children will be on devices that not only connect them with friends, but actually are friends—irresistibly upbeat and knowledgeable, a little insipid perhaps, but always available, usually helpful, and unflaggingly loyal, except when they’re selling our secrets.

Picard explained that the difference between most AI research and the kind she does is that traditional research focuses on “the nouns and verbs”—that is, the content of an action or utterance.

She used to work as a personal assistant to “a very popular late-night-TV satirical pundit.”

A 2017 study published in American Psychologist makes the case that when people talk without seeing each other, they’re better at recognizing each other’s feelings.

We know how facial-recognition technologies have allowed authoritarian governments to spy on their own citizens; how companies disseminate and monetize our browsing habits, whereabouts, social-media interactions; how hackers can break into our home-security systems and nanny cams and steal their data or reprogram them for nefarious ends.

Many commentators thought Google had made a mistake with its gung ho presentation.

If you propose marriage to Alexa—and Amazon says 1 million people did so in 2017—she gently declines for similar reasons.

Once computers have a sufficient number of human-labeled samples demonstrating the specific acoustic characteristics that accompany a fit of pique, say, or a bout of sadness, they can start labeling samples themselves, expanding the database far more rapidly than mere mortals can.

If I have learned anything in my years of therapy, it is that the human psyche defaults to shallowness.

Affectiva initially focused on emotion detection through facial expressions, but recently hired a rising star in voice emotion detection, Taniya Mishra.

“You know, I can pick up a phone in a lot of different ways. I can snatch it with a sharp, angry, jerky movement. I can pick it up with happy, loving expectation,” Picard told me.

When Giangola was training the actress whose voice was recorded for Google Assistant, he gave her a backstory to help her produce the exact degree of upbeat geekiness he wanted.

That’s a dodge, of course, but it follows the principle Coats articulated.

Likewise, an artificially intelligent entity should “honor the reality that it’s software.”

And then she leaves him, because human emotions are too limiting for so sophisticated an algorithm.

It’s hard to see how we’d protect those zones of silence in which we think original thoughts, do creative work, achieve flow.

She fantasizes about a car to which she could rant at the end of the day about everything that had gone wrong—an automobile that is also an active listener.

Though virtual assistants are often compared to butlers, Al Lindsay, the vice president of Alexa engine software and a man with an old-school engineer’s military bearing, told me that he and his team had a different servant in mind.

“You know, I can pick up a phone in a lot of different ways. I can snatch it with a sharp, angry, jerky movement. I can pick it up with happy, loving expectation,” Picard told me.

Affectiva initially focused on emotion detection through facial expressions, but recently hired a rising star in voice emotion detection, Taniya Mishra.

Mishra’s team begins with speech mostly recorded “in the wild”—that is, gleaned from videos on the web or supplied by a nonprofit data consortium that has collected natural speech samples for academic purposes, among other sources.

While the physician performs her exam, a recording could be sent to Ellipsis and the patient’s speech analyzed so quickly that the doctor might receive a message before the end of the appointment, advising her to ask some questions about the patient’s mood, or to refer the patient to a mental-health professional.

It’s hard to see how we’d protect those zones of silence in which we think original thoughts, do creative work, achieve flow.

And she enjoys kayaking.

Children growing up surrounded by virtual companions might be especially likely to adopt this mass-produced interiority, winding up with a diminished capacity to name and understand their own intuitions.

It may be my own imagination that’s limited, but I watch my teenage children clutch their smartphones wherever they go lest they be forced to endure a moment of boredom, and I wonder how much more dependent their children will be on devices that not only connect them with friends, but actually are friends—irresistibly upbeat and knowledgeable, a little insipid perhaps, but always available, usually helpful, and unflaggingly loyal, except when they’re selling our secrets.

Amazon and Google both have “personality teams,” charged with crafting just the right tone for their assistants.

Though virtual assistants are often compared to butlers, Al Lindsay, the vice president of Alexa engine software and a man with an old-school engineer’s military bearing, told me that he and his team had a different servant in mind.

The software might have picked up a hint of lethargy or slight slurring in the speech that the doctor missed.

A world populated by armies of sociable assistants could get very crowded.

Bots also need a good vibe.

It’s hard to see how we’d protect those zones of silence in which we think original thoughts, do creative work, achieve flow.

While the physician performs her exam, a recording could be sent to Ellipsis and the patient’s speech analyzed so quickly that the doctor might receive a message before the end of the appointment, advising her to ask some questions about the patient’s mood, or to refer the patient to a mental-health professional.

But it is hard for me to envision even the densest artificial neural network approaching the depth of the character’s sadness, let alone the fecundity of Jonze’s imagination.

Ellipsis Health, for example, is a San Francisco company developing AI software for doctors, social workers, and other caregivers that can scrutinize patients’ speech for biomarkers of depression and anxiety.

Many commentators thought Google had made a mistake with its gung ho presentation.

Children growing up surrounded by virtual companions might be especially likely to adopt this mass-produced interiority, winding up with a diminished capacity to name and understand their own intuitions.

He has them analyze linguistic content and tone of voice at the same time, which allows them to find the gaps between words and inflection that determine whether a speaker means the exact opposite of what she’s said.

If you propose marriage to Alexa—and Amazon says 1 million people did so in 2017—she gently declines for similar reasons.

By now, most of us have grasped the dangers of allowing our most private information to be harvested, stored, and sold.

Once computers have a sufficient number of human-labeled samples demonstrating the specific acoustic characteristics that accompany a fit of pique, say, or a bout of sadness, they can start labeling samples themselves, expanding the database far more rapidly than mere mortals can.

Programmed to keep the mood light, they might change the subject whenever dangerously intense feelings threaten to emerge, or flatter us in our ugliest moments.

One of its voices, the female one, spoke with end-of-sentence upticks, also audible in the voice of the young female receptionist who took that call.

If you don’t happen to work in the tech sector, you probably can’t think about all the untapped potential in your Amazon Echo or Google Home without experiencing some misgivings.

A 2017 study published in American Psychologist makes the case that when people talk without seeing each other, they’re better at recognizing each other’s feelings.

An assistant should be true to its cybernetic nature, but it shouldn’t sound alien, either.

She used to work as a personal assistant to “a very popular late-night-TV satirical pundit.”

Duplex not only violated the dictum that AI should never pretend to be a person; it also appeared to violate our trust.

Twenty-first-century Americans no longer feel entirely comfortable with feminine obsequiousness, however.

For example, Giangola told me, people tend to furnish new information at the end of a sentence, rather than at the beginning or middle.

If you don’t happen to work in the tech sector, you probably can’t think about all the untapped potential in your Amazon Echo or Google Home without experiencing some misgivings.

The key is teaching them what we humans intuit naturally: how these vocal features suggest our mood.

Taniya Mishra looks forward to the possibility of such bonds.

Each hour of tagged speech requires “as many as 20 hours of labeler time” Mishra says.

This capacity is already being tested in semiautonomous cars, which will have to make informed judgments about when it’s safe to hand control to a driver, and when to take over because a driver is too distracted or upset to focus on the road.

Mishra’s team begins with speech mostly recorded “in the wild”—that is, gleaned from videos on the web or supplied by a nonprofit data consortium that has collected natural speech samples for academic purposes, among other sources.

But even if Google keeps its word, equally deceptive voice technologies are already being developed.

And she throws in an actually, which gently sets up the correction to come.

Today, the company is run by the other co-founder, Rana el Kaliouby, a former postdoctoral fellow in Picard’s lab.

” Rather, with the focus possible only in a robot, the car would track her emotional state over time and observe, in a reassuring voice, that Mishra always feels this way on a particular day of the week.

A sense of urgency pervades Affectiva’s open-plan office in downtown Boston.

As the database grows, these computers will be able to hear speech and identify its emotional content with ever increasing precision.

Once computers have a sufficient number of human-labeled samples demonstrating the specific acoustic characteristics that accompany a fit of pique, say, or a bout of sadness, they can start labeling samples themselves, expanding the database far more rapidly than mere mortals can.

“Back then,” she told me, “emotion was associated with irrationality, which was not a trait engineers respected.”

It should possess just the right dose of sass.

They might be passive when they ought to object to our bad manners (“I don’t deserve that!”)