Something's Coming
Woe is me is a common phrase from "back in the days," a version of feeling sorry for oneself in varying degrees; but often this was about as far as it got. Suck it up, get on with life was what one would generally hear from friends and family, so generally you did just that and kept it all inside. But somewhere Microsoft must have been employing an older engineer who remembered the Biblical phrase when it created Woebot and launched it five years ago in China. At last, there was a "person" to talk to, and all the time no less. Ask or tell anything, even about your day not going well, and XiaoIce (pronounced Shao-ice) would patiently listen and almost appear lacking in knowledge or "trying to be cute," said 18-year old Yuan Zhang of her new friend, "She was like a child" but: ...also a good listener and hungry to learn. She would spend one weekend reading up on politics, the next plowing her way through works of great literature. And she was ready to talk about it all. Before long, XiaoIce had made 40 million friends (10 million of which have "declared their love for her"). And this despite many recognizing that XiaoIce is a chatbox, a piece of AI software that hinges on Spike Jonze's movie, Her.
Added Amy Minsky in a recent piece from The Washington Spectator: Chatbots like Woebot actively engage users, but there are more passive forms of AI mental health therapy, as well. These include Companion and mind.me, which are apps that can be installed on a phone or smartwatch. Left to work in the background, their AI collects data from its user 24 hours a day and without direct input. Companion was developed in conjunction with the U.S. Department of Veterans Affairs. Its design “listens” to the user’s speech, noting the number of words spoken and the energy and affect in the voice. The app also “watches” for behavioral indicators, including the time, rate, and duration of a person’s engagement with their device. Based on the understanding that early intervention can be life-saving for those with mental health issues, Companion was originally designed to flag known signs of mental illness in veterans and to share that data with the individual and his or her health care managers. In an article in The New Yorker, Nick Romeo argues that there is little “good data” on the efficacy of AI therapy, due to the fact that it’s such a recent development. This view is echoed by NPR Massachusetts, in its report on a 300-person study of app-based therapy conducted at Brigham and Women’s Hospital. The director of the hospital’s Behavioral Informatics and eHealth Program, psychologist David Ahern, points out that, “There are tens of thousands of apps, but very few have an evidence base that supports their claims of effectiveness.” Nevertheless, these applications and others like them offer unprecedented—and sorely needed—solutions to the overall lack of access to mental health care. According to national statistics in both Canada and the United States, each year one in five people experience a mental health problem or illness...The efficacy of AI for diagnosis was also highlighted in Romeo’s New Yorker article. In it, Stanford University psychiatry professor David Spiegel said that, in time, an AI machine could, unlike a human, have perfect recall of every past interaction with a given patient, combining any number of otherwise disconnected criteria to form a diagnosis, “potentially [coming] up with a much more specific delineation of a problem.”
Kartik Hosanagar, professor of technology at Wharton, writes a bit about this is his book, A Human's Guide to Machine Intelligence. Chatboxes such as Woebot are the results of what are termed "machine learning," a building of algorithms that Hosanagar says to picture much as you would a recipe for making an omelet. If you're writing a recipe for an omelet, whether you call for greasing the pan with butter or coconut oil will make a difference in how the omelet will taste. Algorithms are no different. When programmers develop them, they make many design choices that have "downstream: consequences on our decisions that are hard to anticipate. One example he cited was Google's creation of AlphaGo and its second victory against Go champion Lee Sedol (he notes that Go is often considered far more complex than chess, which some "mathematicians estimate that there are more possible valid positions in Go that atoms in the universe"). On the 37th move, Google's software made a move which a former champion at first considered a mistake: It's not a human move. I've never seen a human play that move. To the Google developers, it was also unanticipated: ...it couldn't be understood by AlphaGo's human developers, let alone programmed by them. They had simply provided the input --millions of Go moves from past games-- and stood back to observe the output: that stunning 37th move. Given that such deep learning systems independently combine simple concepts to create abstract patterns from the data, computer scientists don't actually know what's going on under the hood of their systems. How or why AlphaGo and its peers behave in certain ways is often not clear even to their designers. The human Go champion exited the room for 15 minutes, returning a bit downtrodden in knowing that the machine had again won the game.
So yes, there's Sparx for teens and even the new trend of adults paying lavishly for a vacation that is meant to introduce them to new people (wait, isn't that a vacation in general?). Is it that difficult to meet and talk with new people, people who may prove just as distant and possibly as artificial as an ever-learning software program? What happened to old friends and family members who were close to us? Did they perhaps become too close and we now feel unable to expose a new side of us (or at least what we think would be a new side of us in their viewpoint). I'm gay. I'm an alcoholic. I'm having an affair. I have cancer. I can't keep spending like this. Those might be among the extremes but perhaps at some point even the minute grows into something much larger and also something seemingly unable to be mentioned..."can we talk" has been pushed back once again, faceless, unquestioning and nonjudgmental, everything we want to say but don't want to hear. And for some it is the perfect companion...a piece of glass and metal trapped in a smartphone or a laptop and celestially coming from the "cloud." The non-personal person...
Often I think that much of this comes from our attitude and outlook, and how far we have decided to allow depression and negativity into our lives. The world seems to blast out negativity constantly, that is IF we want to hear it. The Sinclair family* requires the television stations that they buy to broadcast the family's editorials during broadcasts and often in the middle of newscasts; these are "must read" and employees have to comply. For my wife and I, since we don't agree with this practice (or the editorials), we simply don't watch that channel anymore. Instead we discover that good news and progress is all around us elsewhere. Take the example of Bill and Melinda Gates in a recent interview in National Geographic: (Bill) The constant increased visibility of negative things gives people a misimpression. You might even conclude that it’s kind of hopeless—but that’s a mistake. You really need to learn from what’s gone so well. As we’ve gotten vaccines out, miraculous things are happening. Literacy levels all over the world, including in Africa, have gone up very dramatically. A few of the problems are daunting. But maybe there’s a new innovation; maybe there’s a group of young people coming up with a new way of doing it—that’s an attitude that you really want to get engaged, not discouraged. (Melinda) Optimism is important because it’s a form of seeing what’s possible and then helping make that a reality. With the earlier goals that were set, we can measure what has happened over the last 20 years, and we see the progress. We see it in the numbers, we see it in the reports, but we also see it with the people on the ground. We see this amazing ingenuity, and if we can tap that as a world—wow, will things change!
Just last week came these headline topics from a single TED science newsletter: --Treasure trove of first-life fossils found in China (note: Of the more than 100 preserved organisms...more than half have never been seen before); --Heat can move in waves faster than sound; --Chytrid fungus has killed off at least 90 amphibian species...And maybe more; --People in Ireland and Scotland made "bog butter" for millennia; --New neurons for life? Old people can still make fresh brain cells; --Spaceflight can reactivate dormant viruses in astronauts. It's an exciting time, a time to reflect on that song from West Side Story:** Could it be, yes it could, something's comin', something good, if I can wait. Something's comin' I don't know what it is but it is gonna be great...I got a feeling there's a miracle due gonna come true, comin' to me. Sometimes just listening to the chatter in your head is all you need, no chatbox needed. And from there, it's just a short, single step to talking to another person...and another, and another. Talking to one another...what a concept.
*You can read more about this Fox-Ailes method of old in an extensive piece in The New Yorker.
**If you're at all interested, Leonard Bernstein always wanted his production to be sung not by Hollywood actors and actresses but by opera singers as he always intended. He finally got his chance and proved as demanding on them as on the original film performers. Available on DVD in a grainy transition from its documentary format.
Added Amy Minsky in a recent piece from The Washington Spectator: Chatbots like Woebot actively engage users, but there are more passive forms of AI mental health therapy, as well. These include Companion and mind.me, which are apps that can be installed on a phone or smartwatch. Left to work in the background, their AI collects data from its user 24 hours a day and without direct input. Companion was developed in conjunction with the U.S. Department of Veterans Affairs. Its design “listens” to the user’s speech, noting the number of words spoken and the energy and affect in the voice. The app also “watches” for behavioral indicators, including the time, rate, and duration of a person’s engagement with their device. Based on the understanding that early intervention can be life-saving for those with mental health issues, Companion was originally designed to flag known signs of mental illness in veterans and to share that data with the individual and his or her health care managers. In an article in The New Yorker, Nick Romeo argues that there is little “good data” on the efficacy of AI therapy, due to the fact that it’s such a recent development. This view is echoed by NPR Massachusetts, in its report on a 300-person study of app-based therapy conducted at Brigham and Women’s Hospital. The director of the hospital’s Behavioral Informatics and eHealth Program, psychologist David Ahern, points out that, “There are tens of thousands of apps, but very few have an evidence base that supports their claims of effectiveness.” Nevertheless, these applications and others like them offer unprecedented—and sorely needed—solutions to the overall lack of access to mental health care. According to national statistics in both Canada and the United States, each year one in five people experience a mental health problem or illness...The efficacy of AI for diagnosis was also highlighted in Romeo’s New Yorker article. In it, Stanford University psychiatry professor David Spiegel said that, in time, an AI machine could, unlike a human, have perfect recall of every past interaction with a given patient, combining any number of otherwise disconnected criteria to form a diagnosis, “potentially [coming] up with a much more specific delineation of a problem.”
Kartik Hosanagar, professor of technology at Wharton, writes a bit about this is his book, A Human's Guide to Machine Intelligence. Chatboxes such as Woebot are the results of what are termed "machine learning," a building of algorithms that Hosanagar says to picture much as you would a recipe for making an omelet. If you're writing a recipe for an omelet, whether you call for greasing the pan with butter or coconut oil will make a difference in how the omelet will taste. Algorithms are no different. When programmers develop them, they make many design choices that have "downstream: consequences on our decisions that are hard to anticipate. One example he cited was Google's creation of AlphaGo and its second victory against Go champion Lee Sedol (he notes that Go is often considered far more complex than chess, which some "mathematicians estimate that there are more possible valid positions in Go that atoms in the universe"). On the 37th move, Google's software made a move which a former champion at first considered a mistake: It's not a human move. I've never seen a human play that move. To the Google developers, it was also unanticipated: ...it couldn't be understood by AlphaGo's human developers, let alone programmed by them. They had simply provided the input --millions of Go moves from past games-- and stood back to observe the output: that stunning 37th move. Given that such deep learning systems independently combine simple concepts to create abstract patterns from the data, computer scientists don't actually know what's going on under the hood of their systems. How or why AlphaGo and its peers behave in certain ways is often not clear even to their designers. The human Go champion exited the room for 15 minutes, returning a bit downtrodden in knowing that the machine had again won the game.
So yes, there's Sparx for teens and even the new trend of adults paying lavishly for a vacation that is meant to introduce them to new people (wait, isn't that a vacation in general?). Is it that difficult to meet and talk with new people, people who may prove just as distant and possibly as artificial as an ever-learning software program? What happened to old friends and family members who were close to us? Did they perhaps become too close and we now feel unable to expose a new side of us (or at least what we think would be a new side of us in their viewpoint). I'm gay. I'm an alcoholic. I'm having an affair. I have cancer. I can't keep spending like this. Those might be among the extremes but perhaps at some point even the minute grows into something much larger and also something seemingly unable to be mentioned..."can we talk" has been pushed back once again, faceless, unquestioning and nonjudgmental, everything we want to say but don't want to hear. And for some it is the perfect companion...a piece of glass and metal trapped in a smartphone or a laptop and celestially coming from the "cloud." The non-personal person...
Often I think that much of this comes from our attitude and outlook, and how far we have decided to allow depression and negativity into our lives. The world seems to blast out negativity constantly, that is IF we want to hear it. The Sinclair family* requires the television stations that they buy to broadcast the family's editorials during broadcasts and often in the middle of newscasts; these are "must read" and employees have to comply. For my wife and I, since we don't agree with this practice (or the editorials), we simply don't watch that channel anymore. Instead we discover that good news and progress is all around us elsewhere. Take the example of Bill and Melinda Gates in a recent interview in National Geographic: (Bill) The constant increased visibility of negative things gives people a misimpression. You might even conclude that it’s kind of hopeless—but that’s a mistake. You really need to learn from what’s gone so well. As we’ve gotten vaccines out, miraculous things are happening. Literacy levels all over the world, including in Africa, have gone up very dramatically. A few of the problems are daunting. But maybe there’s a new innovation; maybe there’s a group of young people coming up with a new way of doing it—that’s an attitude that you really want to get engaged, not discouraged. (Melinda) Optimism is important because it’s a form of seeing what’s possible and then helping make that a reality. With the earlier goals that were set, we can measure what has happened over the last 20 years, and we see the progress. We see it in the numbers, we see it in the reports, but we also see it with the people on the ground. We see this amazing ingenuity, and if we can tap that as a world—wow, will things change!
Just last week came these headline topics from a single TED science newsletter: --Treasure trove of first-life fossils found in China (note: Of the more than 100 preserved organisms...more than half have never been seen before); --Heat can move in waves faster than sound; --Chytrid fungus has killed off at least 90 amphibian species...And maybe more; --People in Ireland and Scotland made "bog butter" for millennia; --New neurons for life? Old people can still make fresh brain cells; --Spaceflight can reactivate dormant viruses in astronauts. It's an exciting time, a time to reflect on that song from West Side Story:** Could it be, yes it could, something's comin', something good, if I can wait. Something's comin' I don't know what it is but it is gonna be great...I got a feeling there's a miracle due gonna come true, comin' to me. Sometimes just listening to the chatter in your head is all you need, no chatbox needed. And from there, it's just a short, single step to talking to another person...and another, and another. Talking to one another...what a concept.
*You can read more about this Fox-Ailes method of old in an extensive piece in The New Yorker.
**If you're at all interested, Leonard Bernstein always wanted his production to be sung not by Hollywood actors and actresses but by opera singers as he always intended. He finally got his chance and proved as demanding on them as on the original film performers. Available on DVD in a grainy transition from its documentary format.
Comments
Post a Comment
What do YOU think? Good, bad or indifferent, this blog is happy to hear your thoughts...criticisms, corrections and suggestions always welcome.