“Understanding natural language is such a complex and difficult task, systems like IBM Debater lean on a wide range of systems, each handling a different part of the problem. One system will identify information that helps fuel an argument on one side of the debate. Another will generate the text of the argument. And so on.”
“But understanding language is far more complex. That means systems that perform fairly complex language tasks – like writing a Wikipedia article, let alone engaging in a serious debate on a random topic – may still be light years away.”
Let the debates begin – man vs. machine!
P.S. Sorry about the late posting today.
Microsoft is now trying on the role as moral leader in a tech world now facing increasing public criticism. I think you all know who they are, so let’s just say that Facebook is leading the pack. Microsoft is the only one to avoid sustained public criticism about contributing to the social ills of the last couple of years..
At the same time, Microsoft has emerged as one of the most outspoken advocates in the industry for protecting user privacy and establishing ethical guidelines for new technology like artificial intelligence (A.I.). They have launched a new program. A.I. for Accessibility that will award $25 million over five years to researchers, nonprofits and developers who use artificial intelligence to help people with disabilities.
Echoing a theme he talked about at Microsoft’s conference last year, Satya Nadella, its chief, said that the industry has a responsibility to build technology that empowered everyone.
Artificial Intelligence (A.I.) that is. A recent Gallup survey revealed that the vast majority of Americans expect A.I. to lead to joblessness in the coming decade, but few see it as coming to their own position. “Whether they know it or not, A.I. has moved into a big percent of Americans’ lives in one way or another already (Newport, Northeastern University, 2018).”
“Personal assistants” like Google Home or Amazon’s Alexa, as well as navigational apps, such as Google Maps, Waze and Apple Maps are used most widely among younger and more educated Americans. More than 90 percent of adults with at least a bachelor’s degree or between the ages of 18 and 35 used navigation apps, for example. Gallup’s report reflects only some of the findings of a large survey of nearly 3,300 Americans conducted in September an October of last year. The other findings, released in January, show that more than three in four Americans believe that A.I. will fundamentally change how the public lives and works in the coming decade.
About the same share expect A.I. to destroy more jobs than it creates, though only about one in four were worried about losing their own job?
So it’s only a game as they say, but the geopolitical implications seem obvious. This board game is called Go and I have seen it played in parks around Hanoi, but please don’t ask me to explain it. But I will quote from a article by a Hong Kong reporter that might help shed some light. “Go, in which two players vie for control of a board using black and white pieces called stones, is considered complex because of the sheer number of possible moves. Even supercomputers cannot simply calculate all the possible moves, presenting a big challenge for AlphaGo creators.” But AlphaGo developers did accept the challenge and created the software that makes this game available online.
So far, AlphaGo seems to be the undisputed “artificial intelligence” champion, only being beaten once by South Korea”s Mr. Lee. China’s Mr. Ke seems more resigned to only playing against human opponents. He noted that he would focus more on playing with people saying that the gap between humans was becoming too great. He would treat the software as more of a teacher, he said, to get inspiration and new ideas about moves. Or maybe he should say that he has finally met his match, but when his “match” is basically artficial intelligence, it just may be too hard to admit defeat by a software program? Somehow this all sounds vaguely familiar, like Dr. Frankenstein being outsmarted by his own “monstrous” creation.
AlphaGo is also demonstrating an ability to learn from its gaming experiences. It is not just calculating moves, but learning from its own experiences. That is something that we can all benefit from, so that we can remain smarter than our machines, I hope.
P.S. Happy Memorial Day weekend. Be back on the 31st.
Outsourcing knowledge to Google keeps you away from learning things the right way. Don’t take my word for it. Psychological researchers have been studying the effects on internet dependence on the human learning process. Take your ability to remember, or learn things the right way so that you can recall them at will. And on a personal note, this seems to get harder as you get older. So if you want to stay younger mentally, using Google may be a handy tool, but still keep using your own mental faculties if you want to have people think you really know what you are talking about. How old is Donald Trump? Seventy? He seems to like Twitter better than Google, but he still might like to use it if he wants to fact-check something. I just don’t think he worries about those bothersome facts that much. He does use the TV to watch FOX news, right?
“Using knowledge in the head is also self-sustaining, whereas using knowledge from the internet is not. Every time you retrieve information from memory, it becomes a bit easier to find it the next time. That’s why students studying for a test actually remember more if they quiz themselves than if they study as they typically do, by rereading their textbook or notes. That parades the right ideas before the mind, but it doesn’t make them stick in the same way, you won’t learn your way around a city if you always use your GPS, but you will if you work to remember the route you took last time (NY Times, 5/21/17).
“But why do I worry about all this? And why does Donald Trump come creeping back into my mind. Maybe it is the fact that he is not the “fake President.”
What’s all this about fake news? (remember Rosenne Rosanadana, TV’s Saturday Night Live). I just read the other day that computer experts are using sophisticated algorithms and online data to spot misinformation. So now it seems that we have “machine learning” tools that use artificial intelligence to combat fake news. A growing number of technology experts worldwide are now harnessing their skills to tackle misinformation online. Calls for combating fake news have focused on some of the biggest online players, including American giants like Facebook and Google. Why did we have to wait for this call until after a U.S. Presidential election? Does fake news really have more readers than real news?
I am not really a conspiracy theorist, and maybe the technology and needed algorithms were not fully developed in time for last year’s election, but perhaps it’s just another example of “timing being everything.” Technology still seems to hold a revered place as our best hope for discerning fact from fiction. But many Europeans are not so optimistic. With fake news already swirling around their forthcoming elections, analysts also worry that technology on its on may not be enough to combat the threat.
Remember the old adage, “All I know is what I read in the newspapers.” I guess it’s time to rethink that old saw, or maybe we should literally start reading (and listening) again with a more critical perspective. We must never think that technology can do all this for us.
Sad, but true. Now I don’t know if Confucius ever said that, but it seems that many American inventors and entrepreneurs developing innovative technologies for the U.S. military are finding more support from Chinese investors than from the Pentagon. For example, Neurala, a Boston start-up that makes robots and drones got little response from the American military when it needed money. But it landed an investment from a state-run Chinese company.
Beijing is encouraging Chinese companies with close government ties to invest in American start-ups specializing in critical technologies like artificial intelligence and robots to advance China’s military capacity as well as its economy. The size and breadth of these Chinese-U.S. deals are not clear because start-ups and their backers are not obligated to disclose them. Over all, China has been increasingly active in the American start-up world, investing $9.9 billion in 2015. Chinese investors have money and are looking for returns, while the Chinese government has pushed investment in ways to clean up China’s skies, upgrade its industrial capacity and unclog its snarled highways.
I bet that Donald Trump could personally help Neurala and other American technology start-ups, and make America Great Again. Save U.S. budget dollars by NOT flying his whole family around the world and NOT going to Mar-a-Lago every weekend. Maybe he too could begin investing in America’s future, just like the Chinese!
Please allow me to explain. One of the things that artificial intelligence (A.I.) can do with relative ease seems to be writing music that we all have probably heard as background music for commercials on TV, or in elevators, etc.? It would likely be a gentle piano piece. Its melody is simple, and is unsubtle in its melancholy and tone. What do you expect? It was written by a machine.
Most people working on A.I. have focused on classical music, but many are convinced that composing a short, catchy melody is probably the more difficult task. A compelling song is actually a rare and fragile object. It can only work if all the dimensions are right: the melody, the harmony, the voice, the dress of the singer, the discourse around it – like, “Why did someone write this song?” No one is able to model all that right now, but many are interested in the problem. Some rock groups are experimenting with computer-generated music, but do not foresee their listeners accepting it entirely. “Music fans need to fall in love with musicians. You can’t fall in love with a computer.”
We all obviously have a choice in what we like to listen to, computer-generated or something more “human.” Even with all the contributions that the computer can make in creating music, some still observe that they “don’t know why they want to make music. They don’t have any goal, any desire.”
Now for those of you who like to watch real-time video of rapes, suicides and murders, this blog is very much written to cramp your “style.” I suggest you find some other healthy pastimes to fill your idle hours, and if you cannot, go get some real mental health help. Unfortunately, if you truly do have such an obsession, I am sure you have already found where you can get your fill of such content on some of our most popular “social media” sites. And to make your quest even easier, they are literally at your finger tips as live-stream apps downloaded to whatever handy device you may choose.
As I mentioned in a blog last week, there are some “human” efforts to monitor social media in the advertisement and sale of guns on the Internet. Unfortunately to date it is a very labor-intensive effort requiring constant surveillance by concerned citizens who volunteer their time and energy to “watchdog” this market. In the case of scanning video for for inappropriate content, artificial intelligence (AI) may be the answer to identifying and removing such content in a more efficient way. Software has already been developed that can easily be used to weed out pornography or violence. And the speed of AI’s recognition give it a huge leg up on human monitors.
Perhaps the biggest challenge in monitoring inappropriate content is that it’s all in the “eye of the beholder” as suggested by one privacy expert. But let’s not fall for that old bromide. Rape, suicide, murder?
Now we all know that Siri is a very smart piece of artificial intelligence (AI), but have you heard about Viv? She (I think that’s the voice gender, or maybe you can choose) is even smarter. Her creators like to call her a virtual assistant more than simple piece of AI. For the past four years the same people who brought you Siri have been working on creating Viv. She who will do more than just give you requested information. Using simple voice commands, she will actually do things for you, like ordering a pizza delivered to your doorstep. Please don’t ask me why ordering a pizza is the most critical task in assessing the performance of a virtual assistant. I guess you just get hungry putting all those pieces of code together?
In all fairness, Viv has moved on from ordering pizza. She can order a car, flowers, turn lights off and on in your home, and is talking with other potential partners to unite many businesses into a single, unbroken conversation: television companies, car companies, media companies, and makers of smart refrigerators, etc. Forget all those apps, unless you really like living in an earlier technological era.
It’s the latest and easily the biggest time-saver in our busy and more connected world. So what are you going to do with all that free time? Looks like you will have more time to do things in the real world.
Let technology do it for you. Now you can go to your new improved Google calendar and find that “Goals” has scheduled all those things you have to do everyday. “Goals” is the name of the software that offers you a menu of goals to choose from on a daily basis. It can then scour the white spaces in your calendar for available times and will map out a schedule. Of course you will still be in charge of deciding if this is all okay with you and let “Goals” make all the final arrangements. Why not?
I don’t know about you, but all of this does scare me in my “old age.” I really do have more spare time these days, and I would hate to have it all filled up by a virtual assistant of some kind. I can still pretty much plan my calendar from day to day, and more often think in terms of month to month at a more leisurely pace. And we can all thank artificial intelligence (A.I.) for making this all possible. Yes, that’s right, your new personal assistant is really a robot after all. And a very hard-working one at that. He/she (your choice) can even sort your email into high and low priority items and weed out the junk. Unfortunately, this appears to have freed up time for people to send more work email!
So be careful what you wish for, or in this case, what you really never wished for, but Google decided you needed. Sherry Turkle at M.I.T. worries that “human” software agents can diminish authentic human interactions. I don’t think robots ever worry about that!
Need someone to go shopping with? Well maybe that’s not done so much any more, and you probably don’t need someone after all if you have Chatbot. And you really don’t need a salesperson to help you since you will be online, and Chatbot will be there programmed to answer all of your anticipated questions. At this rate there may not be any brick and mortar stores (malls?) to stroll through in the near future. What’s going to happen to all of those gigantic shopping malls and parking garages? As more shopping goes online, maybe they will morph into “distribution centers” where your purchases are shipped to your home, saving you from the inconvenience of having to go to a store to shop.
This is the future that many social media entrepreneurs are banking on. “Facebook said it was opening up Messenger, it’s own messaging app, so that any outside company – from Applebee’s to Zara – could create a bot capable of interacting with people through the chat program.” And Facebook may be the logical and most profitable place to begin. It already has 900 million regular monthly users of Messenger with more than 15 million businesses having an official brand page on Facebook.
So this is shopping in the virtual world. But don’t forget that you still need a home with a real address to have everything shipped to. Or maybe you can just get a really big post office box? After you buy all that stuff online, you may have to invest in a bigger home that will truly become your castle!
They really do need us. Just when I was starting to worry that Artificial Intelligence (AI) would become so pervasive that it would eclipse our limited human capacities, I now read that the tech behemoths (Apple, Amazon, Microsoft) are now hiring real people to make these virtual assistants sound more human. Microsoft, for example, has hired poets, novelists, playwrights and former television writers to be members of their writing teams. So technology may help us become more literate after all, or at least will provide more employment opportunities for those who are literate. Oh, I forgot to include comedians in the list of desired writing team members. That makes me happy!
This is very good news for job seekers in the U.S. By 2025, 12.7 million new U.S. Jobs will involve building robots or automation software; by 2019, more than one-third of the workforce will be working side by side with such technologies. That’s only three years away (Forester Data). What a change in the work place! In the old days you just had to worry about getting along (or not) with the humans you worked with. Now you will have to get along with your “artificial” colleagues as well. I’m not sure that they will become supervisory personnel in the future, but I would guess that many of you are thinking that you have already worked for some real “bots” in the past.
Some early studies of human-robot interaction have found that attempts to make robots seem more humanlike can inspire unease or revulsion instead of empathy. Maybe we all just have to learn to “get along,” but I don’t know if robots can be programmed for that?
Don’t start counting on receiving monthly checks from the government quite yet. The basic idea here seems to be that as robots begin doing more of the work traditionally done by humans in our economy and jobs dry up, we should ensure that American workers’ wages will not be negatively impacted. How about $1,000 a month? And on top of that, we would all be free to become artists, scholars, entrepreneurs or otherwise engage our passions in a society no longer centered on the drudgery of daily labor. What a concept!
Blame it all on artificial intelligence, if you are really looking for someone or something to blame. So I guess it’s all about the intersection of artificial and human intelligence (just made that up). We will be increasingly living in a world where we can use as much of either as we may want. But I still think we need to exercise our natural intelligence on a regular basis to keep its “moving parts” working. I don’t think artificial intelligence works on this same principle?
So excuse me as I go back to reading my daily papers, and drinking my coffee to fuel my brain’s synapses, and try to figure out how to escape the “drudgery of daily labor.” I don’t think robots have to worry about that.
P.S. I will be taking an early spring break next week, but will return on March 21st. Thanks for reading my posts, or scanning them if you are a robot?