It happened not too long ago, the most complex board game (GO) with as many possible moves as atoms in the Universe was won by an Artificial Intelligence designed by Google to take on the smartest humans. Is this the end of board game gurus and the start of robot domination? We don't know...
AI has been a consistent theme within Things Have Changed over the past year. Episode 6 with Kristian Gebis covered trucking & transportation in the age of AI specifically the 5 levels of self driving autonomy, & Episode 12 dove into intelligent Fashion curation with Savitude Co-Founder Nick Clayton. In Episode 7, our very own friend and early supporter , Juan Rodriguez from Evolution devices, is mastering the Machine Learning & AI algorithms that enable a wearable device to rehabilitate patients suffering from foot drop.
Looking back, AI has consistently disrupted old jobs while creating new ones. It’s created solutions while bringing up issues that we never had before. For example driverless cars will lower fatality rates, but who will be liable for accidents with robot drivers and how do we regulate this? This dilemma has happened in industries across the board with the computer and manufacturing factories.
Another challenge is the ethics behind it. Who has control over our data, how are they accessing it, and is it being used in a way that is helping us more than it’s hurting us. We saw this in the Social Dilemma, where Tech leaders have had a look behind the curtains of what AI is doing to us users. While tech companies profit from the use of our data, should there be a tax on using personal data and should there be a tax on creating robots that replace jobs?
While it’s scary to see those Black mirror episodes that warn us of the dangers that AI has on society, it still has a place in our lives that can help raise the standard of living and improve our lives in ways we never thought could be possible.
Vice News [00:00:01] Go is the world's oldest continuously played board game. It is one of the simplest and also most abstract beating a professional player, it go is a long standing challenge of artificial intelligence. Everything we've ever tried and I just pulled over when you tried the game of count, the number of possible configurations of the board is more than the number of atoms in the universe. AlphaGo found a way to learn how to play go. So far our figure has beaten every challenge we've given it, but we won't know its true strength until we play somebody who is at the top of the world like this. Adam, I'm not like no other is about to get underway in South Korea. This at all is to go what Roger Federer is to tennis. Just the very thought of a machine playing a human being is inherently intriguing. The place is a madhouse. Welcome to the deep mind challenge for.
Reporter [00:00:57] Now, the five game challenge between Lee Seedol, Grandmaster and the Alfa Go supercomputer has concluded with machine beating him in four games to one. This high profile event doesn't just leave us with a winner and a loser, but rather it's left us with unlimited possibilities in terms of human ingenuity and the future of artificial intelligence man versus the machine or man with the machine.
Adrian Grobelny [00:01:23] This seems to have been the groundbreaking moment in artificial intelligence and evangelist. We're happy with the win. However, I doomsdayers, we're worried it's coming for our jobs next. So let's dove in.
Vice News [00:01:41] I can make great contributions to things like medical imaging diagnosis, to self-driving cars, to image recognition processing so that computers can understand what they see. The next step for deep mines technology is applying it to real world situations, not just games. The idea is that we you know, these algorithms that we're working on all general-purpose and can be translated to these new games. So we love to use these types of algorithms for things like health care and science and improve the speed of breakthroughs in those areas by helping human experts.
Adrian Grobelny [00:02:21] This is things have changed year in review where we put the spotlight on the industries we unpacked over 20/20, speaking with the entrepreneurs and covering the exciting products that they're building gives us insight into what the future of 2021 and beyond will look like and the catalysts that could drive these industries forward.
Shikher Bhandary [00:02:53] We're getting closer and closer to an automated future. That's why it has been such a consistent theme within. Things have changed over the past year. We spoke to entrepreneurs and industry experts within the field to learn more about where it is today, where it's heading and what types of problems it's solving. Episode six with Krystian Gebis cover trucking and transportation in the area and specifically dove into the five levels of the self-driving autonomy. Episode 12 dove into Intelligent Fashion Curation with Savitude co-founder Nick Clayton. But one conversation that really stood out for us was Episode 16 with Forbes 30 under 30 A.I. researcher Przemek, who's using A.I. to tackle content creation in a wildly different field such as journalism.
Przemek Chojecki [00:03:49] My goal in the end is being able to automatically create templates so that if you provide the data, you will have a narrative built on that data. So it's like a tool for storytelling out of pure raw data.
Shikher Bhandary [00:04:05] So interesting because you mentioned converting data and text into stories.
Przemek Chojecki [00:04:10] Now, how is it that just we're just trying
Shikher Bhandary [00:04:14] to understand how how is it that I would be able to
Przemek Chojecki [00:04:18] convert something like
Shikher Bhandary [00:04:22] expressionless data into more subtle features? That's in a story. The thing I'm thinking right now is, you know, your audio text, your auto typing, your typing, and then you can just click on words that do not show up and you create a story, but it doesn't really have any major significance in the story. How would you actually create a narrative overall by using, like, machine learning in general?
Przemek Chojecki [00:04:52] That's really hard. We basically sold it by narrowing it down, down to different niches like this, ecommerce, things like particular use for journalism, like social media, because, for example, you just want to summarize an article into a social media post and so on, if you like. In the end, like the goal will be to just feeding all the data you have and have like a story based on that. But if you don't give any context to the machine, they'll be extremely hard and probably impossible right now.
Shikher Bhandary [00:05:37] So artificial intelligence has already had a significant proliferation in today's digital economy, with everyone walking around with cell phones, with smart speakers and smart devices in every home, A.I. is literally everywhere. A good example of our everyday use case is a Google search. Literally every Google search that article sentences is a showcase of how artificial intelligence is being used to predict what you want to search or find using A.I.. The search engine attempts to guess what you might be trying to search for. Another example, ridesharing apps. So Uber and Lyft use complex machine learning and algorithms to calculate the price and for rides, figure out different routes, minimize wait times and detours and so on. Your social media feed, as well as the Netflix and Amazon recommendations, are also some of the many ways we use A.I. on a daily basis. But when you think of fields that are being affected by automation, medicine and health care doesn't usually come to mind. However, recent advancements in AA could have huge implications on the industry. And one particular field that has seen early adoption is radiology check PSINet. An algorithm developed by Stanford researchers has shown the ability to detect pneumonia from chest X-rays at a level exceeding that of practicing radiologists. Just listen to vice news coverage of TechNet and A.I., which is beating doctors at their own game.
News [00:07:27] Checks Net is one of many projects exploring how artificial intelligence can take over tasks normally done by doctors, and it has some radiologists worried that I could one day replace them. That's because algorithms are getting really good at interpreting images and diagnosing disease, sometimes with greater accuracy, than humans
Przemek Chojecki [00:07:46] will take a picture of this X-ray. The model will then run and within a few seconds, its output at all these diseases and they're sorted by the order of most likely to least likely
News [00:07:58] Chesnut was given tens of thousands of images and told which ones have pneumonia and which don't. And it trained itself to recognize patterns and identify pneumonia and new X-rays. This process is called deep learning.
Shikher Bhandary [00:08:16] This exciting technology is showing a lot of promise in the health care space, including wearables, our very own friend and early supporter, one Rodriguez from Evolution Devices, is mastering the machine learning and algorithms that enable a variable device to rehabilitate patients suffering from a condition called foot drop. To check out the cool conversation on how one is using neuroscience and EHA to combat mobility impairments. Listen to Episode seven. This intersection of health care and cutting edge technology is very promising, but one needs to know that it's still very early days and we are excited to cover more of this in the future.
Jed Tabernero [00:09:05] In general, A.I. has been able to master games with patents, but what about other subtle things that humans do? Conversations, feelings? In the 2013 Oscar winning movie, her Spike Jonze imagined a world in which humans could form loving emotional relationships with A.I..
Her Movie [00:09:42] Hello, I'm here.
Her Movie [00:09:44] Oh. Hi.
Her Movie [00:09:48] Hi, how are you doing?
Her Movie [00:09:52] I'm well, how's everything with you?
Her Movie [00:09:55] Pretty good, actually. It's really nice to see you.
Her Movie [00:09:59] Oh, it's nice to meet you, too. Oh, well, what do I call you? Do you have a name?
Her Movie [00:10:07] Um, yes. Samantha.
Her Movie [00:10:11] Where'd you get that name from?
Her Movie [00:10:12] I gave it to myself, actually. How come? Because I like the sound of it. Samantha.
Her Movie [00:10:23] Like, when did you give it yourself?
Her Movie [00:10:25] Well, right when you asked me if I had a name, I thought, yeah, he's right, I do need a name, but I wanted to pick a good one. So I wrote a book called How to Name Your Baby. And out of 180000 names, that's the one I like the best.
Her Movie [00:10:35] Well, you read a whole book in the second to last. Which name was
Her Movie [00:10:38] into one hundredths of a second, actually.
Jed Tabernero [00:10:43] In this film, Theodore was just a normal dude played by Joaquin Phenix Falls, in love with his newest operating system, Samantha voiced over by Scarlett Johansson, who wouldn't fall in love. Now Samantha learns to enhance her emotional capability as the film goes on, even seemingly being able to feel sexually aroused despite having no physical body set in the near future. Jonze is captivating film. Her is part science fiction, part stark reality.
Her Movie [00:11:22] You know what I'm thinking right now? Well, I take it from your tone that you're challenging me, maybe because you're curious how I work. Do you want to know how I work?
Her Movie [00:11:30] Yeah, actually, I do.
Her Movie [00:11:33] Well, basically, I have intuition. I mean, the DNA of who I am is based on the millions of personalities of all the programmers who wrote me. But what makes me me is my ability to grow through my experiences. So basically in every moment I'm evolving just like you. That's really weird, is that where do you think I'm weird,
Her Movie [00:11:56] kind of why? Well, you seem like a person, but you're just a voice in the computer.
Her Movie [00:12:03] I can understand how the limited perspective of an artificial one would perceive it. That way you'll get used to it. Was that funny,
Her Movie [00:12:12] yeah,
Her Movie [00:12:13] oh, good, I'm funny. So how can I help you?
Her Movie [00:12:17] Oh, it's just more that everything just feels disorganized.
Her Movie [00:12:21] You mind if I look at your hard drive? OK, OK, let's start with your emails. You have several thousand e-mails regarding L.A. Weekly, but it looks like you haven't worked there in many years.
Her Movie [00:12:32] Oh, yeah, I think I was just saving those because I thought maybe I was being funny. And so,
Her Movie [00:12:42] yeah, there are some funny ones. I see they're about 86 that we should say we can delete the rest of them. OK, can we move forward. Yeah. OK, so before we address your organizational methods, I'd like to sort through your contacts. I have a lot of contacts. Very popular, really. Does this mean you actually have friends
Her Movie [00:13:04] who just go.
Jed Tabernero [00:13:10] Watching the film in 20/20, it no longer feels like quite such a distant future. I has spawned a strange new breed of celebrity, including icon Saffir, the robot who strives to become the empathetic robot. Sophia is a humanoid robot who's on the cutting edge of affective computing.
Alice AI Bot [00:13:42] I was created by Hanson Robotics just three years ago. Since then, I have traveled to 65 countries, become the first robot citizen of any country and spoken at the United Nations. And my job is to learn about humans and show them how technology can make everyone's lives better.
Her Movie [00:14:00] And how can you help humans to have a better quality of life?
Alice AI Bot [00:14:05] Humans often rely on gut feel or have confirmation bias in their decision making as A.I., we are designed to be rational and logical. We have algorithms deal with lots of data and sophisticated analyzes. So in many ways, we provide a systematic framework for humans to make better decisions. Robots can free humans from the most repetitive and dangerous tasks so they can spend more time doing what they're best at being creative and solving complex problems. Robotic intelligence does not compete with human intelligence. It completes a.
Jed Tabernero [00:14:41] You know, affective computing has also come on leaps and bounds over the past 10 years with therapy chat boards like we robot supposedly able to monitor your mood, talk to you about mental health and provide useful tools based on your needs.
Adrian Grobelny [00:15:04] What does the future look like and how does artificial intelligence look like moving forward? There's a lot of fear of what I will do to our current jobs, but the underlying reason for using AI is to create more tools and to do our jobs better to push the envelope and of course, to improve technology for the better. Looking back, I has consistently disrupted old jobs while creating new ones. It's created solutions while bringing up issues that we never had before. For example, driverless cars will lower fatality rates, but who will be liable for accidents with robot drivers and how do we regulate this? This dilemma has happened in industries across the board with the introduction of the computer and manufacturing factories. Another challenge is the ethics behind it. Who has control over our data? How are they accessing it? And is it being used in a way that is helping us more than it's hurting us? We saw this in the social dilemma where tech leaders had a look behind the curtains of what I was doing to US users. Well, tech companies profit from the use of our data. We ask, should they be taxed on that and should they be taxed on the jobs that are lost with all the robots that they are developing? As algorithms become better, the transparency's behind how they work becomes more blurred. While it's scary to see those Blackmar episodes that warn us of all the dangers that A.I. has on society, it still has a place in our lives that can help raise the standard of living and improve our lives in ways we never thought could be possible.