37 Burst results for "ai"
Fresh update on "ai" discussed on That Sounds Fun with Annie F. Downs
"Answer your mail and tax any day having well. It's just such a treat. I feel like so many people listening we are all all of us are leading somewhere and you kind of have stepped into this spot in our world. Where people did you hear what carry new upset about this. And you're hearing you upset about that and about that okay. It's not going particularly well way. I believe that leading at home is probably harder outs right. Yeah yeah. It's hard ironic. One of the great joys of my life is my youngest son. Who's twenty five works with me and my company. Yeah so that is a blast and a half awesome. Okay how it. I'm sorry. I don't know the answer. But how did you get in this spot. How did you get in the spot. Where we're you're the one that we go to for leadership seth. Oh gosh it's just accident it's all grace. They look an awful lot alike often any. I burned out famously. Infamously fifteen years ago. And i realized i had no hobbies so i developed some hobbies and people like you incredible voices for leaders to have hobbies because i had no life leadership was my life and it's not a hobby anymore but i really enjoyed writing and i had started a blog back in two thousand seven when i launched. A new church connects his church. It was complicated thing. But i had to communicate publicly so i thought well i'll try blogging so i did that and then you know we launched church. It got off the ground. It was fine. And then i kept it alive and i just really enjoyed it and did it as a little hobby and to motivate myself. I had written my first solo book almost a decade ago. Now on change. And i had read michael hyatt's platform and realized okay. Publishers don't sell books authors. Sell books. So i'm like how much they will put a cover on it and and put it together for you and they'll keep them all in your garage right forever and somewhere else and so i'm like okay. Well maybe i should get serious because i enjoy leadership. I'd always done teaching and consulting in that kind of thing very low key drive forty five minutes. Go to a church basement. Tell an elder board what they should do. They tell you you're wrong. They give you a big bell and you go home right. So i mean i'd done that for years and then i made a goal of like okay. Hopefully in the first year one hundred thousand people show up otherwise i will be motivated and within a couple months one hundred thousand people showed up and then it became my side hustle and then about six years ago i shifted gears and started doing leadership development fulltime. But honestly i don't understand it fully. And i think it's just like i learned everything the hard way i went to law school. I went to seminary. They don't teach you how to run a law firm at law school. They don't teach you how to run a church at seminary and so i'm just trying to figure out. Okay what did i not know. And how can i get people fast tracked and it's become what has become today. Yeah i i mean it very recently. We i went back and listen to yours about executive assistance. I went back and listen to some years about leadership in your office that affects your ministry and where it is and and so i wonder for you. How are you determining what leadership conversation needs to be next couple things. If i'm struggling with it chances are a lot of people are i do have. I don't know whether it's a gift or just an aptitude. If you look my strength finders strang finders future is my number one. I don't let present. I live in the future. So i'm always trying to connect the dots morning i was reading on. Ai how can they. I make you a better writer. I don't know whether it can. But and then i signed up for this service. That does i still read all my own stuff but like well what is. Ai insisted writing. Look like i think a lot about the virtual office. I released a course on virtual office about a year before the pandemic on my guy it was like. Hey wayne future. Ten years from now people are going to work from home. It'll be remote and then the world blew up mean. Some of that is very natural and then we do have data. I always tell my team. The internet doesn't lie and we read the comments. We follow people on social. We try to anticipate what are the real problems. That leaders are facing today. And what can we anticipate people struggling with tomorrow and then we tried to produce resources within our wheelhouse like. We don't do everything but i do. Leadership change personal growth. And if it's something. I feel that we people we know can speak in to. I tried to do and then as you know with podcasting you can be an expert in a lot of things because i have to do is shut up and ask some questions. It's like oh we can go all over the place that i know nothing about so part of that is like free consulting for me. Yeah just like invite people like you on. And then i learned a bunch of stuff and take notes. Well carry every episode. You do is free consulting for all of us so thank you very much for doing that. I'm so thankful how have you seen. I'm really interested that you know recently. I think it's new york times but you can correct me. 'cause i think you probably know said that. Like fifty five percent of people in the workforce are wanting to leave their.
The Not so Digital Workforce
"You may think of the digital workforce as zoom meetings and shed google docs but this trend encompasses a wide range of industries and types of work. This labor refers to a really wide suite of different types of work quite often The moment is being used to refer to digital knowledge. Work so any works. That's that can be undertaken through computers. I virtually remotely roth than having to be in a specific geographical location. That's david vissel. David is a human geography at the university of melbourne and he researches the changing relationship between people and place. There's a wide spectrum of other types of works that could equally be referred to as digital works so the economy in in cities. So things like uber and delivery and all of those new types of services that we're seeing springing up in in our cities that are absolutely reliance on networks of connected mobile phones and algorithms that drive that drive both the workers and consumers so even sectors threats we traditionally associate with being very different and very not digital say things like mining for example are increasingly using. Ai and different types of autonomous developments. So yes a labor certainly a massive consideration through across a lot of different sectors of the moment and it's very variable bull people participating in the digital workforce than ever before this rapid change is something. That's come out of necessity with the emergence of the pandemic but as david explains this influx of flexible and digital workers has an impact on the way how cities function well hit potentially involves all of us in terms of the effects that it has so even if you don't work at all and no doubt you purchase things and you use different online services so even consumers are using dish labor.
Fresh update on "ai" discussed on The TWIML AI Podcast
"Thanks so much for joining us. And if this is your first time. I invite you to subscribe in apple podcasts. Spotify youtube or wherever else. You might be listening to the show. All right everyone. I am here with jens wide sal. Yes why is a senior research team lead at bory alice. A inch y. Welcome to the tuomo. Ai podcast thank you for having me sam. Also i am really looking forward to digging into our conversation will be talking about touring which is a reason project that you've been working on that does text to sequel but to get started. I'd love to have you share a little bit about your background and how you came to work in machine learning our interest so i did my garage at university of toronto in cumbersome science and math and stats took on a number of courses during undergrad. And i'll change you to work in the computer vision lab run by david fleet at ud so add it to some research projects there and afterward went onto my phd. Was david costa virus by her son and did my phd in in processes not incumbent revision so boston processes is classified arm based on parametric models. That learn very quickly from small number of data points and my work was focusing on scouting. The compute aspect of the gp onside. I also work limited on adversarial robustness of mammals. So this was the time when people just dot the you can actually apply to sign in perceivable probation to immigrants on then change the class labels always dot a can you change more than just possible can actually change the features to make the features like another image label and surprisingly turns out. Actually you can. You can actually take a picture of me and then apply a perturbation. Change the future of me to look like few troll a car for example not just any car at a particular car of particular covering dot polls So that was very surprising at the time. Always we thought the I reserve Be solved very quickly. Turns out help me more for the from truth. Things do unsolved problem. These follow that space. Not super closely. is still motivates A lot of things off. I do especially Later on after. I joined our allies. Look at the literature and people found out on not only can do this on vision can do this on the not as well and obviously they are. The probation is not imperceptible. Change to pixels but some extra taxed and also he can do this. Indoor physical word applies on patras to that are not changing pixels but some patterns that people don't pay attention to and there are actually very hard to get rid of so what that tells is These old theoretical Studying understanding this resort tax were not actually capturing with really under the hood like these people were considering Tax with a some sort of a perturbation is in a ball centered at the image and look at the robustness model. Outweigh about clearly. Doesn't capture all the other ones and daloa looks like is What the model automotive represent. The data represented the award is different fallen human from people's representation and the author aligned and Looks like they're models are picking up on sort of ashore cosworth spear correlations or other high buffa associations dot a. We don't use and only looks like that's the problem and it looks like one has to really go beyond pattern matching to really to be able to to get to the root of this problem to look got dr in a model that can reason that can try to discover the the the hypothe- relationship datta people use a in recognizing understanding reasoning. So donald stuff. Moore's my thought at a time. And i think it was also thinking chairman of by lots of people in the field and So that led me to work on an l. p. some law closer to reasoning. I felt it's languished already model of the word also because browse his part of robot. Canada it's There's a lot of actual thaw tax to a nappy inside a bank. So but before. We dive in Maybe tell us a little bit about morales and the charter you mentioned. That borealis is part of our now. What's the overall focus of the organization. So us was actually has always been part of it was started as department of our bbc. Okay andy go. Now is to technology to value to the bank and also advanced science so we do a combination of work and fundamental research and alicia a lot of our voluntary search. So tell us a little. Bit about touring and the motivation for it. How did the project get started right. So is this natural. Language database interface is a demo of anguish database interface built. And it's really just putting a lot of our word on some parsing space together. In this academic demo so netra language database interface the from application perspective the pin uses to a law a nontechnical users to interact with structured data. Set is there's lots of inside endure and You know who want to give out change for nontechnical users to to get those insights and from a research perspective. It's a very challenging natural english Problem because the underlying problem is you have to parse pasta in english or had our next languish than convert to see cole. And we all know. Natural language is ambiguous machine languages on bigger after resolve all amputate. He yard a too harsh correctly. Furthermore was different from compared to on other program. Language is the mapping. From adams. To see cole is under specified. If you don't know the schema really depend on what is the structure of schema and so he still model has to really learn how to reason using it. And in order to resolve all that may retail and correctly predicted the sequel and lastly this printer model some. You don't want to just work on this domain one. To work on demand is on databases. You're never seen before. So without st cross domain across database part of it and dodgers very challenging. Guess it's completely different. Distribution wants moved to different.
Seth Dobrin Talks About Trustworthy AI
"We're gonna talk about trustworthy a i. It's something that is increasingly in the news and concerns a lot of people. Ibm has a product called fact sheets. Three sixty that i understand is going to be integrated into products. Can you tell us what fact sheets three sixty is. And then we'll get into the science behind. Yes so let me start by laying out what we see is the critical components Trustworthy a at a high level Three things there's a ethics there's govern dated ai and then there's an open and diverse ecosystem an ai ethics is fully aligned with with our ethical principles that we've published with arbin dr ceo co leading the initiative out of the world economic forum. And i'm adviser for essentially open sourcing our perspective on a ethics from a govern data in ai perspective. It falls into five buckets. So i is. Transparency second is explain ability third is robustness. Fourth is privacy and fifth is fairness and so the goal of fact sheets is to span multiple of these components and to provide a level of explain ability. That is needed to drive adoption and ultimately for regulatory compliance. And you think of it as a nutritional label for ai where nutritional labels are designed to help us as consumers of prepackaged foods to understand what are the nutritional components of him. What's healthy for us. What's not healthy for us. Factually is designed to provide a similar level capability for a.
"ai" Discussed on AI Today Podcast: Artificial Intelligence Insights, Experts, and Opinion
"Everybody even has but different people have about about technology right be everybody's people have different levels of fear of technology and the thing about fears that they may be emotional. But there's always something to it right you know. He has unreal as some of these things might be. Some people are afraid that you know artificial general intelligence. Agi if you attended any of our foundations courses this is sort of the strong ai. The idea that we could build this one intelligence system that can do anything that we that we give it to rather than what we have now which are all narrow. Ai which are individualized ai. Thinks that can only do the task that we have specifically train them to do right and only then with a certain level of accuracy. There's no nobody's built sort of general intelligence but there's a fear that somebody will or some organizations will and that system will become really strong and will take over everything right that that's a fear right. The other fear is that robots will take over their jobs. You know whether it's the physical robots and are moving things around in the real world or the software robots that are doing these. You know these tasks machines right. There's their fears about that right. There's also a fear of loss of control. People are afraid that they're gonna lose control over over their privacy over data their decision making you know their right of choice. You know all sorts of stuff people are worried about. They're afraid that algorithms will will make decisions against them and there may be bias or maybe you know some some other aspects of control you know small number of people control over these are these are fierce right and and you can't dismiss these fears because these these fears are there. The fears are real. You know the thing that they might be people might be afraid of. That may or may not be real right now but the fierce themselves are real right. And so we're we're afraid of these things and and so we have to when we when we think about what we're doing. They asked if you have to address the fierce because if you don't address the fierce then even as you might think of them as non rational fears you know. We don't have a super intelligent. So why are you afraid of it right now but the our fears and there there are fears are real and so we have to do some things in our. Ai systems to make sure that people don't have in the back of their mind this concern. You know that that may or may not ever happen right. And of course you have issues of over-concentration of data right so on the flip side. We do have real concerns. In addition to these fears are mostly driven by sort of What people are thinking in their head is as to what may or may not happen. We actually have legitimate concerns about ai. Based on what we have experienced right we have issues of lack of transparency. These are we have machine learning systems that are making decisions and it is true that we in many cases we have a very limited amount of visibility into what's actually happening. Why did that decision happen. You may not have a satisfactory answer right. The other legitimate concern is that while the systems may be sort of neutral. Neither good nor bad. they're just technology..
Interview With Patrick Bangert of Samsung SDS
"So patrick i'm glad to be able to have you with us on the program here today and we're gonna be talking. Ai at the edge particularly in the world of medical devices. Which is i know where a lot of your focus is here. We're gonna get into some of the unique challenges of leveraging data and ai at the edge in the medical space. But i want to talk first. About what kinds of products. We're talking about people think medical devices. Okay well medtronic is tracking my blood sugar on the side of my arm and you know. Then i've got a big cat scan machine kicking around over here. What kind of devices does your work involve with. And and his edge relevant From your experience. Thank you for having me on the show pleasure to be here. We are dealing with medical imaging devices. So if you have a smart watch on your wrist. That's not what we deal with. Even though those are very useful of course to measure your exercise and sleep patterns we're dealing with technologies like an ultrasound and mri is not an x ray. And what's called digital pathology which is where a biopsy is removed and put on a microscopic slide. Those kinds of technologies produce images that are relevant to telling you whether you're sick at all hopefully not or if you are what kind of disease it is. And so the job of computer vision in this case is to detect whether is a disease diagnose what it is to find out where it is to find out how big it is advanced in if cancer stage one. Three how advanced it is. And all of these outputs can of course be created. Virtually instantaneously by executing artificial intelligence models at the edge and the edge in this case is the device itself. Yeah okay so. Some devices are huge. Mri scanners take up a whole room. As some devices are quite small ultrasound. Machines view could transport it in your suitcase and so there's obviously also price difference here but nonetheless. All of these technologies do produce an image that that is then analyzed by
Social Commonsense Reasoning With Yejin Choi
"All right everyone. I am on the line with jin. Choi eugen is a professor at the university of washington. Yajun welcome to the air podcast and excited to be here. Thanks for having me. I'm really looking forward to digging into our conversation. I'd love to have you start by sharing a little bit about your background and how you came to work in the field of ai. Right so i primarily work in the area of natural language processing but like any other feels of ai. now the boundaries become looser losers and. I'm excited to work on the boundaries between language and vision language and perception and also thinking a lot about the connection between a i and human intelligence and what are the fundamental differences in that in terms of knowledge and reasoning And so let's go a little bit deeper into that. Talk us through like some of the ways that you take on those topics in your research portfolio. What are some of the main projects. You're working on the things that you're exploring right so currently i'm the most excited about the notion of commonsense knowledge and reasoning. This was in fact the only dream of a field. The in seventy eight as people love to think about it and tried to develop formalism for it. It turns out it's really trivial for humans but really difficult even for the smartest people to really think about how to define it formally so that machines can execute it as a program so for a long time. Scientists assumed that it's Doomed the direction. Because it's just too hard so i didn't really thought about commonsense for for a long time and then it's only in recent years. Some of us got excited to think about it again. Which is in part powered by the recent advancements of neural modell's that is able to understand large amount of data.
Microsoft Teams Is Getting Hybrid Meeting Features, Including CarPlay Support
"Announced several updates for teams including audio only support for apple. Carplay is also a cameo feature coming to powerpoint live to insert teams camera into a slide deck slide and ai powered speaker coach. Coming in twenty twenty two. That will offer speaking tips and automatic correction tools for video in the coming months. Teams will also add support for intelligent cameras from. Oem's like jabra neat polly and ye link able to track speakers during a
Apple Watch Executive Takes Over Secretive Car Project
"Apples always pushing the edge. And i thought apple kind of watched walked away from from cars but the apple watch executive is taking over secretive car project now. Just two days before hideaway. Doug field ahead of apple secretive car project that tech china's tap apple watch exact former adobe. Co kevin lynch to take his place so in the latest changing of the guard for the project known as project titan which is rotated leaders about as much as reporting shifted focused They replace individual so he's been working on this since july when he was brought in to help develop the vehicle software. He's been with apple since two thousand thirteen. Curiously bloomberg rights at lynch still reports to apple's cheap chief operating officer jeff williams and not to john daria the company's head of ai. So we'll keep a watch on what's apple doing anything apple build some apple bands will
Deep Reinforcement Learning for Game Testing at EA With Konrad Tollmar
"Conrad woke him to the tuomo. Podcasts thanks sam. Thanks for inviting us to be here. I'm really looking forward to digging into our conversation. We'll be talking about As the audience might imagine the intersection of and games before we do. I'd love to have you share a little bit about your background. I mentioned what is k t h. Okay teaches royal institute of technology in stockholm. It's a technical university where i did my undergraduate as well as might be hd. So i i think my interest for a i started longtime ago starting with computer vision. I always been passionate about photography. And i saw them. There was an opportunity to combine my kind of interest for photography than webs kind of my academic. And the so. That's kind of my starting point here. Nice and tell us a little bit about the kind of research that interests you in your professorship and on your graduate studies so my phd more symbolic media spaces and we build different kinds of interactive in viramontes to connect places with vdi streams but also being able to use sensors to convey other kinds of information. If you're close or if you're in the proximity of a space for that led me and benchley to explore that further or after my ideas and i spent some time working smart and interactive environments some over this work for play and some were for more like everyday use and i think some of us could remember recall. The kind of demos sue sorted out the mit's media on the late nineties.
Ethics Panels Reject and Delay Biometrics, AI Projects for Google, IBM, Microsoft
"With ai. Ethics chiefs at google microsoft and ibm published by reuters looks at what. Ai projects these companies reject in september. Twenty twenty google's cloud unit rejected a financial firm looking to us to determine credit worthiness and earlier this year. Google blocked features from analyzing emotions. Ibm turned down a client request for a more advanced facial recognition system microsoft place limits on software mimicking voices. All three companies said they welcomed clear regulation for
The Future of Direct Response Marketing With Internet Marketing Pioneer Rich Schefren
"What do you see in the next ten years in direct response. We're going now. You know in my opinion. Just from what i'm seeing from the market as we speak like a more individualized approach for sure in terms of selling. You know obviously the changes with you know targeting and so forth with apple and so forth where you seeing like direct response going in the next ten years. Okay will some of the obvious stuff right like which like i i'll spend a few minutes. Everyone focused on stuff. That is less obvious So obviously data's going to be huge right. And i think that like some of those things that i did from two thousand seventeen to two thousand nine hundred ninety gora before we launched my thing. Right was I got gave the opportunity to look at enterprise level Ai tools that we haven't seen anything like in our world yet right And the which made me very aware very early on. Just how much like a is completely worthless about any data. And it's marginally better if it's someone else's data what really makes it valuable as your data about your customers and what they're doing right and unfortunately most entrepreneurs are not even capturing the first party data that they're entitled to for example like most people don't download all of their information from their ad accounts. And if you lose your account you lose all that data and that's gone forever goodbye right so you know first party. Data is going to be increasingly more important. That's one of the reasons why. I'm going down the road that i'm going. I need to know all that information myself about my customers. What their preferences are what. They're looking at. What turns them on what turns them off. I can't rely on a platform nor is it wise to rely on a platform. I mean the platforms are really out for themselves at the end of the
How Technological Advancements Are Changing Consumer Research With MediaScience CEO Dr. Duane Varan
"Dr duane walk the martic tech podcast. It's bigger thanks for having me excited heavy. As our cast excited to talk a little you know of the more technical side of marketing. This is the mark tech podcast. Normally kind of focus on the mar part. And you're going to bring some tech influence here in the sense of machine. Learning artificial intelligence the more sophisticated technologies we use. Let's start off talking a little bit about media. Science in the description of this podcast. I mentioned biometrics facial expression i tracking. Eg g. those sound like really complicated technologies. how are they actually being used in marketing. I mean they are complicated. Of course the issue that we address in our research is that when you're talking about marketing above all your taxes back human emotion but the tools that we use to get to human emotion usually depend on self report in other words whether it's a focus group for a survey or interview were relying on woke people. Tell us about their most journey. The problem is people lack an understanding of own motionless journey. So when you ask a person a question about how. They feel about something what they're giving. You is the rash on reputation of what they think. They must be feeling. And that's actually far removed from their actual emotional encounter so what we do at media. Science is we want to measure that emotional response directly rather than being just depended upon what people tell about it. So the tools that you mentioned are all tools that are designed to get at measuring that emotion directly. I mean they are fairly complex. One of the reason they're complex is because they very pressing the person so you can't do this against a generic set of measures you have to actually calibrate for the individual. And then you have to actually let the that individual's response relative to their universe so to speak so that you can situate them in terms of what it means for them against the data but very exciting because it just exposes layers of data that we don't see otherwise
A Powerful Intersection of AI and Robotic Process Automation With Merve Unuvar
"So marvan. I'm glad to have you here with us on the show and i know we're diving into the topic of rpa intersection with a i. I think given the coverted era is a lot of thinking about gaining efficiencies about finding opportunities for automation when you're working with big enterprises obviously. Ibm works as many of the largest firms in the world. How do you walk people through finding those pockets where automation could make a difference. What does it look like spot opportunities in workflows yet. Thank you that. This is a very interesting area. Especially as he emphasized during this pandemic a company has realized that some of the workflows could be rethought through given most of their workforce with two remote working right so before we discussed this topic further. I'd like to open up the definitions of key concepts here for the audience. So what is it business workflow. It's basically an execution of business processes that contain tasks information and paperwork related to all of these right and then they're passed from one person to another to achieve a business school better. It could be alone operable for a bank or could be a claim submissions furnishings company. So this usually moms one or more people and a hub can best leverage automation in these workflows needs to be thought through in a few dimensions so the first one is from overall process and the workflow performance point of view so in order to analyze the performance. Right manner i. We need to understand the end goal of workflow if he thinks through the same mortgage scenario is the goal to sell more loans or is it to process loans faster or it can be combination of these metrics but we need to really define the key performance indicator or the goal of these workflows and then start monitoring the performance towards these goals and one of the very obvious waste of flying. The pockets of automation is then to find the bottleneck tasks in these workflows that will impact the
Jaron Lanier on the Future of Humans and AI
"You're considered the founding father of virtual reality. Do you think we will one day. Spend most or all of our lives in virtual reality worlds. I have always found the very most valuable moment in virtual reality to be the moment. When you take off the headset and your senses are refreshed and you perceive physicality afresh. You know as if you were newborn baby. But with a little more experienced he can really notice just how incredibly strange in delicate and julia impossible. The real world is Sue the magic is and perhaps forever will be in the physical world. Well that's my take on it. That's just me. I mean. I think i don't get to tell everybody else how to think or how to experience retreat. At this point there have been multiple generations of younger people who've come along and liberated me from having to worry about these things But i should say also even in a what. I called it mixed reality back in the day in these days. It's called augmented reality But with something like a hall and even then like one of my favorite things to augment a forest. Not because i think the forest needs augmentation but when you look at the augmentation next to a real tree the real tree just pops out as being astounding you know it's it's interactive. It's changing slightly all the time if you pay attention and it's hard to pay attention to that but when you compare to reality all of a sudden you do and even in practical applications My my favorite early application of retrea audi which we prototype going back to the eighties. When i was working with dr joe rosa and at stanford med near near where we are now. We made the first surgical simulator and to go from the fake anatomy of the simulation which is incredibly valuable for many things for designing procedures for training things then to go to the real person. Boy it's really something like Surgeons really get woken up by the transition. It's very cool. So i think the transition is actually more valuable than the simulation
Exploring AI With Kai-Fu Lee
"All right everyone. I am here with kaifu. Lee chi food is chairman and ceo of innovation ventures the former president of google china and author of the new york times bestseller superpowers. And we're here to talk about his new book which will be released next week. A twenty forty one kaifu. Welcome to the tuomo. Ai podcast thank you thank them. It is great to have an opportunity to speak with you. I'm looking forward to digging in and talking more about the book before we do though i'd love to have you share a little bit about your background and how you came to work in the field of ai. Sure i started With my excitement in back in nineteen seventy nine. When i started my undergraduate at columbia i worked on language and vision at columbia and then i went to carnegie mellon for my team at which develops the first speaker independent speech. Recognition system based on machine learning actually Some the earlier thesis in machine learning in nineteen aba. I also developed a computer program that the world's fellow champion is all in the eighties. Very early years after mike graduation from Cmu i talked there for two years than i joined apple and led a a lot of apples. Ai speech natural language and media efforts later joined sgi and then microsoft where i started microsoft research asia in beijing in nineteen ninety eight which kind of became one of the best. Tom research labs in asia. Later i joined google and ran google china for four years between two thousand and five in two thousand nine. We did do a little bit for how they i mostly was Really developing google's presence in china in two thousand nine. I left google and started my venture capital firm assign ovation ventures and at san ovation ventures we invest in the bow for the ai companies. We were about the earliest and probably invested in the most companies we invested in about seven unicorns in ai alone and with a few more Yet to come so they're excited to be in the era i it's Was not so hot during much of my career. But glad scooby with the catch. The recent wave and participate in it.
VW Debuts the ID Buzz Autonomous Van
"Volkswagen debuted its. Id buzz autonomous e van. The twenty twenty one. I a mobility event in munich. This features autonomous systems developed by argot. Ai vw plans to use. Id buzz as the platform for its full-scale commercial. Ride hailing and delivery operations plans launch germany. In twenty twenty-five
Dr. Craig Stanfill Defines Artificial Intelligence
"Welcome day frames to very special discussion here. In america i with the man who knows about things that people discuss. But i think they don't know what they're talking about. It see sexy concept of summits a concept. It is artificial intelligence. He is dr. Craig stanfield the author of a new fictional work. That has i think iran a factual message if not more than one. That's called terms of service subject to change without notice stock stanfield. Welcome to america. I well thank you for having me on. Dr gorka a pleasure to be here. So can i just start with the basics. I'm not a techie guy. This is a science fiction kind of dystopia in future. I am a science fiction. Guy loved science fiction. Philip k dick bladerunner all that stuff star wars you name it. But let's start with the factual state of the art. What is the truth about artificial intelligence. What does it mean and right now. Twenty twenty one where all we in terms of artificial intelligence. What is going on right now. Is that artificial. Intelligence is being used by technology. The start the beginning. What is artificial intelligence for lehman. Can you define the term. What what is it is it is that you know thinking machines. What is that artificial intelligence. A professor of mine in grad school settings whatever the artificial intelligence says and that's always historically had a rather flexible definition in the present day. What mostly means is what i would call data science database artificial intelligence which is an area of research that are acted in the early nineties late eighties and the basic idea is this. You've got a routine decision that needs to be made and the way you get a computer to make it. Is you get a human. To look at a bunch of data and transcribe the data into what you want the to do. And so very powerful machine. Learning algorithms have been developed that will across a wide variety of topics replicate what that person would have done. And so we all know that that alexa and so forth and other services can transcribe voice and very good job of it and the way they did that was. They took a bunch of people speaking. And then somebody would transcribe it at eventually. The ai learns what it what it is that you said the same thing with translation from english to german whatever that some of the automatic translators do and it's all based on a monkey. See monkey
Australia Rolls Out AI-Powered Phone Detection Road Cameras
"Recently brought a system into play for spotting people who were using cell phones while driving or failing to buckle their seatbelts. And yeah hey getting people to not wreck their cars by talking on the cell phone and getting to getting die in a fiery crash. Because they're not wearing a seat belt. That's a positive thing and what they did was. They took camera photos of people's car through the windshield. So you know you go under a camera. Takes a picture of you sitting in your car. And then they show a bunch of those two people and say. Is this person wearing a seatbelt. Yes or no. Is this person talking to cell phone. Yes or no. And i think that they've now just put Put this into play in australia. And you know in terms of what they're doing here specifically you know who can argue with. Let's not have fiery crashes but It's kind of spooky. And where does this go if they can do that. What else can they do. Well exactly are. They going to take photos of you eating to doughnuts when you shouldn't have to donuts. Is that going to be a when you see. What's happening in australia right now. With with kofi vaccines tracking your movements having to verify that. You're on lockdown. This i think is perhaps the message of your book. Terms of
What Makes Artificial Intelligence 'Intelligent' With Dr. Craig Stanfill
"Is the difference between a program or a an effective pattern in in in computer vice artificial intelligence. Where is that i'm familiar with. Is the turing test. That helps you identify where the something has achieved that level impractical terms what makes artificial intelligence intelligence are not interested in what the intelligentsia says you as the expert. The author of this new book. What makes it. Actually worthy of. The word intelligence dr stanfield. What makes it worthy of intelligence is that you will producing a result by a machine learning rather than programming. A programmer has to sit down. You know so if a then b and here's the formula for the circumference of a circle or just set a bunch of rules and write a bunch of code and the problem is that it's very brittle the person who is writing. The code can't possibly anticipate everything that might happen in hundreds of thousands or millions of examples. And it's hard to come up with rules. People for example for years and years tried to write programs to do translation language. The language translate german to english or english. German people were doing that back when i got my phd. That's you know was a an order form of. Ai called rule based programming or just to write a program to do the translation and that really never got very far because it was very difficult to produce the rules and it was very time consuming and it didn't necessarily give good results when they started using machine learning and take a bunch of examples and then train. Something like a neural network to Do the same thing. it is basler more capable. They can do things that you were never able to write programs to
Is the AI Market Saturated?
"My first question is is the market saturated and without picking winners. What products us rising to the top. Good that you're asking this question right now in general timing of the world because here we are for those who are listening to podcast. August the twenty twenty. One people might be listening to this year from now. So this'll all seem really quaint. To those in the future but the markets actually in the midst of consolidation. Saying we're actually starting to see a lot of acquisition activity and we do track over one hundred vendors and machine learning platform space about seventy two of which meet the minimum threshold of viability. There's lots of startups in the space. We love startups. We have an affinity for companies of all sizes but when we're looking at companies who are buying products and services we tend to look at those companies that have either at least ten customers or have at least ten million in revenue or at least ten million dollars in venture capital they if they have like to customers and no venture funding raised in a little bit of revenue than. We're like just grow little bit more a little bit more. So this is about seventy two companies. At least that are in that that john rao of course all the cloud vendors are in that space. The major cloud vendors microsoft. Ibm google amazon And a few others. So those were recalled the cloud sas machine learning as a service vendors basically and then there's a whole other category of pure play machine learning platform vendor so you may be familiar data robot or did i do in that space a bunch of others that are kind of trying to pull together all the components of what's required to put machine learning and advanced analytics solutions into play and increasingly. What they're doing is they're growing through Both building out their product suites and through acquisitions so she did robots but on a tear lately did i. Two as well as been been really growing raise very significant round recently but the answer is that this market is actually starting to
"ai" Discussed on AI in Financial Services Podcast
"So ilan today we're talking about maturity and readiness. I appreciate you as a emerge a subscriber with us here but also is somebody with some very unique experience working in one of the largest companies in america actually bringing ai to life. Let us know how you think about this topic of maturity and kind of how you assess it. Hi dan thank you so much for inviting me. It's been an honor to be here with you or yup. Yeah so my job is very unique in the sense that i'm not necessarily evita. Science is my job is to understand. The needs of big enterprises would save fifty to half a million employees and how they can best utilize machine learning and opie computer vision in their lines of business yet. Some of this enterprises can have a hundred lines of businesses. Each one of them will be in different levels of ai maturity yes so my job is to come in and understand what problems of trying to solve and how they i can actually solve that up so in order to do that. I have actually leveraged the work that you guys have done at emerge. I learned from you. Guys in to assess the level of maturity from different lines of businesses mainly mainly three components. That luke for one is the skills are available in that line of business to is the coach or the business culture of the company out of business as well as the importance of data. So let's deal to each one of those in terms of skills. It's very important that there are unavailable skills of course but also it's also important to understand that. Ai is a team sport.
"ai" Discussed on AI in Business
"We have a learning bosmo moses. We have an amount of on boarding process sort of tutelage that occurs in the beginning. Now we're we're wondering all right well. Where does artificial intelligence fit into all of our various sales data in order to potentially layer value onto what on boarding should look like in the future where it begin to fit in. What are the parts not. Everything's getting revolutionized but some bits are. What are those bits. It's interesting what we're seeing. Now is the ability to record and be an all these calls just given where technology as a hook into calendars and plug into zoom. We now the ability to capture these interactions. At a way that we had done before. We're actually in is interactions and conversations. But that's a lot of video content to go through if you were to go through that in real time it. It's not something that the average new hire or manager qaddus were. Ai plays a really critical role. Is identifying those very unique moments. Those moments that drive outcomes the ability to understand what does best next step. Oh quite or how should we asked discovery questions. Or what sales methodology are we following in. How should i. He handle each of the different sections of or an objection the ability to go through the new hire and say i want to hear all the pricing objections. Let me listen to the last. Ten calls were price objection. Came up here. Our best reps answered it and just literally listen to that thirty second piece jumping in jumping out quick comment that ability to quickly here so many of these edge cases dramatically impacts voted on poor people faster. Eso and i can imagine if. I'm a sales manager. If i get a sense of new rep. Extra new rep y where they're getting hung up or where i feel like they're getting hung up. I might ask them to learn more about that part of the process right. Listen to how. Susan handles pricing questions. You know listen to how. Jim opens a call and and really bonds with people quickly. You know that kind of thing right now some of that just grabbing some recordings from jim doesn't necessarily require ai but clearly is being used here somewhere. What you're saying makes me think that. Nlp is able to extract from minute would say fifteen fifty to fifteen minutes. Fifty two seconds to you know Eighteen thirty is a pricing with confidence level of eighty five. We think that this audio clip is is pricing. That's literally what happened in my head when you said that but it probably doesn't work exactly like that. How does it work. But that's smart. We have our team of researchers yarn a phd on our research team as well but we've deeply studied the language of sales and all the different type of language that drives positive outcomes so pricing and the product mentions and objection handling. These are terms that we've studied and we built these topics into our ability to track those moments. We will pre identify. Hey that's an at risk deal. We've identified at risk language. Let's flag it. Let's call it out. Click of a button. Go right to that moment or several moments in a call so these pre identified list of topics that you want to drill into is really the special sauce dot it okay. So tell me if. I'm right or wrong here where is fitting in is. Were recording all the calls. We're leveraging nlp on top of it and we're able to determine when to certain topics come up. Maybe even a little bit about sentiment. I'm not sure and then some kind of flags for at risk deals in my mind. I'd love to know what the lingo is at risk deals..
"ai" Discussed on AI Today Podcast: Artificial Intelligence Insights, Experts, and Opinion
"Healthcare and for education in using more energy with the help of ai with the with the shoveling process and something that has been called the date wallet where where we can manage Getting enable citizens do manage their their data capital or something that is currently merging as as day to capital city. The three layers that we have come up with. And i think the the the key ovation is in the third layer that while we see a lot of countries that they they developed institutions and resources and In key areas where they want to focus on. This is something that came up. Maybe because of our Our structure of a lot of ideas coming from the bottom up that we need things that we can actually focus on and then showed results in for the everyday people that brings it closer to their to their lives and make it relatable. So maybe that's just a quick recap of the. I'm the strategy. Great yeah thank you. i know. It's a very comprehensive strategy. So it's not always easy to summarize just a few minutes that was great and then i know that follow up questions dig a little bit deeper into some of these including this one so one area of the strategy focuses on education competence development and societal preparedness. I know that there's also a goal in this strategy to create in a innovation centre as well. Can you share with us. Some more details around this picture so as i said this kind of competency development in social preparedness that it's a key factor that that the technology is there. We could gain a lot of value from this if a lot of people knew what this was. And we're not afraid of it. Did not have misconceptions are didn't have the first association to be the terminator seven big job to to transform a societies thinking from the judges of of what they haven't here to a realistic view of what is really happening although this is a very personal thing because an artificial we a lot of cases identify ourselves with our intelligence into..
"ai" Discussed on AI Today Podcast: Artificial Intelligence Insights, Experts, and Opinion
"And we also wanted to to define some specific objectives that that's are going to be relevant for us. We had some high level objectives that we would like to achieve this fifteen percent of gdp growth for coming from a and also having one million people who are finding more value added jobs with the help of ai so not maybe not replacing drop. Maybe not new jobs. But having an ad on competence that they would be able to use a enabled tools are are can collaborate with robots in a way that they can be more focusing on the creative or the human league arts. This was a huge. Wouldn't say debate but it's a lot of ongoing discussion about how to how to do this. So the end result of of this. This kind of thinking was that we built a strategy in a three layer Format the first layer. The foundational layer were six pillars that we defined that may indicate from all the things that we learned from all other species and all of the things that how can we prepare the economy and society for this and the three retailers are what we call the value chain That how can we kick start the the data economy which was a huge work. How do we do the research and development. And how do we do the ai. Adoption or applications and this is a cycle that the more applications. We do the morning that we have more developments. We can have the more applications. We can have end and locally in the country and also in all of these stages integrating with available ecosystem in every part we defined new institutions new actions new resources that are allocated to different kind of purposes. The second sleep dealer is work. We called framing. Pillars are education end of this society the fourth pillar the infrastructural kind of investments and the regulatory unethical frameworks. Think these are kind of self explanatory. 'cause we have to invest in education either in a society as a whole either in changing the educational system. One of the big things that we were discussing is that. How can we manage the the education of the currently Working people since in ten years. Someone who already worked for ten years is gonna still be off time in. They're not even half time at their at their carrier. So it's not a level a question of changing in the universities or even at the at the elementary or high school levels that it's a.
"ai" Discussed on The TWIML AI Podcast
"Translating these technologies into the clinic in ways that are helping everyone not just a select population and then also working on systematic scoping review which is kind of like looking got all the literature in primary care. Ai and trying to see what whether the gaps in like what are the things that research should be doing more often to really be. Proponents are helping progress. The field more towards like more equitable use of ai. Call in the scoping review. That is kind of a literature review and search and it sounds like it would be relatively easy given your sadistic that. There's actually no literature on the topic. Yeah so our strategy has been yeah first of all. It's really point that there's like no literature we're just like oh my gosh like we can't do. This is useless. Like we're gonna like mini lesson. One hundred articles. Summarize no one. No one's gonna care so the goal here is we're gonna look at all the articles on a in primary care and then from there kind of like. How many articles actually talk about perpetuating disparities. How many articles even do some kind of like a subgroup analysis. Like are they even looking at how their algorithm performs across different types of patients or they just saying. Oh here's like one score. And then you know you can imagine a scenario which actually has been studied and shown for x ray classify. Where like you know. We extremely well on the white patients but then on black patients or modernize other marginalized communities are just going to have like way lower like scores on performance and You know a part of the issue is like a data issue like you know we have a lot less data but also the data is like fraught with a lot of historical and institutional biases inherent in the sources. That people need to create more aware of. I think what are some examples of those one of them is just sky. Cardiovascular risk for example black patients have lot higher risk of developing cardiovascular disease. Not because of necessarily any biology just because of you know we have poor access to health care worse. Worse access healthcare worst like asked education all of factors like targeted advertising about like sugary drinks. And things like that. There's so much that is out of control out of people as they grow up and live their lives because of this. There's like an association right like you might have an algorithm that just like really highly like put someone at just constantly recommending someone to you know get more interventions on their cardiovascular health even though it might actually make sense just like say you have the white and black patients in a picking up cues based on race or something like that. Yeah line house. As opposed to actual indicators from their physical condition and there's also this this issue of like missing data like some people just aren't really good at for good reason like they have other things to be worrying about like putting food on the table or something like that and they're not as well able to like get to the clinic in like see the doctor..
"ai" Discussed on The TWIML AI Podcast
"Of knowing our own limitations continuing to investigate looking into the pasta look to see how it is. Technologies have already been used at what the implications also beacon developed new kinds of foresight or prediction into the future of how those tools, and that will give us aware of demarcating what it is. We consider in that kind of scope and what it is the. Scope of externalities in some sense is the one that we wanted to have based on the changes that will come from. The tells a little bit about how the the paper and the kind of framework is is organized. How do you introduce this issue of decolonialization in the context of AI and? Run through all of the folks and ways that it touches folks in the eighties. Them. Just said the paper begins by saying that actually one of the things we need to do is expand the idea of what we're thinking of when we think of ai that ai must both be this kind of? Object this technical subject that we are always talking dealing with. So much of my own research in at the same time, it also needs to be this kind of a subject itself where we are looking at the structures, the systems that kind of match. That onto playing, it's a with that idea of object is an object, an AI as a subject. You can now see that you need a very broad kind of tool with which to deal with that and if. You just said in the scope of. is going to talk so many things in the future what we need, what we are missing is the tour of foresight into the future to understand how technology impacts the future. So then the paper begins to set up that actually is missing tool of foresight. And how we can develop that missing one missing to is to use the advantage that we have of historical hindsight and so necessarily then that makes the bridge to this idea of a colonial history which is shared between all of.
"ai" Discussed on Pulse of AI
"I just hope my fellow citizens will be rational. On those issues if we're going to get covid nineteen or the next and inequality pandemic after that under control. And Prep where do you stand what are your thoughts around the need for legislation for responsible I've verses corporations adopting standards on their own. Fortunately, there are already a lot of laws and rules in place that would drive ethical and responsible behaviour for corporations and some of the May. Obviously most of them were were created before A lot of the use of AI were anticipated with the implications are the same using ai to generate false and misleading statements. which is now the deep fake mindset that is now possible and very hard to detect but also goes to you know some of the laws that are in place to protect people for libel and false information and things like that. So it's not as if there aren't laws they may have to be adapted somewhat to anticipate some of the uses of AI, but it's all. Responsibility of government and industry to behave responsibly where they I. I also think when it comes to bias, a lot of people focus on bias and they I and whether it discriminates I think we missed the point that people discriminate and much of the data. The trains AI systems is delivered in given by people who themselves have unconscious bias. It's not all of us to behave ethically to not discriminate and to ensure that the systems run our businesses don't discriminate in. So I think the idea that this is something that's an AI problem or should be delegated to technologist is the wrong idea. It should be managed by rules and regulations and make sure we don't. Discriminate and that people who built systems using a I other technology you know should be held responsible when they do So there are rules that can be enforced and then I think there's also awareness and responsibilities that can be brought in I work with companies all the time to establish responsible governance with bringing the.
"ai" Discussed on Pulse of AI
"Anytime the day I become useful and practical, and in daily use the stop calling it a as. Far As we're concerned, it is still a either I, as a science has been a continuum. We have made great progress, and we will continue making great progress. I. Don't think I should be this one point in our minds that we believe is the all knowing all. You know a capable that we compare it to I think we should. We should look it as a continuum of improvements in our in our tools, and if we do that, then we won't be disappointed in. Hopefully, we won't. We won't have another AI winter. And see the power of what about you Jordan? I stand by the paper. I wrote in two thousand six. which was before the great blooming of up deep learning? I still think that we need to look at how intelligence arrives in nature. Office at Ecosystem. Insect colonies. And so forth and try to understand. The form of intelligence than adapation. That has occurred especially in evolution understanding evolution has open ended. Creative structure. I think he's probably key open problem. That I see for. Today. So, Again, that's just me. Baby back has this compensation has gone? Well! Thanks for joining me today, both of you. Thank you! It was fun. Buzzer. Well there, you having I hope you enjoyed the discussion and don't forget. Follow me on twitter at the poll survey and connect with me. I'm Lincoln until next time..
"ai" Discussed on Pulse of AI
"At it <Speech_Male> really is a set of a <Speech_Male> religious belief. <Speech_Male> <SpeakerChange> Is Point <Silence> with the? <Silence> Profits. <Speech_Male> <SpeakerChange> <Speech_Male> You <Speech_Male> know claiming it's <Speech_Male> coming at a time <Speech_Male> ten years thirty <Speech_Telephony_Male> years like Ray Kurzweil. <Silence> <Speech_Male> <Speech_Male> <SpeakerChange> Larry <Silence> University. <Speech_Male> <Speech_Telephony_Male> And and <Speech_Male> the the Guy <Speech_Male> Behind the book Super <Silence> Intelligence. <Speech_Male> <SpeakerChange> <Speech_Male> I'm not <Speech_Male> a believer. I mean I believe. <Speech_Male> <Speech_Male> That if <Silence> you have a game player. <Speech_Male> <Speech_Male> And you add more <Speech_Male> CPU time <Silence> to the game player, <Speech_Male> it becomes <Silence> stronger. <Speech_Male> <Speech_Telephony_Male> But <SpeakerChange> we've known <Speech_Male> that. <Speech_Male> For thirty <Speech_Male> years we've known <Speech_Male> that <Speech_Male> if you can go deeper <Speech_Male> with Alpha Beta <Speech_Male> search <Speech_Male> you get a stronger <Speech_Male> player. <SpeakerChange> <Speech_Male> Okay, so <Speech_Male> it's not <Silence> surprising. <SpeakerChange> <Speech_Male> <Speech_Male> That we had <Speech_Male> deep blue. <Silence> K., <SpeakerChange> win. <Speech_Male> It's <Speech_Male> not surprising <Speech_Male> that we have very <Speech_Male> strong. Players <Speech_Male> go in <Speech_Male> at backgammon <Speech_Male> and so forth. <Speech_Male> <Speech_Male> What super <Speech_Male> telling says <Speech_Male> is <Speech_Male> if you add more <Speech_Male> <Speech_Male> compute power <Speech_Male> to a strong <Speech_Male> game player. <Speech_Male> Say a human level <Speech_Male> gameplay. <Speech_Male> Now double <Speech_Male> <Speech_Male> the speed of the CPU <Silence> when it go deeper <Speech_Male> while <Speech_Male> now by definition <Silence> it stronger. <Speech_Male> Than humans, <Speech_Male> <SpeakerChange> it's super <Silence> superhuman. <Speech_Male> The <Speech_Male> fallacy <Speech_Male> is the idea that you could <Speech_Male> add more computer <Speech_Male> time to <Silence> chess player. <Speech_Male> And <Speech_Male> it will suddenly <SpeakerChange> start <Speech_Male> doing symbolic <Silence> capitals. <Speech_Male> <SpeakerChange> <Speech_Male> All <Speech_Male> the is are in their <Speech_Male> own silence. <Speech_Male> and. You can increase <Speech_Male> the strength of <Speech_Male> AI. Silo, <Speech_Male> but that doesn't <Speech_Male> let it hot between <Speech_Male> silence. <Speech_Male> And that's probably <Speech_Male> the greatest open <Speech_Male> problem <Speech_Male> for a is <Speech_Male> a problem of transfer. <Speech_Male> <Silence> Of Tasks. <Speech_Male> <Speech_Male> It's a hard problem <Speech_Male> for learning. <Speech_Male> <Speech_Male> people <Silence> seem to do it. <Speech_Music_Male> <Speech_Male> I learned how to <Speech_Male> balance a checkbook <Speech_Male> and now I got <Silence> a job as an account. <Speech_Male> <Advertisement> <Speech_Male> I can transfer <Speech_Male> what I learned for balancing <Speech_Male> my checkbook <Speech_Male> to balancing <Speech_Male> the books of a small <Speech_Male> company I can <Speech_Male> convert to <Speech_Male> balancing the books of a <Silence> large company. <Speech_Male> Is <Speech_Male> Not doing that <Speech_Male> and no matter <Speech_Male> <SpeakerChange> what you <Speech_Male> think, the singularity <Speech_Male> is, it's <Speech_Male> the idea. <Speech_Music_Male> <Speech_Male> That more <Speech_Male> computer time <Speech_Male> will lead to <Speech_Male> explosive <Speech_Male> generalization <Speech_Male> of I <Speech_Male> had not <Speech_Male> shown anywhere. <Speech_Music_Male> That's not <Speech_Male> <SpeakerChange> a stick <Speech_Male> evidence. <Speech_Male> Showing back <Silence> to be the case. <Speech_Male> <SpeakerChange> So <Silence> <Advertisement> I'm not a belief. <Speech_Male> <Advertisement> <Speech_Male> <Advertisement> So one non believer, <Silence> what I <Speech_Male> couldn't <Speech_Male> put it better. I <Speech_Male> <Advertisement> think I! <Speech_Male> <Advertisement> Totally <Silence> <Advertisement> agree with Jordan. <Silence> <Advertisement> <SpeakerChange> <Silence> <Advertisement> <Speech_Male> <Speech_Male> <Speech_Male> But let me ask you this. <Speech_Male> If you look at <Speech_Male> singularity with a capital <Speech_Male> S. right, <Speech_Male> you know the way singularity <Speech_Male> university in <Speech_Male> thoughtless do, but <Speech_Male> we're kind of at singularity <Speech_Male> with small <Speech_Male> less, <Speech_Male> and that gets back to <Speech_Male> this <Speech_Male> demystification, <Speech_Male> of Ai <Speech_Male> on, <Speech_Male> and if you <Speech_Male> look at it from human <Speech_Male> computer, you know <Speech_Male> human augmentation <Speech_Male> to a <Speech_Male> augmentation of humans. <Speech_Male>
"ai" Discussed on Pulse of AI
"An expert in deep learning. I don't want to. I can tell you people who criticize it. All I would say. Is that like any monopoly or oligarchy? Eventually the chinks start to shop. So and I think that Likud. In particular is very sensitive to over claiming that's going on in the field. Because of claiming is something that precedes I'd disinterest in a I. And I think he certainly wants prevent that, so so there are. Still on on an annualized basis, there are still very interesting results. Coming out of the learning, but just not the kind of quote. Artificial General Intelligence. The generally i. That's been that's been hype. The end in that's not showing up anytime soon. And I think that the leaders of the sealed of of the learning a now realizing tamp down on that soda. That sort of Speculation. So, but back you talk machine learning deep learning evolutionary. which approaches are better for which types of problems? Yeah I mean there's definitely today. There's no one-size-fits-all and so even even when we're applying deep learning systems, there's a major design phase that goes depending on the domain, and how close or far that domain is from state of the art, deep learning systems, and when you even look at deep learning systems for say text analysis in natural language, and compare them to image analysis and others, the architecture you know, the design of the deep learning system is vastly different and so even within deep learning. There's no one-size-fits-all a let alone a other AI. Systems there is still an engineer discipline for the most part with more promise of being able to. Retain some level of autonomy after deployment, but even there it doesn't mean that you turn switch and the I just runs forever and you don't need to. Modify it or or make changes of through times for sure you know, we still have job security. It will don't like computer programs at change on their own. So with a straitjacket that old machine, learning, then which is you have a training phase, and then you basically, if something that goes into read only memory and.
"ai" Discussed on Pulse of AI
"Now both of you work and our enthusiasts, I. Guess is a good term. Evolutionary I. Maybe I'll take a stab at that. oftentimes, a lot of these techniques Jordan talked about come together. To. For us in in a solution when you for example, look at how a deep mind beat the world champion and go. It wasn't just steep learning. It was a combination of deep learning research and some other techniques that made that possible. End They often complement each other one of things that we found. Is that. There's an element we you can only. I can name it as creativity that's missing in a lot of these machine. Learning based techniques such as a neural networks because they do what we call a sort of a hill climb when we do back propagation when we train these systems. And by that I, mean that they can incrementally improved their behavior by looking. We can look at the minor changes that might actually make improvements and build on those changes, whereas when you actually use the algorithms evolutionary computation You're maintaining several. Candidates solutions at the same time so in effect. You're actually looking at multiple. Possible Solutions. And not just doing a hill climb on each one, but learning from the distribution of those solutions as to what your search space looks like in the first place, and that allows for creativity. And, so it makes sense for us to bring in evolutionary computation along with some of these other techniques, such as neural networks and deep learning and other techniques to strengthen. Some of these solutions that approach we call. He Blew Sherry I. And the broader categories blistering computation. Jordan you have anything to do that or I mentioned evolution in general. Grab bag of AI because. As a field has sort of defined self away from neural networks and pollution and other biologically inspired search mechanisms. In favor of? Basically in favour of Algorithm only. and. As only after the great success of neural network that has now welcomed into the tent. So far evolution is still outside the tent of what's defined. By the way the Church of AI. That's right. It mentioned, but it is probably. The most likely to succeed if any approach day I after all intelligence is a result of biological evolution. To the extent that we can understand that. Not just the optimization, but also the emergence in the creativity. Degenerative of natural systems we might be able to have A. Computer based process evolution that results in more and more of what we might measure as intelligence, but that's still hasn't been done so I can't really claim. That has been done just that. That's really where the answer might lie. So as a follow on question that really quick, so you mentioned hidden in Luke Kuhn right Didn't they just recently? Come Out, and we're talking about sort of the limitations of deep learning. And we're talking about you know there needs to be some new approaches, is this? Can you talk about that? Some of the limitations deep learning, and is this you know sort of is evolutionary i. To overcome those limitations. Are Not.
"ai" Discussed on Pulse of AI
"Our Own intelligence, and maybe life as a whole, so Jordan talked about open ended evolution. For example, it's fascinating topic that goes beyond just human intelligence. And I, also totally agree with Jordan with respect to the hype around it often. A systems are not compared to a line that's even human intelligence like a single human intelligence compared to everybody's collective intelligence, or maybe even a fortune teller, and that's where we keep stumbling into these AI winter's. Over and over again, the expectation is set to high think. Now before we get into some of the bigger issues around the demystification of AI. Let's start at ground level. You know when we talk about I. It's sort of this all encompassing term. It's kind of like saying. I work in science or you know I work in medicine, right? What what are the differences between machine learning deep learning evolutionary, a I I mean. Break those down force if you will and Jordan maybe you could start and help us think through that. the kind high vector is for some deep learning, which is a scaled up version of the neural networks of the mid eighties. They didn't scale very We lost a lot of interest in them, and then starting at around two, thousand, ten, two, thousand, twelve. A people discovered some tricks. optimizations that allowed the neural networks to scale up large systems. And this Correlated with the beginning of the distribution of GPU's. As compute engines, not just as graphic cards. And so together these two things, the development of techniques for large scale neural networks, the delivery of large scale neural networks on clusters of GPU's led to this renaissance in neural networks and. Also to the touring. Award to Jeff Hinton. And Yon Likud. This past year. I though is more than just machine. Learning a good part of AI is in machine, learning, probably sixty or seventy percent, but the other have to do with problem solving in representation and search not search like Google search. Search through the space of chess moves to figure out. What's the best position to take? At, so there's a lot of techniques that go into. Hey, I besides machine, learning, but deep learning is currently the hot area..
"ai" Discussed on Pulse of AI
"Co evolution open ended evolution, which we try to replicate the magic of learning. That took place on. Fantastic, and what about you walk? Yes I, too have spent my career in a I since the early eighties, and but as an. I also did three startups in I the first one had to do with natural language interfaces, and the technology ended up in Siri not in doing natural language for the past. Thirteen fourteen years I started. A company called sentient where. We worked on a revolutionary computation. spun off Fund and we split off a website opposition company, and then rejoined forces with cognizant. We're a I'm the VP of evolutionary eye for cognizant, and he blew is what we consider. The tip of the spear in Cognizance Ai Strategy. Exciting. I'm looking forward to the show, so there is a lot of confusion out there on. Jordan, What is after all? Wrong there's really two different meanings of the word one. Is this mystical fictional idea? Of machine that would have human capabilities. They commanded data robot for example. And the second, the more realistic definition of is at collection of tools and techniques. For adding. Intelligence to our, computer systems. And at right, now is a pretty hot face. Using. Research results machine learning, specifically deep neural network learning. and. That's what led to a lot of the current hyphen confusion. But in every era of AI, some successes lead to media over. Compensated hype. And that often leads to. A loss of interest eventually. The AI winter as we call him sometimes. Winner was worse than disinterest. Disinterest leave us alone, but here I went to. They come for us. We don't want anybody coming forth so back. What do you think about that? What what what is I? Do agree with that or things like to add or. Oh, absolutely I agree with it. I think it's a set of tools inspired by human biological intelligence. It pushes us. Actually define what we mean by intelligence, various aspects of intelligence respect is interesting but did also has very practical applications. And often maybe not always often. Ai Systems are systems in which we pose the problem and we expect the system to solve it. and and therefore they do represent a a sea change to the way we were used to using software where we actually build the solution as part of the software. So. Yeah, that's to me. That's that's very interesting also gives us a sense. Of what we mean by intelligence for example gives us a away to reflect into..
"ai" Discussed on Artificial Intelligence (AI Podcast) with Lex Fridman
"Ai and that sort of I guess bought me a ticket to be involved in all of the amazing things that are going on in I research right now I do know a few people who I would consider pretty expert on both fronts. And I won't embarrass them by naming them but there are there are like exceptional people out there who are like this the the one the one thing that I find. Is A is a barrier to being truly world-class on both fronts is is. The, just the the complexity of the technology that's involved in both disciplines now. the the engineering expertise that it takes to to do truly frontline hands on AI research is really really considerable. The learning curve of the tools just the specifics of just whether it's programming the kind of tools necessary to collect the data managed data to distributed, compute all that kind of stuff. Yeah and on the neuroscience I guess I'd there'd be all different sets of tools exactly especially with the recent explosion in you know in neuroscience methods. So but you know so having said all that I I think. I think the I think the best scenario. For both neuroscience and AI is to have people who interacting who live at every point on the spectrum from. Exclusively focused on neuroscience to explore exclusively focused on the engineering side of ai but but to have those people. Inhabiting a community where they're talking to people who live elsewhere on the on the spectrum and I be I may be someone who's very close to the center in the sense that I have one foot in the neuroscience world and one foot in the world, and that central position I will admit prevents me at least someone with my limited cognitive capacity from from being a truly true having true technical expertise in any domain but at the same time. I at least hope that it's worthwhile having people around who can kind of. See the connections between unity the Yeah. The the merger intelligence of the community I guess nicely distributed. Is Useful. Okay. Exactly. Yeah. So hopefully that I mean I've seen that work I've seen that workout well at mind there are there are people who I mean even if you just focus on the I work that happens at deep mind it's been a good thing to have some people around doing that kind of work. WHO's PhD's are in neuroscience psychology every every academic discipline has it's.
"ai" Discussed on The AI Podcast
"But how should the US approach trying to keep up or just keep advancing with AI, and how might the US and China approach their relationship when it comes to the future of AI Shum? I think this is really not an arms race because u s clearly stronger in researching technologies and China is stronger in implementation monetization. So if I were looking at it without regard for the fact that their two countries, I would love to invest in companies that have American researchers and then Chinese entrepeneurship implementers with China's Mark. And day the provider that would be ideal combination and other points way to look at it is that there are really two parallel universes. The US market and the China market, the two universes. Don't cross the Chinese companies don't sell to American customers or vice versa. So the gain of Chinese AI will never come as an at the expense of American company, so given to parallel universes, even if we don't work together that much at least it's important to look at the other universe. Right. Learn your lessons from it. So I'm kind of looking at the current trade disputes has not very interesting or effective thing on a I I think a will move forward. China cannot be stopped from being very strong in a implementation and monetization. And that there's so many ways to work together that. If the trade dispute could end I would love to see those connections to be rebuilt. But even if the two countries were to move on their separate ways, I think China will develop its own industry as what the US in parallel right now, it's going to ask about this later later in the chat. But I think now's probably better time your company, your current company sign ovation ventures focused originally believe primarily on Chinese companies, but also has become one of the first Chinese VC firms to really establish a presence in the US, can you speak a little bit about what you're focused on. What's move your portfolio companies are doing, and is there any of that what you were just describing that kind of ideal scenario with US research and Chinese implementation going on anywhere in your firms companies. Well, we're very proud to have been one of the earliest tech VC's very early stage series AM be in China. Our strength is in. In understanding technology trends, get in before the other v c it's so we've been investing for five years as I mentioned to you AI really became hot in the last two and a half years. So the first one half year. So we did it very quietly while we uniquely knew that when saw the opportunity and with greatly benefit is some of our earlier investments are now have become unicorns. So we have a total of fifteen unicorns that we have invested in starting from series A and B so not not just getting in after they're even the corn and times ago. Yeah. Exactly. And also just within the I the core. I companies like Thomas driving for finance AI semiconductors we've invested in five companies that have become unicorns just in I and their total valuation is about twenty one billion dollars south. So we've been a clearly the. Most successful a investor in China, and we're very proud of it and wanted to do a lot more. Our original plan was to invest in the US. We have made some investments some of which have demonstrated the US China's energy we've invested in a company called fictive, which is Silicon Valley based and can help American product innovators used a China, Shenzhen hardware winkle system with invested in some AI companies that use technologies that have been used in China, for example, one company developed a grammar checking error correction essay checking software, very very niche company. But for Chinese many Chinese learning to write in English, right? That's a great tool that could be used we've invested in a company called the wonder workshop, that's a great educational tool and China's one of their biggest markets we've helped him go the China. Now going forward. I think Winnie to understand how the safest and export control and those issues whether it would still may sense for us to continue investing in the US. Now, we've only put five percent of our capital in the US. So we can go higher or go lower..
"ai" Discussed on Gigaom AI Minute
"This is the minute brought to you by Byron Reese. And Prior A and I talked a little bit about explain ability how people want an explanation on why AI makes a decision. I talk about how explanations imply understandability and that some decisions by a is may not be understandable by humans. Why would this be? In large part, it's because I models are systems and their systems with an enormous number of levers if you were to ask the question. Of about whether on the Earth broadly speaking. Then, you have to say well, there really isn't. A single Y of anything that happens there the oceans and there's the solar winds and solar activity in this vegetation and there's all of these other factors. But in addition to the complexity of it. is every single factor in is interdependent on other factors? So. There's no way to understand just part of the system. The only way to understand the system is to understand the entirety of it system and how everything within it interacts. As an models become more and more complex. This probably isn't. A feasible thing. Your other. which has a thermostat and heater and a few basic components in a system. That you can understand and they I'm Mata with billions of pieces of. Data thousands of different fields and different weightings and all the rest maybe a system that's beyond understanding..