Dr. Fei-Fei Li, the Chief Scientist of AI/ML at Google joins Melanie and Mark this week to talk about how Google enables businesses to solve critical problems through AI solutions. We talk about the work she is doing at Google to help reduce AI barriers to entry for enterprise, her research with Stanford combining AI and health care, where AI research is going, and her efforts to overcome one of the key challenges in AI by driving for more diversity in the field.
Dr. Fei-Fei Li
Dr. Fei-Fei Li is the Chief Scientist of AI/ML at Google Cloud. She is also an Associate Professor in the Computer Science Department at Stanford, and the Director of the Stanford Artificial Intelligence Lab. Dr. Fei-Fei Li’s main research areas are in machine learning, deep learning, computer vision and cognitive and computational neuroscience. She has published more than 150 scientific articles in top-tier journals and conferences, including Nature, PNAS, Journal of Neuroscience, CVPR, ICCV, NIPS, ECCV, IJCV, IEEE-PAMI, etc. Dr. Fei-Fei Li obtained her B.A. degree in physics from Princeton in 1999 with High Honors, and her PhD degree in electrical engineering from California Institute of Technology (Caltech) in 2005. She joined Stanford in 2009 as an assistant professor, and was promoted to associate professor with tenure in 2012. Prior to that, she was on faculty at Princeton University (2007-2009) and University of Illinois Urbana-Champaign (2005-2006). Dr. Li is the inventor of ImageNet and the ImageNet Challenge, a critical large-scale dataset and benchmarking effort that has contributed to the latest developments in deep learning and AI. In addition to her technical contributions, she is a national leading voice for advocating diversity in STEM and AI. She is co-founder of Stanford’s renowned SAILORS outreach program for high school girls and the national non-profit AI4ALL. For her work in AI, Dr. Li is a speaker at the TED2015 main conference, a recipient of the IAPR 2016 J.K. Aggarwal Prize, the 2016 nVidia Pioneer in AI Award, 2014 IBM Faculty Fellow Award, 2011 Alfred Sloan Faculty Award, 2012 Yahoo Labs FREP award, 2009 NSF CAREER award, the 2006 Microsoft Research New Faculty Fellowship and a number of Google Research awards. Work from Dr. Li’s lab have been featured in a variety of popular press magazines and newspapers including New York Times, Wall Street Journal, Fortune Magazine, Science, Wired Magazine, MIT Technology Review, Financial Times, and more. She was selected as a 2017 Women in Tech by the ELLE Magazine, a 2017 Awesome Women Award by Good Housekeeping, a Global Thinker of 2015 by Foreign Policy, and one of the “Great Immigrants: The Pride of America” in 2016 by the Carnegie Foundation, past winners include Albert Einstein, Yoyo Ma, Sergey Brin, et al.
Cool things of the week
- Terah Lyons appointed founding executive director of Partnership on AI article & site
- Fully managed export and import with Cloud Datastore now generally available blog
- How Color uses the new Variant Transforms tool for breakthrough clinical data science with BigQuery blog & repo
- Google Cloud and NCAA team up for a unique March Madness copmetition hosted on Kaggle blog
- AI4All site, they are hiring and how to become a mentor
- Cloud AI site
- Cloud AutoML site
- Cloud Vision API site and docs
- Cloud Speech API site and docs
- Cloud Natural Language API site and docs
- Cloud Translation API site and docs
- Cloud Machine Learning Engine docs
- TensorFlow site, github and Dev Summit waitlist
- ImageNet site & Kaggle ImageNet Competition site
- Stanford Medicine site & Stanford Children’s Hospital site
Additional sample resources on Dr. Fei-Fei Li:
- Citations site
- Stanford Vision Lab site
- Fei-Fei Li | 2018 MAKERS Conference video
- Google Cloud’s Li Sees Transformative Time for Enterprise video
- Past, Present and Future of AI / Machine Learning Google I/O video
- Research Symposium 2017 - Morning Keynote Address at Harker School video
- How we’re teaching computers to understand pictures video
- Melinda Gates and Fei-Fei Li Want to Liberate AI from “Guy with Hoodies” article
Question of the week
Where can I learn more about machine learning?
Listing of some of the many resources out there in no particular order:
- How Google does Machine Learning coursera
- Machine Learning with Andrew Ng coursera and Deep Learning Specialization coursera
- fast.ai site
- Machine Learning with John W. Paisley edx
- Machine Learning Columbia University edx
International Women’s Day March 8th
International Women’s Day site covers information on events in your area, and additional resources.
Sample of recent women in tech events to keep on radar for next year:
Where can you find us next?
Mark will be at the Game Developer’s Conference | GDC in March.
MARK: Hi. And welcome to episode number 117 of the weekly "Google Cloud Platform Podcast." I'm Mark Mandel, and I'm here with my colleague, Melanie Warwick. Melanie, how are you doing today?
MELANIE: I'm great.
MARK: Thanks for joining me for yet another week.
MELANIE: And still in the actual studio.
MARK: Yeah, you're still actually in the studio, though I know you've been running around a bit. We have, I think, one of your favorite people this week.
MELANIE: I know. I am super thrilled. Our interview this week is with Dr. Fei-Fei Li, who is the chief scientist of AI ML, Machine Learning, at Google Cloud. She's a visionary in the space, and it was incredible to be able to have her come on. So I'm looking forward to that interview, sharing that with the public.
MARK: Yeah, it was really great, sitting there, and listening, and understanding at least about a quarter. So it's great.
MELANIE: You hung in there really well.
MELANIE: You did a good job. As always, we start with our cool things of the week, and then we end with a question. And so, for this week the question is actually-- if you want to learn about machine learning, what are some-- and we'll list a couple of options for you to explore. But before we get into that, let's talk about the cool things of the week.
One of the cool things of the week is that there's this new nonprofit that was started called Partnerships on Artificial Intelligence. And Terah Lyons, who's new, as a founding executive director of this group, is basically leading this group to help ensure AI is applied in ways that benefit people in society. To quote her, she said, "We will have to work very hard to make it so that the benefits of what we are creating are broadly distributed and that technologies like AI are applied to solve grand challenges instead of being wielded exclusively by privileged and narrow special interests."
What's interesting about this is that this specific nonprofit was pulled together by Apple, Amazon, DeepMind, us, Google, Facebook, Microsoft, IBM.
MELANIE: And they've got board members that have representatives from universities, foundations, and groups like the American Civil Liberties Union. So it's a really interesting organization that's come about and is pushing for AI ethics to kind of throw back to our machine learning bias and fairness discussion a few weeks back. So that's one cool thing.
MARK: Nice. I am going to mention something that actually I think I've been wanting for a really long time. So if you use Cloud Datastore, fully managed export and import with Cloud Datastore is now generally available. If you've ever used this tool probably for a long time, you'll remember the old admin tool that was existed-- it was sort of a cron job and a [? bunch ?] of jobs. It wasn't the best experience one could possibly have had. I'll put it that way.
No, it was good. It did the job, but this tool looks to be a whole lot better. You'll have a newer API. There's integration with IAM. And previously, like, you were looking at much longer times for doing backups. These backups should be happening a whole lot faster as well. So if you want to check it out, we'll have a link in the Show Notes to show through how you can export and import entities through the documentation. It looks like a much nicer and easier process than what we've previously had, so I'm super happy about that.
MELANIE: Yes. And there's also a blog post out there on how Color, the company, uses the new variant transforms tool for breakthrough clinical data science with BigQuery. This is interesting from a genomics perspective for those out there who are in that space. So Color is a health services company that offers affordable and accessible genetic testing, and they wanted to be able to do more heavy data mining on the data that they have to better understand their client DNA.
So variance is a popular task in genomic analysis, and it's basically identifying a variant of a specific genome from some reference genome. And so, BigQuery was something that they went with because apparently it's very mature in terms of its ability to handle the type of data sets that they're working with, because it's a fully managed data warehouse, and it has strong genomic data import mechanism. And the variant transforms is a recently released tool that's basically-- and it's an open source tool. It's out on GitHub, and we'll have a link.
But it's basically able to handle massive amounts of genomic data directly and load it into BigQuery for you, and it works with Datastore too, as far as I understand. Well, yeah. We'll post that out there for you guys.
MARK: Nice. Lastly, one of our teammates, Eric Schmidt-- not that Eric Schmidt-- different Eric Schmidt-- developer advocate, Eric Schmidt.
MARK: I feel very sorry for him for that reason.
MELANIE: Although in some ways it could probably be a plus, too.
MARK: Yeah, pros and cons, right-- just pros and cons.
MELANIE: Maybe he gets into places better.
MARK: Yeah, maybe he does. Google Cloud and NCAA team up for a unique March Madness competition hosted on Kaggle. I believe that's basketball.
MELANIE: Yes, that is basketball. You and I both are not sports people apparently. Anyways, yes. And this is their fifth annual one that they've done, where you're building up models to predict and, I guess, build out your March Madness.
MELANIE: So that was launched actually last week and closes March 15 in terms of submitting your predictions and submitting your model.
MARK: And there's a $100,000 prize pool.
MELANIE: Yes, the prize pool-- and to note that that is split out over, I think, first place is $25,000, second place is $15,000, and third place is $10,000-- which actually, when you think about it, doesn't really add up, so I wonder where the rest of the money goes. But anyways--
MARK: It's fine. It works out.
MELANIE: It's fine. It's something like that. So yeah, you should check it out. We'll put the link in there if you are a big basketball fan and also love data.
MARK: Yeah, awesome. All right, well, why don't we go chat with Dr. Fei-Fei Li?
MELANIE: So today's podcast, we are excited to have join us Dr. Fei-Fei Li, who is our chief scientist of AI and ML for Google Cloud. And she is the associate professor in the computer science department at Stanford, who is currently on sabbatical. So welcome.
FEI-FEI LI: Thank you.
MELANIE: And before we get started-- we know about your background, but we would like to have you tell, in your own words, who you are, and what you do.
FEI-FEI LI: I guess I consider myself an AI technologist and researcher foremost. I've been a professor for most of my professional career. And before taking the sabbatical, I've been the director of Stanford AI Lab. And here at Google, I am chief scientist of AI and machine learning, overseeing bringing AI technology to the enterprise world through Google Cloud.
MELANIE: That's great. And in terms of doing the work that you're doing at Google Cloud, what are some of the things that a chief scientist-- do you do?
FEI-FEI LI: It's an incredibly fascinating world. Here, Cloud is the biggest computing platform that humanity has ever invented to deliver computing service to people in businesses. So on this platform, our goal is to really democratize AI services and products to as many businesses and as many people as possible. And so Google Cloud offers a suite of AI products-- from APIs, to customized models, to more closed collaboration and partnership services-- to really try to enable businesses to solve their critical problem through AI solutions.
We also, in the meantime, advance AI technology, because as you get close to solving real-world problems, new challenges arise, and that inspires the development of new AI technology. And we work closely with Google's AI research and development teams throughout the entire company to advance AI as well.
MELANIE: Nice. And I know that there was a recent release of, like, AutoML, and [? other's ?] been the advancements with TPUs and TensorFlow. What are some of the things that you've seen, especially from, like, applied AI, that you think are very relevant in terms of what's important for businesses to be aware of?
FEI-FEI LI: Yes, so that's a great question. I want to first take AutoML as an example, because that's a product that really, truly excites me. It bears the meaning and mission of democratizing AI beautifully. We see that more and more businesses would like to use AI as a service to improve their products. And much of the power comes into making predictions and doing smart analytics of the data the business have.
But it used to be-- still is true-- that AI's a pretty technical field. It's also young enough that not every business has the kind of AI talent that Stanford or Google has, so the barrier of entry is pretty high. So we've been thinking about how to lower the barrier of entry for businesses to use AI. So let's take Images as an example. If you are a real estate company, maybe you want to tag different kinds of houses or rooms. If you are a retail clothing company, you want to take an image and tag shoes or different kinds of shoes or clothing. If you are a nonprofit tracking wild animals, you would like to take photos and tag leopards versus giraffes. So there is a lot of need for this kind of AI work.
Well, one way to solve the problem is through what we call APIs. These are pre-trained models that Google engineers have built that work fantastically, and companies can just upload their images and get, say, a label from the pictures using these API models. But there's one thing that API doesn't do for you, which is customization. The API results come from a pre-trained model that has a pre-designed set of labels. It's a lot of labels, thousands of labels, but it's still pre-designed. But if you truly care about just tracking wild animals or care about satellite imagery, you are now in the realm of the need for customization.
And in fact, lots and lots of businesses and people need customization, and they don't have that capacity to do it from data curation all the way to creating the model. So we saw that as an important opportunity and created this new product called AutoML, starting from image understanding or image tagging. It's really almost a plug and play-- that customers can give us their data that they want to label, and we build a specific customized model for them to use. I'm not going to get into the details, but the model is built on some of the cutting-edge Google AI technology.
MELANIE: This must be really exciting for you, especially considering I know you started out researching around computer vision-- and to see where it has gone from when you started to where we are now, and how it's being more democratized as it's being made more accessible. And I know you were one of the inventors of ImageNet, which is this well-known competition that was really pivotal in the advancement of computer vision, which came to a close last year.
And one of the questions I wanted to ask you as just sort of a sideline is, how do you feel, seeing something that you've built that helped advance a space that matters so much to you? How do you feel now that it's closed? What does that mean to you?
FEI-FEI LI: So first of all, I think ImageNet is generally a group effort. Many of my colleagues' students are part of the core creation team of ImageNet, and as a team, we are very proud. We're proud that ImageNet has made a contribution-- computer vision. It is one of the driving forces of what we see as the deep learning revolution. And closing down the competition in the academic world is really just the beginning of a new chapter.
First of all, Kaggle is now hosting ImageNet, so the democratization of ImageNet data and this competition itself lives on, and it's reaching to more developers. It's also a sign of the progress. Scientific discovery and innovation is always built upon the continuous renewal of ideas, and ImageNet served a important historical purpose. When we decided to finish the academic competition, we recognized the field is ready to embrace new ideas, so that's where we are with computer vision and machine learning.
MARK: For those people who aren't as familiar with, say, ImageNet, can you explain a little bit about what that is? And also the transition that it's gone through-- you said it's gone into Kaggle now.
FEI-FEI LI: So ImageNet was a project my students, collaborator, and I did starting in 2007. The end product, by 2009, is a gigantic image data set of 50 million images labeled-- over 22,000 categories of objects and real-world objects and things and concepts-- and publicly released to the world of AI research and education. The goal of making ImageNet was to reboot machine learning research and rethink the relationship between algorithm and data.
It was built at the time that data was not as recognized as the key ingredient of AI systems. ImageNet's contribution, I think part of it is, because of that, it enabled the family of algorithms called neural network to show its potential. And because of that effort, we think it was important to encourage the academic and research community to benchmark their algorithms on ImageNet, so we created the ImageNet competition starting 2009-- actually, starting 2010, and that competition served as a good platform for international teams to benchmark the progress we're making in computer vision.
And the transition in 2017 is that we recognize, just as I said before, we've come a long way. The problem that ImageNet set out to do, object classification and detection, is I wouldn't call largely solved, but we've made so much progress that we need to create a space for new problems and new ideas--
MELANIE: --to inspire new ways to solve these types of problems or to solve different problems.
FEI-FEI LI: Exactly. That's why we transitioned it from an academic competition to a Kaggle data set.
MELANIE: So in terms of AI, I've seen you talk about how-- you like to emphasize that it's not artificial. Why the emphasis that it's not artificial? And is this related to-- also, in reaction to AI hype, do you have any thoughts around that that you'd want to share?
FEI-FEI LI: Yes, I now tend to say that I share with my students and colleagues that the field of AI is nothing but artificial, and what I really mean is that this is a deeply, deeply human pursuit because AI as a technology is going to influence and impact our human lives and human society in very profound ways. And as technologists, I think it's really important to recognize that.
Humanity has never created a technology that resemble ourselves as much as AI does. And of course, part of ourselves is being able to function with the level of logical reasoning and computation, but part of being human is also recognizing the social relationships, the emotion, the wanting to do good, and all of that. And since AI as a technology is so powerful and impacting so many different sectors of the world, we want to recognize its human impact and also prepare ourselves to AI's human impact.
MELANIE: So off of that human impact and also off of what you were saying about ImageNet and trying to look into the next phase of what research would look like-- I know you've also talked about, sort of, the new goalposts being around natural conversation, collaboration, emotional perception. What do you think, in terms of where things are going, and what types of benchmarks are needed for those types of goals?
FEI-FEI LI: Yeah, that's a great question. I think that if you look at the complexity and the intricacy of human intelligence, being able to classify or group concepts is only one of them. That's mostly what ImageNet did. But whether it's our visual ability or our navigation ability or our natural language communication ability or comprehension, reasoning, emotion, humans are really complex. So I think in order to think about the next step of AI, it's really important to go beyond the simple tasks that we are starting to excel at but to recognize the complexity of intelligence.
There are already some interesting budding research. Google and many other groups are looking deeply at things like for example, natural language dialogue. For me, that's a very interesting area, where you go beyond just labeling parts of sentences to actually enabling a machine-human dialogue in meaningful ways. For our vision, starting to tell the story of a picture, starting to reason on a picture, starting to understand the events, the human activities, the object functions and all this would be the next step.
MELANIE: And take what we've excelled at and do what you've been talking about in terms of democratizing it and making it accessible for people who use it.
FEI-FEI LI: Right.
MELANIE: So your research, one of the areas that you've been focused on lately, is more in the health sector. Can you tell us a little bit more about what are some of the things that you're exploring in that space?
FEI-FEI LI: Right. So you're referring to some of my collaborations with Stanford School of Medicine and also Stanford Children's Hospital and adult hospital. So a few years ago we thought deeply about AI transformation, and one of the most exciting transformations that everybody hears about is self-driving cars. And this is the result of advances from sensing technology to perception technology to machine learning technology and all that.
And then when we recognize that there is actually a parallel opportunity in health care that has more analogy to self-driving car than most people recognize-- it's really about the space of care. It's the how do we enable care in a more effective way? For example, in hospitals, there are a lot of things going on, and work could be more efficient. Errors should be avoided. Patients can be monitored further. Or in elder homes, can we enable more intelligent, ambient sensing and machine learning to help families and clinicians to take care of our elders and all of that? And we start to recognize, wow, this is the kind of technology that's needed in these potential services.
It's very similar to self-driving cars. It uses smart sensors to gather information from the environment. And then back end, it uses computer vision and machine learning techniques to recognize activities and to make decisions. So we started to collaborate with Stanford Hospital and also a senior home facility in San Francisco to explore how we can use AI to help avoid health care workflow errors or to help aging seniors, especially early detection of dementia and all that.
And a specific example is a project we're doing with Lucile Packard Hospital at Stanford, where we put smart sensors or depth sensors in the hospital unit to help doctors and nurses monitor their hand hygiene practice. It turned out the lack of hand hygiene or hospital-acquired infection is one of the leading killers of our patients in hospitals. It kills more people than car accidents per year.
And how do we continuously monitor the behavior of proper hand hygiene practice is actually a very difficult task, and it's expensive to deploy people to watch people. And that is very sparse. It introduces bias. So we thought, the same kind of depth sensor a self-driving car uses to watch the pedestrians and the cars on the road could be actually put into the hospitals. So we installed these smart sensors that are also privacy preserving and use computer vision technology in the back end to be able to continuously monitor and observe hand hygiene practices in the hospital unit.
And that was one of the first kind ever done in that way, and we're very excited to see some initial results showing that an automatic observing system using AI does as good, even better, actually, than human observers, but it doesn't get tired. It's not biased and all of that.
MELANIE: That would be great. And this was a recent implementation, so the results haven't come back just yet.
FEI-FEI LI: We actually-- it's in submission to a publication.
MARK: Nice. I'd love to take a slightly different tack here. I mean, this work sounds fantastic. But you've been in the AI and machine learning industry for a really long time, and you've watched it grow. And you were talking a bit about barriers to entry and, like, the speed at which it's progressing.
If people are coming in new to machine learning and AI now, and they're interested in getting involved, where should they go? What should they do? Like, what path do you think they should take?
FEI-FEI LI: Yeah, good question. I get that a lot as a professor. So I think it depends on what you're interested in. If you are interested in machine learning as a research, then I think it's important to get involved in the research projects. I assume that's when you have already some technical background.
If you're interested in machine learning implementation and software engineering, there is already a lot of online courses, and Google's TensorFlow and the TensorFlow community has a lot of materials that can help to get your hands on to the coding and software engineering of the site. If you're interested in using machine learning as a way to create interesting startups and entrepreneur opportunities, start thinking about the pain points of different businesses and products and find areas where using smart data analytics or machine learning techniques can help solve a pain point.
But machine learning and AI is not just technical. That goes back to my original point. There is actually a lot of related areas. Machine learning and policy, right? Or AI and education, AI for sustainability, AI for economics-- there's a lot of now new areas that are at the cross section of AI technology and human impact, so a lot of people, even without technical background, can get involved and contribute.
MELANIE: You know, I know that you get asked a lot about the AI-- sort of what are the bigger challenges, and you spearhead efforts around diversity inclusion. And you have this wonderful organization, AI4ALL. Can you tell us a little bit about that and what it does?
FEI-FEI LI: Yes. So AI is a field that largely stemmed from computer science, and I think it's pretty obvious by now that the general field of computer science and AI has a critical issue of lack of diversity. We lack women. We lack under-represented minority. All the publicly released data, whether it's Silicon Valley companies or academia, we're seeing a very small percentage of women and even smaller percentage of underrepresented minority in this field. And I think that is a pity, because involving more diversity would make our field better.
And we talk about several reasons. One is for economic and labor reason. We need more and more CS and AI talents, and we need to tap into all walks of life. It's also an innovation creativity reason that when a diverse group of people are coming together, solutions and ideas are more innovative. And it's also a fairness and justice reason. When we want the technology to resemble or to represent the values of our collective society, we want all walks of life to be involved in developing and creating it.
For these three reasons, at Stanford about four years ago, we started this program to encourage diversity group of high school students to participate in the study and early hands-on research of AI. And we created this program called Stanford AI Lab Outreach Summer. And at Stanford, we were focusing on high school girls around 10th grade, and we bring them on campus for two to three weeks in the summer. And they study AI. Also, they study AI and its human mission and human-centered impact.
And that has been, very luckily, successful, so we decided in 2017 to expand it to a national nonprofit organization-- we call it AI4ALL-- to really replicate the Stanford model and create summer camps nationwide. And so we started with Berkeley last year, where they reached out to low-income families and students, focusing on robotics. 2017, we're expanding to Princeton to under-represented racial minority students to study AI and policy, and we are expanding to CMU this year as well to reach out to rural population.
And so we're expanding to six different sites. 2018 now, we'll continue this effort.
MELANIE: And we plan to include, in our Show Notes, too, ways for people to link in and find out how to get involved.
FEI-FEI LI: Thank you.
MELANIE: I want to make sure-- is there anything else that you wanted to share or talk about from AI, from the research, from recommendations for people to think about how they get involved, how they can get into the space?
FEI-FEI LI: AI is incredibly exciting. I guess I'm biased, but being an AI technologist and also educator for 20 years, I think seeing this field has grown from a niche academic field to become one of the biggest drivers of the fourth industrial revolution is really extremely exciting and rewarding. And in this process, I think we are just at the beginning, even though we hear a lot about AI and its progress so far, but we really are just seeing the beginning of the power of this technology.
So I would love to see more people getting involved and different parts of the world-- technologies, technologists, humanists, policy makers, social scientists, law makers-- all come together and have deeper and multi-way dialogue and to design this technology and also guide the impact of this technology.
MELANIE: Great. Thank you.
FEI-FEI LI: Thank you.
MARK: Yeah, thank you so much.
MELANIE: Thank you again, Dr. Fei-Fei Li. We are really grateful that you were able to come on and talk to us about your work, what your research is, what your experience was of how you felt about ImageNet. All of it was wonderful. And if anything, there were so many other questions we still wanted to ask. We were limited on time, so we worked with what we had.
MARK: Yep. But you're welcome back any time.
MARK: Definitely. So why don't we move on to our question of the week? And I'm going to ask you, Melanie.
MARK: So I am interested in AI and ML, in theory, and I would like to learn more. Where can I go?
MELANIE: You mean, I haven't beat it over your head enough in this podcast? So there's a couple of resources, and we'll definitely add those in the Show Notes. And I'll probably come up with some more when we post this.
But one in particular that comes to mind is-- most people will talk about Andrew Ng famously has a Coursera course that's out there that many will go to on a regular basis when they are initially starting out.
MARK: I've done half of that.
MELANIE: Sweet. Good job. Maybe you'll get to the other half at this point.
MELANIE: fast.ai is also a group out there that has some wonderful content, especially in terms of deep learning and just helping to really break down the basics around some of the math and the machine learning in general. Google just launched a course on Coursera, saying how Google does machine learning. So we will also include that in our Show Notes, but there's definitely a number of resources. These are some key ones that come to mind, and yeah, we'll make sure we share what we have.
MARK: Nice. It looks like enrollment for the Google one starts on March 12.
MELANIE: Good point. All right, well, I think that's it for this week. And anything coming up for you? I don't think you've talked about anything that you had this month.
MARK: No, it's GDC. It's all GDC all the time.
MELANIE: That's all you do?
MARK: Yeah, it's all do. So come say hello. We'll be at the booth. I won't be at the Developer Day, but there is a Google Developer Day that's going to be talking about a bunch of different Google products on the Monday, but I'll be at the booth various times from Wednesday to Friday. We're sponsoring Women in Games International, so I'll be at that party as well. That's probably about it for the moment. I'm sure I'll have some more stuff coming up.
MELANIE: Awesome. Well, Happy International Women's Day a day early since we'll be releasing this on the 7th, and I hope everybody out there has a good one.
MARK: Yes, absolutely.
MELANIE: And I think that's it for us for this week.
MARK: Yeah, for this week-- so yeah, Melanie, thank you for joining me for yet another week on the podcast.
MELANIE: Thanks, Mark.
MARK: Thank you all for listening. And we'll see you all next week.