2019-07-18-The-Future-of-Accessibility-Powered-by-AI

# The Future of Accessibility Powered by AI
## Presenter:  Heather Dowdy  - Thursday, July 18, 2019
[Source recording](https://www.youtube.com/watch?v=xmXQchGyibY&t=302s)

**[Heather]:** 
>> Thank you for not reading all of my bio.  I don't like to hear about myself.  The Future Of Accessibility Powered by AI.
Today's agenda    Karen?
>> Is captioning okay?
Is it 6:31.
We're good.  We're good.  All right.  Want to discuss today a couple of things.  One, talk about trends.  Trends that point to the need for innovation.  I want to talk about the AI for Accessibility program.  Listening to few of you all and what you do, I think you will find it interesting.  I'm going to end with ideation time.  I'm here to share what I think is good stuff and I'm hoping to take back stuff good stuff from you all.  Best part of a meet up is hearing from you and taking back ideas.  We will have ideation time and Q&A at the end.
On this slide here, I have a picture of me and my father and beautiful Chicago cold winter weather.  This was last November.  I had a story published about me and my family and growing up.  I pretty much was born to be an ally.  I'm a child of deaf adults.  I have been fluent in American Sign Language since six years old.  My dad who is deaf.  This article is entitles lessons from south side, the dad who helped raise an empathetic engineer.  Talks about how I fell in love with technology and assistive tech that was in house growing up.  I didn't know that was going to lead me to be an year, I just knew that would help my family.
>> That has led me to a career that is in tech.  I've been in tech and accessibility ten for years now.  Started here in Chicago.  Where I am from.  I worked at Motorola for a decade and help lead our product development there.  Working with engineers and lawyers.  Pretty much everyone.  At end, only one in the company working on is accessibility.  You make a lot of friends from marketing to legal.
I heard from industry colleagues all the great things that Microsoft was starting up again and new wave of focusing on accessibility.  I went to Redmond outside of Seattle.  Talking and working to make all of our internal and external Microsoft websites accessible.
Particularly for our employees with disabilities.  I love, love, love that we are really about empowering every single person on the earth to do more.  That starts with people who are developing the products that we make and sell.  Rewarding and lots of work.  When you think about it, you and I that can see, we are pretty much don't have a problem able to download our paycheck stuff and regular things.  Your colleague that has trouble doing that is not something that should be.  I want to encourage everybody that there are lots of small steps you like to take to be an ally and champion.  Doesn't mean you have to be technical.  Ways that can improve user experience for everybody.
Now we are going from mobile to web and now AI.  I feel like AI is new frontier in accessibility.  For me, hey, this is the opportunity for us all to make sure that the future is accessible.  By being part of this big wave of creative things that are being taken up.
I'm a techie.  I love talking about innovation.  I love that I'm amongst my people here.  People that are thinking about the why and thinking about the how and looking at it from end to end.  What I have on this slide is talking about the cycle of innovation.  How many times it begins with human need, right?  And then you take technology advancements and make progress and that continues to push itself in this resolving circle.
So sounds pretty simple for a chart.  This is how technology    or innovation is done, right?  Particularly for people with disabilities we know that progress has been stagnant.  We also know there is great opportunity.  That's what I want to talk about for tonight.
For today, I'm going to talk about the why.  I think when people understand the why there is a need for something, it will be much, much easier to get everyone on one accord and will go faster.  In talking about that one part of that innovation cycle is talking about human need.  We know the statistics.  More than a billion people have disabilities in this world today.  Many of them lack access to three things.  Access to assistive technology.  One in ten people have access to assistive technology that they need.  In other countries, that is that and disparity is higher.  Something to think about.
Then, lack of education.  25% of adults with a disability have a lack of education compared to 8%.  We are thinking about how to close these gaps.  And then employment.  I'm sure that many of you all know the unemployment statistic for people living with a disability which is twice that than people living without a disability.  Did you know it's up to 80% for people with autism.  Something to think about.  Something that drives the need for innovation and there is a trend here.  Disability is something that affects all of us.
So talking about AI.  I'm really curious to figure out where we are in the audience.  How many of you kind of already feel like you know a lot about AI?  Could you raise your hands?  Teach a course on it?  I saw about    I saw hands and then half a hand and put it down very, very fast.  No worries.  Another question.  I'm wrapping up this week at disability in conference.  Anybody attend?  Pretty awesome conference.  We talked about the human need particularly for people with disabilities.
Wanted to talk briefly    don't worry.  Not going to go too much into nuts and bolts of AI.  A lot to pack in here.  Like I said at beginning, I want to hear from you all.  There is so much going on that I doubt that there are very few people that haven't heard about the word AI.  It's certainly a buzzword.  And certainly the AI is technology that can see, learn and reason.  That's what makes it so cool.  Typically, we teach machines, you know, what to do.  The thought that these machines could learn on their own is certainly a cool concept.
In the largest sense of the word, we are seeing cool applications particularly of machine learning which is a little bit different.  Speaks to the fact that there is so much data available whether that's through your local government or whether that's Google or any other tech company collecting data on accounts that you use.  So much data that can be used to fuel innovation of new and creative solutions.  In talking about some of the AI advancements, want to briefly touch on the fact that AI is not just this new thing.  It's the fact of rate at which it's advancing that has main it mainstream topic today.  You see here sat that talk about breakthroughs on object recognition and language understanding over the past three and four years and how we are finally getting to a point where we are sort of becoming on parity with a human. 
Couple of years ago if you listened to any earlier demos of bots and assistants, you could tell the difference between a machine and a human.  Nowadays, not so much anymore.  Particularly not when interacting with your favorite voice assistant with Siri or Alexa or Cortana.
That led to some progress, right?  Just talking with gentleman here earlier about seeing AI.  How many have heard about that?  Thank you very much.  I really, really appreciate it.  This is a cool app.  Proud of the fact that born out of engineers at Microsoft hacking.  Not on their free time during a company hack a thon.  That's a big part of our culture.  Speaks to whole design thinking process and how we can use technology to come up with anything.  The Seeing AI app is on iPhone.  Free.  Highly recommend for you iPhone users that you download it and tinker with it.
Helps my colleagues who are blind at work and daily life pretty much turning visual world into audible experience.  Has features such as recognizing handwriting which is cool when a teammate of mine wants to read personal notes from her husband.  Doesn't want to ask someone else to read it for her.  There is a currency channel and some other object recognition.
I've got a really quick video here.  Let's see.  It was playing a second ago.
It's loading.
>> Microsoft research project for people with visual impairments.  Turning the visual   
Donna CART Captioner:  Is this captioned?
>> App recognizes [Indiscernible].  Describes the people    including their emotions.  It 22   
>> Let's try this again.
Donna CART Captioner:  Video is hard to hear.  I hope you have captioning on.
>> People with visual impairments [Indiscernible].  [off microphone].
Donna CART Captioner:  I apologize.  I cannot hear the video to caption it.
  It's captioned.  I can see it now.  Thank you.
Standing by.  (Video playing).
>> All right.  Taking a second to get back to presenting mode.  Cool app.  I love that everywhere I go, people are telling me how they use that app which I think is a really important thing.  One of the features that I like is with bar code scanner, actually provides audible queues when you are getting close to the bar code.  Which is important.  Just because you have that feature and you have a bar code, how does a person who is blind actually know how to position the camera?  Those details matter in term office user experience.
All right.  Come on.  Thinking way too much.
All right.  Let me kill PowerPoint. 
>> While I'm doing this, that's one example that has been cool to implement using some prebuilt Azure services called cognitive services.  How many of you use sub titles and captioning during presentation?  Five hands.  Maybe six.  I highly recommend that you turn it on.  Because it's definitely improving a lot.  It's not on par with having a live person captioning yet.  That's still something that we recommend for events but could be useful in meetings, yes.  I see a hand?
>> Is it transcribe everything now or not?
>> Not yet.  But that is a really good proposal that I will talk about.
>> Yeah.  All right.  Let's get back into this.  And really simple to enable, you go to the slide show ribbon and then you can select an option to make that on always.  Other advances, just during video, we have Microsoft stream which will provide automatic transcripts.  We use this for town hall streaming videos which is cool.  I like it because I don't have to remember everything that was said.  I'm an audio and visual learning.  Exception benefit that helps a lot of people.
And this is another hack that came out of Microsoft.  This solution helps students with dyslexia.  I like it because has some important features in terms of playing language.  Provides suggestions.  Also breaks up some of the words into syllables as students are following along.  In terms of providing access, this feature being used in millions of classrooms has led to increased test scores by at least 10 points.  Really important in terms of results.
>> And thousand I want to bring up my    now I want to bring up my special friend, Adam, from our Chicago off to talk about, okay, those are some of the things that Microsoft is focused on in product level, let's bring it more local.  Let's talk ate partnerships that Microsoft has with Chicago.  Better put this on too.
>> Okay.
>> Can I touch you?
>> You may.
>> Hi, thank you, Heather, and welcome home.  Appreciate having you here.  My name is Adam.  I'm with Microsoft in the Chicago office.  We have been doing a lot of different things in Chicago in accessibility and inclusion space.  I think if there was one interesting thing that I would like to talk about is this one over here, G3ICT for smart city digital inclusion.  Some of you will recognize some of the people in this picture here.  Pat, Danielle and Karen who a lot of people know as commissioner for mayor's office for people with disabilities. 
What this project does, if you think about cities today, every city is trying to become a smart city.  And they are also trying to become digital smart cities which means that they are trying to move more and more of their services to be digital.  What this does is this makes sure that as cities are moving to become digital smart cities, they are inclusively digital smart cities.  What boils down to is we are working with G3ICT to build a digital inclusion maturity model so that cities can, first of all, benchmark where they are in terms of digital inclusion on maturity models and specific advice on how to move up to maturity models.  Karen was great to look at all staff in her city to be interviewed and benchmarked.  We were doing this simultaneously in Chicago and New York in hopes when finished with this, this will be a great tool of any city of any size to make sure they can bring anybody along when becoming digital.
[Applause]
>> Thanks, Adam.  I love him so much.  We have some amazing teammates.  Lots of folks working on accessibility which is cool.  Other picture on this slide is screen shot from a blog article from University of Illinois which I went to school.  It is talking about the accessibility lighthouse program.  Don't know if you have heard about it, announced late last year.  There is a partnership with Microsoft and UIEC by making classrooms accessible by bringing in live captioning and setting up the software in the room so that all students can get what they need.  There will be more coming out of that.
I don't know.  How many people went to U of I here?  Okay.  Three.  Not a lot of people.
>> [off microphone].
>> Oh, UIC.  Cool.  I remember for me when I was going to University of Illinois in Urbana, Champaign.  Not easy thing to navigate the accommodations and accessibility.  To see the effort to make UIC one of most accessible universities in the country is refreshing and about time as well.
I've already talked about, okay, few of things Microsoft is doing in terms of products.  Now I get to talk a little bit more about this grant program.  And a lot of that is behind the mission of Microsoft which I've already stated is how ever person and every person on planet to achieve more.  We are leading in a lot of ways as part of that.  Perhaps you will notice that Microsoft is big on privacy and security.  And that doesn't stop when comes to when you compare the different cloud technology providers.
This slide here just talks a little bit behind the bigger picture with AI.  You have a lot of these cool solutions, right?  The other side of that coin, lots of people talk about, well, you know, how should AI be used?  Will it take of jobs of captioners for instance be having this embedded into PowerPoint and other tools.  There are lots of components to think through when creating any type of solution with AI.  These are categories that need to be top of mind as far as what Microsoft put forth.  That's fairness, reliability, privacy, inclusiveness and transparency and accountability.  This could be a slide in itself.  In AI for accessibility, we are driving home the inclusive part.  We want Microsoft AI platform to be the most inclusive platform there is. 
That's why we started really thinking.  We thought about the question, how do we want the world to look in five years?  Microsoft is leading on a lot of different fronts.  And it would be nice to just say, hey, we are incorporated AI into our products.  That's good enough.  It really is more than that.  And it's time I actually dream again.  That's the point of technology breakthroughs.  When we come up with this technology, accessibility is just opens up doors for new use cases that we have not considered or not thought of when that technology was developed.
How do we want the world to look?  I want to make sure my captions are on this time.  I'm going to play a video that talks about one of the problems that we were really excited to solve using AI.
(video playing that is captioned).
>> Cool.  So that was Zyrobotics.  It's a start up in Atlanta, Georgia that is using some of our cognitive services, specifically custom speech to create a STEM based reading fluency app for students that have speech disabilities.  What I love about that particular product and what they are shooting for, they are looking at the fact that, hmmm, it can be pretty expensive to have occupational or speech therapist.  And particularly in areas where there are students of color, or other areas where they have a lower socioeconomic status.  How do we scale the need for therapists with all of these students?
They are using AI to be able to provide feedback to a student when reading in order to highlight exactly what partner struggling with.  What part they missed.  That feeds into a dashboard that the teacher as well as the parent can track together and talk about how exactly the child is progressing in their fluency.
>> All right.  So in thinking about how we want the world to look in five years, Microsoft came up with this AI for good.  This umbrella programs that AI for earth, AI for accessibility and humanitarian action.  In particular, today, I want to focus on AI for accessibility.  With AI for accessibility program, we are here to empower start ups, researchers, entrepreneurs, NGOs.  Literally anybody that has an idea or wants to create a solution to empower people with disabilities and one of our focus areas of employment, daily life, communication and connection.
How we're going to do this and how we started to do this is Microsoft invested $25 million over the next five years.  In this program and engineering resources, in terms of providing technology grants.  Providing some of the knowledge and expertise of Microsoft engineers to help build these solutions and using the big and vast Microsoft network to scale.  Which is absolutely amazing.
Yeah, you    kind of showed you briefly about some of the thing we are building and there is a lot more, particularly in accessibility space.  We really think that having people with disabilities and having people and start ups that are rooted in the disability community willing to work with people with disabilities to design solutions with and by, that that will be the next breakthrough in terms of progress.  Particularly for me, I don't know for y'all, having done this for ten years, I'm tired of talking about the unemployment rate.  It has been the same number for a long time.
This is an opportunity to use technology to accelerate that.  That's why I'm super, super excited to hear from you later on.  And think about solutions and perhaps people in this room that might apply for a grant.
So there are two types of grants.  There is one that is fast track.  Azure credits only.  Using our cloud platform.  I can talk more about this later as well.  Using our cloud platform, you know what you want to do.  You know how much you need.  This is compute power to get it done.  We have Azure credit only grants.  We have Azure credits plus cash.  That goes towards a data scientist perhaps to collect data or something related to engineering related work that is needed to build your solution or your idea.
How do we evaluate grants?  A major tenant has to be people with or by people with disabilities.  If you are not technically savvy, match up with someone that is.  Perhaps you are.  Match up with disability community to make sure what you are designing for meets the need.  Other criteria that I have listed on the slide.  Microsoft is giving away money.  That's nice.  Must be really, really hard.  It's like, I'm telling you, every single thing we evaluate on as part of leading business operations.  We look at all of the applications, we have a board.  And so we really do read every single one.
Believe it or not, it's AI for be accessibility program.  Some of those ideas don't need AI to get implemented.  That's what we are after.  We are wanted to prove what can be accelerated through use of AI technology.
Employment, daily life, communication and connection.  How are you going to empower people with disabilities in one of those focus areas?  Data.  Data, data.  Machine learning requires lots of data.  And the truth is, particularly in the disability community, there isn't a lot of it.  I tell people that it's really easy for me to have lots and lots of pictures of a plant for instance.  It can take longer for me to sit down with an individual that has a speech disability and record some speech data.  Two different things.  And, again, we are looking for more and more data in order to be this innovative platform that others can build off of.  Not only are we looking for folks to bring forth the data, we are looking for them to be collaborative.  Don't hog it for yourselves, you need to share it.  That's how we are going to go further together as African proverb says.
Each grant is for one year.  Possible to renew the grant.  Whole review process for that.  What can you accomplish in one year?  What will you have to show at the end of that year?  And will you idea live beyond that year?  And capable and showcasable.  Capable, do you have the skill set on your team or do you need to pair up with someone else?
I feel like you guys are super quiet.  I hope you are taking this in and thinking of an idea.  Again, this program is open to    normally when I say $25 million in the room, there are a lot more excited people.  I don't know if you ate too much or it's evening or the eat.  One of those three things.  All right.  I got to remember where I'm at.  That's all right.  That's okay.
I just wanted to talk about awe few example projects.  Which is cool that I'm getting to talk with you all about it.  Taken some time for us to be able to talk about it.  One project is from Vanderbilt University.  At their center for autism.  It's a big network with businesses and young adults with autism as well as the university.  So their solutions and their project is to build a virtual interviewer that would help in preparing for job interviews by providing a mock job interview scenario.
With this virtual interviewer, they would be able to adjust questioning based on if applicant with autism is perceived to have a difficult time, not making eye contact.  Being able to provide that realtime feedback in order to help that individual prepare to individual for job well.  That's taking in all of these different sensors and data from the actual person with an autism while interviewing.  What I love about it is, it's not saying, you should adjust.  No, interviewer, you adjust to the person that you are interviewing.  Really cool about that.
Then there is    that from the Vanderbilt University since that job interviewing, that is employment.  Next is future project in daily life.  From eye and ear.  They are with Harvard Medical School.  They are solving the last 3 on feet problem    30 feet problem.  You have a smartphone app that can recognize objects.  You need to get on a bus and you are blind and you thought that the bus stop was right in front of you.  That happened to be a bench.
That bus stop is a couple more feet away.  And you miss your bus by a couple of feet literally.
And so they are using computer vision in order to help people who are blind with navigation and particularly, they are looking at being able to navigate to bus stop signs in 12 cities globally.  Really cool stuff.
Each of our grantees are using different number of combinations when comes to technology    the Azure.  Some are adding it to an existing app or product and then running with that.  Not a lot of work    extra work needed.  Azure has done a lot of work for you.  I'm going to encourage you all, there is a link towards the end, you can set up a free Azure account today and tinker with demos that we have online.  Lots and lots to play with.  Some are using our machine learning studio.  Some are using the custom Azure machine learning and some have own machine learning algorithms.  Lots of flavors when comes to expertise with machine learning.  How much you want to fine tune, how you structured your data.  We can talk a lot about that.  I wanted to give you a flavor for that.
We have cool products as part of the Azure suite.  We have grantees that are working on the chat bots.  I was talking about that earlier with someone.  I'm going to pause here.  Just make sure everybody is still with me.
Does anybody have any questions or something in particular you want me to cover when we are talking about the AI itself?
Question over here.
>> [off microphone].
>> Question on the floor was, are these products accessible by screen reader users?  Very good questions.  Most of cognitive things are.  They have people with disables build on Azure, we are able to give the feedback back to Azure team to make sure it is the most inclusive platform.
>> [off microphone].
>> On MacBooks?
>> It's agnostic.  Every platform.  Yeah.  I don't go to    you can go to the Azure portal and consume the technology that way.  Question here.
>> [off microphone].  My understanding is that a lot of machine learning, people use the tensor flow API to build their applications.  Would you be able to use your Azure credits to run [Indiscernible] Microsofts [Indiscernible].
>> Question for those on the call, can you use tensor flow for machine learning solution or do you have to use Azure products?  Yes, you can use tensor flow.  A lot of connectors between the different platforms.  Azure is open to tensor flow as well as a few others as well.  Yes is short answer.  Question in the back?
>> [off microphone].
>> So the question was, how have these products been user tested?  These that I'm showing right now on the screen are sort of the building blocks from Azure.  When comes to grantees themselves and the product that is    products that they are making with this technology, yes, we are very much encouraging them and connecting them to the disability unite if they don't already have some connections for user testing.  That is very much part of their 12 month timeline.
And a question over here.
>> [off microphone].
>> Great, question.  So the question was, he noticed that we give a lot of grants to universities and do those grants    are they just for professors or do we give them to fellow students or fellowships?  The answer is that anyone can apply and anyone can get a grant.  What I will say is that given how universities kind of structure who can work on a project, typically there is a lead professor.  That's just how the university does it.  That's not necessarily a requirement to us.  We have given grants to individuals themselves, to students themselves.  Anybody who has an idea executive capability to implement it, we're open.  Open.  Completely.
Yes?
>> [off microphone].
>> Very, very good question.  Overhead costs which is what I deal with a lot in separation    operations.  The grants that we give to universities do think account for overhead costs.  In some cases, yes, if we need to, we will account for that.  We have    when there is an application process that I will talk about too.  Give you the link.  I will let you look it up yourself.  That's where you capture some of that information in terms of what budget you are asking for.
Two sides before is example highlighted projects.  There are many others on the websites and video that I showed with Zyrobotics, that's a start up.  Not all universities.  Those are highlighted products to give you an idea.  Over here.
>> [off microphone].
>> You guys are asking all the questions in my wheelhouse.  Wait until I go back and tell my team.  Question is, in exchange for grant, how does IP work?  IP belongs to grantee.  Microsoft does not take the IP.  If there is a special case where we enter in special agreement or collaboration, that is a separate contract.  We want to empower the community to provide solutions.  But IP stays with the grantee.
Question here?
>> You can access products without [off microphone].
>> Oh, do you need a grant to be able to use Azure?
>> No, you do not.  There is free account to tinker with.  At a certain point depending on how much of Azure you need whether that's virtual machines or lots of stuff in storage to get in to, pretty nice to have it paid for.  That is the point.  Never heard anyone turn down free money.  Question up front?
>> Maybe in the requirements.  The grants that are issued, is there a typical range and dollar amount that you award it also    can the grants    are they typically for non profits or can be grants [Indiscernible]   
>> Oh, great question.  See, you all were paying attention.  You were giving me resting face.  You answered all the right questions.  Question was, is there a range for cash part of the grant options and then also do we give money to NGOs or organizations non profits.  Range is funny one.  We don't public ranges    publish ranges.  If your idea is good enough, bring it.  A lot of people hear $25 million and say, hey, I will ask for $2 million of that.  It's like "Shark Tank".  You don't go on "Shark Tank" without having done your numbers.  Eat you up.  You just really have to document exactly how much you need and why.
Particularly for Azure for you question too in the back, there is an Azure calculator.  You don't need a lot of money for the Azure part.  Engineering resource related things, you might.  Collecting data.  We believe in paying participants.  Part of cash can be used for that.  And in terms of non profits, yes, we do, again, give out money to anybody and everybody.  Non profits are included.  If you have capability, sure, submit an application.  If you don't have the technical capability, make sure that you partner with someone else that does.
Question in the back.
>> [off microphone].
>> Question was be a the Zyrobotics project and asking about results.  How kids faired since the release of that product.  Quick answer is that they are in development.  It's an iterative process.  We pretty much closed out our first year of operating the program.  And so there are about three to four cycles of awards per year.  They have a year and different start points.  They are still working on it.  That's part of our timeline and plan with them is that we get specific on metrics.  If you can't measure it, you didn't do it.
One question here and then here and then I will keep going and you will keep me on time.
>> If did the grant ideas have to be unique.  If someone else has already done that idea, next people that apply less likely to receive grant?  Second part, if there are grantees that need essentially the same data.  Way to coordinate that data can be shared...
>> Thank you, excellent questions.  If there are already something out there similar to your project professional, will we fund it pretty much?  Does need to be unique.  We want to have lots of proof points that we can make some serious progress with technology.  And what I will say that I was going to get to later and there will be a link    or here is a link here.  A/K/A [Indiscernible].  And you will be able to see all the projects.  Not just the ones that I highlighted today.  I encourage that you all take a look at what we have funded and will announce a few more at end of this month and see how your idea differs or similar to them and figure out what is unique angle.
Unless particularly in different region, we will not fund the same thing.  There are cultural differences and nuances being in different region.  This is global program.  Second question was on similar data.
So if we have different grantees collecting similar data, how do we facilitate that?  Great question.  Because when you go on our site, you will see that they are spread out across all of six disability segment that is we cover.  Neuro, diversity, speech, disability, deaf and hearing and vision and, yeah.  I think I had them all.
And so we do as part of open data set part that I was talking about in terms of requirements, since we want everybody to be open.  We do facilitate some of that data collecting if they are willing to share it back with Microsoft.  Some folks are willing to share it back to different data set.  That's fine.  If it's not with Microsoft.  As long as it's open so that someone else can build on top of it is what we care about.
Okay.  One more question and then I need to move on.
>> [off microphone].
>> I think that's Google.
>> What?
>> That's all right.  They are competing.  They have a program too.  We will be talking about ours on a website.  I'm going to keep going.  Okay.  Yeah, these are the things that we concerted in terms of how we want to have the world look like in five years.  We are making sure that we have people building with and by people with disabilities.  Having that global influence that we touched on in the questions which was great and collaboration piece that we touched on.
As part of ethics that I talked about at beginning of being inclusive in terms of AI and AI technology that's being built, I won't go into examples.  I'm pretty sure that you are red people and there is bias.  Bias in us as humans building the technology.  When you think about people with disabilities and the fact that there is a lack of data in our community, we really, really need solution that is are going to increase data collection so that we can make sure that we can feed the next wave of AI technology.  When making a hiring bot, needs to have data with people from Down Syndrome.  We need folks that are willing to collect that data.
Lastly, really about how do you want the world to look in five years, I would really love perhaps after this in the last 30 minutes if you want to talk about ideas that you have, I would love to provide feedback and you can also reach me and I will make sure that you have my e mail address.  Heather.Dowdy@Microsoft.com.  I encourage you to apply for a grant.
I'm going to skip over other parts that talked more about the challenges in each of our focus areas of employment, daily life, communication and connection.  I think we already hit on that.
I'm leaving you all with lots of resources.  I want you to really consider tinkering with Azure if you have not done so already.  Really easy to do if you go on to the site itself and see some of the tutorials that they have.  You can go and play with it.  If you apply, put in [Indiscernible] so I know it is connected in this event.  Look at the website.  Look at other things.  There are so much more that I could have talked about in terms of project.  It's interesting.  What we've got going on.
The next application deadline will be in November.  You got time to think of something.  And that's pretty much it.  Also, spread the word.  Tell other people about this program.  Maybe they are entrepreneurs or there is a start up that is looking for funding.  Spread the word that way we get lots and lots of great ideas circulating.  Thank you.
[Applause]
>> Thank you, Heather.  And also thank our sponsors for hosting us tonight.
[Applause]
Now important to networking portions you are going to return to begins now.  You do have an half an hour and we ask that [Indiscernible] plenty of food.  If you are still hungry, get some food.  We appreciate not having to store that.  Thank you, everybody.  Thank you for coming.
>> Thank you.