<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by Michael Bruner on Medium]]></title>
        <description><![CDATA[Stories by Michael Bruner on Medium]]></description>
        <link>https://medium.com/@michael.bruner3?source=rss-506382294dc2------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/0*CnGtVKkANH1uopT6.jpg</url>
            <title>Stories by Michael Bruner on Medium</title>
            <link>https://medium.com/@michael.bruner3?source=rss-506382294dc2------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Wed, 18 Nov 2026 07:34:53 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@michael.bruner3/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[Evolution Documentation]]></title>
            <link>https://medium.com/@michael.bruner3/evolution-documentation-ef517c308957?source=rss-506382294dc2------2</link>
            <guid isPermaLink="false">https://medium.com/p/ef517c308957</guid>
            <category><![CDATA[virtual-reality]]></category>
            <category><![CDATA[performance]]></category>
            <dc:creator><![CDATA[Michael Bruner]]></dc:creator>
            <pubDate>Tue, 14 May 2019 22:39:26 GMT</pubDate>
            <atom:updated>2019-05-14T22:39:26.700Z</atom:updated>
            <content:encoded><![CDATA[<p>Evolution is a multimedia dance performance at the University of Texas that is put together through a collaboration between multiple courses. I’m in a visual programming course where we are learning Touchdesigner and will be designing interactive visuals for one of the dance performances. I’ll be using this Medium post to document my process as work continues.</p><h3>General Plan</h3><p>Our team is made up of two visual programmers, a lighting designer, and a choreographer. We want to work as closely as possible with each other so we create a cohesive and compelling piece. From our initial meetings we’ve agreed to have our piece focus on competition and idolizing of an object and how this can become a source of conflict between people. We decided to have a prop on stage that symbolizes these idolized objects in our lives. To give the prop significance and impact on stage we’ve decided to track the object and embed LED lighting within it.</p><p>From these meetings it is clear that the three main challenges for myself will be tracking the prop, controlling the prop lighting wirelessly, and creating flexible interactive visuals that respond to the position of the prop on stage.</p><h3>Visual Reference</h3><p>These are some images that inspire me are helpful in informing the kinds of visual effects I want to create. Light painting, flocking, and lines are all visual elements I want to include in our performance. These can interact with dancers as they move around with the prop and create interesting and dynamic patterns.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/424/0*5frGthDQikCGH8jY" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1000/0*jcb3H52ixYytHQ-G" /></figure><p>This is a reference image for the kind of shape we want our prop to be.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/220/0*wTnKTscm4tlVjqRh.png" /></figure><h3>Wireless LED Prop</h3><p>For the prop the biggest challenge was figuring out a wireless lighting solution that was responsive and controllable through touchdesigner. I decided on using a wifi enabled micro-controller and using UDP as the communication protocol. I did all of the technical research, programming, and wiring for the prop. I wrote these two articles below that cover the specifics of the programming and wiring processes.</p><ul><li><a href="https://medium.com/@michael.bruner3/pixel-streaming-tutorial-3a702fc3da45">Pixel Streaming Tutorial</a></li><li><a href="https://medium.com/@michael.bruner3/led-prop-construction-958414884547">LED Prop construction</a></li></ul><p>The spatial tracking of the prop will be done using the HTC Vive tracking system. A single tracker will be attached to the top of our prop and allow us to get fast and precise positional data. We want to use this data to control light effects on the prop and to effect the projected visuals.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*o6IYBwufR3ezuU2k5zBrMQ.jpeg" /><figcaption>An image showing the prop with the Vive tracker and projections</figcaption></figure><h3>Touchdesigner Network</h3><p>When creating the touchdesigner network it was critical from the beginning to keep everything organized because we had a lot of data coming from different places. I started by thinking through the needs of the project and creating a rough layout of how everything should be organized and connected. This gave me a great starting point for the project and allowed me to create a structure from the very beginning that allowed for flexibility.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/842/1*o14SUvZSl8ZiDD2P9SsiVg.png" /><figcaption>Our Initial Touchdesigner Network Plan</figcaption></figure><p>A Key feature and visual effect I wanted to include in this performance was a flocking system that would follow the dancers throughout the choreography. I knew this was an effect people had worked on before so I did some digging and found some already built code and shaders that helped me achieve the effect I was aiming for. Below is an article where I go through the different resources I found and how they were useful to me.</p><p><a href="https://medium.com/@michael.bruner3/touchdesigner-flocking-boids-c34dca49afc2">Touchdesigner Flocking Boids</a></p><p>My final network ended up being similar to my initial network plan but like every project new features and ways of organizing came up during the process. Here I walk-through my final touchdesigner network and explain how it is organized and the different visual effects we ended up using.</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2FB-Btwea4iQA%3Ffeature%3Doembed&amp;url=http%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DB-Btwea4iQA&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2FB-Btwea4iQA%2Fhqdefault.jpg&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/953147fb3cd41a51670e5e46ccdf8df1/href">https://medium.com/media/953147fb3cd41a51670e5e46ccdf8df1/href</a></iframe><h3>Rehearsal</h3><p>Footage from our Dress Rehearsal.</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2FZOHxVI7sdWE%3Ffeature%3Doembed&amp;url=http%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DZOHxVI7sdWE&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2FZOHxVI7sdWE%2Fhqdefault.jpg&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/fbe8b746bf57ab283e6e5cbfebf0fec0/href">https://medium.com/media/fbe8b746bf57ab283e6e5cbfebf0fec0/href</a></iframe><h3>Performance</h3><p>Footage from our first performance.</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2FRm3awCcTfSM%3Ffeature%3Doembed&amp;url=http%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DRm3awCcTfSM&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2FRm3awCcTfSM%2Fhqdefault.jpg&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/a564fef701a1adfd4103c1dd8b566841/href">https://medium.com/media/a564fef701a1adfd4103c1dd8b566841/href</a></iframe><p>Overall I was very happy with how the project turned out and the level of quality our team achieved. The Projections at the beginning could have been more present but I think we did well with creating a sense of building and changing action throughout the performance.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=ef517c308957" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[LED Prop construction]]></title>
            <link>https://medium.com/@michael.bruner3/led-prop-construction-958414884547?source=rss-506382294dc2------2</link>
            <guid isPermaLink="false">https://medium.com/p/958414884547</guid>
            <category><![CDATA[arduino]]></category>
            <category><![CDATA[led]]></category>
            <dc:creator><![CDATA[Michael Bruner]]></dc:creator>
            <pubDate>Tue, 14 May 2019 17:00:46 GMT</pubDate>
            <atom:updated>2019-05-14T17:00:46.286Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*hWTv4sSpuR6ByZuwMOnsCQ.png" /><figcaption>Our finalized Prop with the lights on!</figcaption></figure><p>This article documents the construction of a wireless LED prop that was used for a multi-media dance performance. I’ll be covering the LED installation process in this article and have another article covering the programming and control of the object from touchdesigner.</p><p><a href="https://medium.com/@michael.bruner3/pixel-streaming-tutorial-3a702fc3da45">Pixel Streaming Tutorial</a></p><h3>Part List</h3><p>This build requires only a few components which makes it really easy to put together. Here is what we used to drive the LED lighting.</p><p>Wifi Board: <a href="https://www.adafruit.com/product/3405">Adafruit Huzzah32</a></p><p><a href="https://www.adafruit.com/product/1781">3.7v Lithium Ion Battery</a></p><p><a href="https://www.amazon.com/ALITOVE-Individual-Addressable-Programmable-Non-Waterproof/dp/B01MG49QKD/ref=sr_1_6?crid=283DXNXT7YTX4&amp;keywords=addressable+led+strip&amp;qid=1557851909&amp;s=gateway&amp;sprefix=addressable%2Celectronics%2C159&amp;sr=8-6">Addressable LEDs</a></p><p><a href="https://www.amazon.com/ALITOVE-Female-Connector-WS2812B-SK6812-RGBW/dp/B071H5XCN5/ref=sr_1_6?keywords=3pin+jst+connectors&amp;qid=1557851972&amp;s=gateway&amp;sr=8-6">JST Connectors</a></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*2OxGtP_pAaWDmgP6APk4Lw.jpeg" /><figcaption>Here is the board and battery connected to the top plate</figcaption></figure><p>The Battery plugs directly into the board and gets automatically recharged when the micro-usb is plugged in.</p><p>We used a total of 80 LEDs in our prop. We split the LEDs into groups of 4 and attached jst connectors to each of these so that we could easily build the object and snap the connectors together as we went. Cutting and soldering all of these connections was the most time consuming part of the process.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*O2V5qw1bSfTDb790NjF1GA.jpeg" /><figcaption>All of the LEDs cut and soldered</figcaption></figure><h3>Putting it Together</h3><p>Our prop was built out of multiple modular pieces which made it easy to construct and wire simultaneously. The LEDs have a sticky backing which we used to attach to our 3d printed connection pieces which held together our entire object. As we put the object together we connected the JST connectors together in series to ensure that by the time the whole object was constructed our wiring was completed.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Dn3mm4g3xPKeorJYyaIv5A.jpeg" /><figcaption>All of the pieces put out to assembled</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*pyO_zMSf2FGGbK3ObHE51g.jpeg" /><figcaption>In the process of putting it all together</figcaption></figure><h3>Final Result</h3><p>Here are some videos showing the assembled props with different lighting patterns going through it.</p><p>Some of the JST connectors are visible through the frosted acrylic and cutting them down shorter could have minimized this from happening.</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2FFyxRAC2qRXQ%3Ffeature%3Doembed&amp;url=http%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DFyxRAC2qRXQ&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2FFyxRAC2qRXQ%2Fhqdefault.jpg&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/03adb262e9de1bc096a53b3babb1eee8/href">https://medium.com/media/03adb262e9de1bc096a53b3babb1eee8/href</a></iframe><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2FEXKBd744cRQ%3Ffeature%3Doembed&amp;url=http%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DEXKBd744cRQ&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2FEXKBd744cRQ%2Fhqdefault.jpg&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/a1fd605b5428d3691f8d9ecd97541fd8/href">https://medium.com/media/a1fd605b5428d3691f8d9ecd97541fd8/href</a></iframe><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=958414884547" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Touchdesigner Flocking Boids]]></title>
            <link>https://medium.com/@michael.bruner3/touchdesigner-flocking-boids-c34dca49afc2?source=rss-506382294dc2------2</link>
            <guid isPermaLink="false">https://medium.com/p/c34dca49afc2</guid>
            <category><![CDATA[flocking]]></category>
            <category><![CDATA[unity3d]]></category>
            <category><![CDATA[touchdesigner]]></category>
            <dc:creator><![CDATA[Michael Bruner]]></dc:creator>
            <pubDate>Tue, 14 May 2019 03:42:12 GMT</pubDate>
            <atom:updated>2019-05-14T03:42:12.911Z</atom:updated>
            <content:encoded><![CDATA[<h3>Touchdesigner Flocking Research</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*xKLOryOyB-joteRP.jpg" /></figure><p>This is a collection of demos and github resources that are related to boid flocking systems.</p><p>This is a web article that has an interactive simulation that runs in the web. It also breaks down the basics of logic of boids and how they function. All behavior is driven by local interactions but through this large scale patterns and motion is created.</p><p><a href="http://www.harmendeweerd.nl/boids/">Boids: Flocking made simple</a></p><p>A video showing a flocking system combined with the leap motion in touchdesigner. It lacks a reference to the flocking algorithm they are using to drive this.</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2Fe--L3pArf-c%3Ffeature%3Doembed&amp;url=http%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3De--L3pArf-c&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2Fe--L3pArf-c%2Fhqdefault.jpg&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/fc02ccf5603575d830513eabd8d5a3de/href">https://medium.com/media/fc02ccf5603575d830513eabd8d5a3de/href</a></iframe><p>This is a github project for Unity which showcases multiple implementations of different flocking systems, both CPU and GPU based. It is a great reference and tool to use and can be easily placed into you Unity projects. Pretty impressive that it is all happening inside of a shader.</p><p><a href="https://github.com/Shinao/Unity-GPU-Boids">Shinao/Unity-GPU-Boids</a></p><h3>Referenced Articles</h3><p>Flocking articles and references I came upon.</p><p><a href="https://www.derivative.ca/forum/viewtopic.php?f=4&amp;t=9241">Touchdesigner object avoidance</a></p><p><a href="http://www.derivative.ca/Forum/viewtopic.php?f=22&amp;t=9151">Touchdesigner GPU flocking</a></p><p><a href="https://www.derivative.ca/Forum/viewtopic.php?f=36&amp;t=3794&amp;p=20130#p20130">Boids with Sops</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=c34dca49afc2" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Pixel Streaming Tutorial]]></title>
            <link>https://medium.com/@michael.bruner3/pixel-streaming-tutorial-3a702fc3da45?source=rss-506382294dc2------2</link>
            <guid isPermaLink="false">https://medium.com/p/3a702fc3da45</guid>
            <category><![CDATA[arduino]]></category>
            <category><![CDATA[diy]]></category>
            <category><![CDATA[esp32]]></category>
            <category><![CDATA[touchdesigner]]></category>
            <category><![CDATA[led-lighting]]></category>
            <dc:creator><![CDATA[Michael Bruner]]></dc:creator>
            <pubDate>Wed, 20 Mar 2019 05:02:00 GMT</pubDate>
            <atom:updated>2019-03-20T05:02:00.938Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*vezmE6spmFP85ZOGZtqiKw.jpeg" /></figure><p>In this tutorial I’ll go over how to wirelessly control addressable LED’s with touchdesigner. This is great if you want to have dynamic and reactive battery powered lighting that you want to be able to control from a distance. To get things up and running we need a few pieces of hardware and then we need to get our software and code configured.</p><p>Here’s the <a href="https://github.com/mikebru/Wifi-LED-Touch">Github link</a> with the code</p><p>To get this working I did a lot of googling and looking at what other people of done. This <a href="https://github.com/EnviralDesign">github </a>was particularly helpful for both the arduino code and touchdesigner setup.</p><h3>Hardware</h3><p>For hardware the three necessary components are a wifi board, I used the <a href="https://www.adafruit.com/product/3405">adafruit Feather Huzzah32</a>, a<a href="https://www.adafruit.com/product/1781"> 3.7v battery </a>(can also use USB), and <a href="https://www.adafruit.com/product/2561">addressable LED’s </a>. For my setup this is all the hardware that was required.</p><p>I’ve read that the LED’s need a 5v data and power and the Huzzah only has 3.7v so you need to get a <a href="https://www.adafruit.com/product/1787">3v to 5v level shifter</a>. However, I ran my setup without this and it works fine. Something to check out if you try this and it’s not working.</p><h3>Arduino Code</h3><p>To get going we’ll need to get some libraries installed so we can run everything. The main things are getting the latest version of the <a href="https://www.arduinolibraries.info/libraries/fast-led">FastLED </a>library and getting the <a href="https://learn.adafruit.com/adafruit-huzzah32-esp32-feather/using-with-arduino-ide">Huzzah32 board</a> added to the Arduino Ide.</p><p>In the github repository you’ll see multiple Arduino sketches, that use different communication protocols. These were for my own experimentation to see what worked best. I recommend using the UDPLED sketch as it had the most reliable results.</p><p>At the top of the sketch there are some variables you want to configure, the number of Leds, the data pin, and the network name and password.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/397/1*dKL8kRq1yLCfLgMLD_d6jQ.png" /></figure><p>The bulk of the program takes place in the main loop. For each pixel we are sending three bytes of data (r,g,b) so our inputBuffer length is (NumLeds * 3). We read in packets via UDP and wait till the inputBuffer is filled and then use a for loop to assign the rgb values to each LED. We then call FastLED.show() to flash the changes to the Led strip.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/681/1*6oOTx2mm806t6N3hbhfp3g.png" /></figure><h3>Touchdesigner Setup</h3><p>Inside of touchdesigner our main task is to take video data (TOPs) and convert it into a single channel (CHOP) that we can send out over UDP to the correct IP address and Port.</p><p>For converting the data we first crop down the image to the number of pixels we have in our strip. We then do a topTo chop conversion to get the information as channels. Three selects are used to separate the rgb values into distinct channels, after which the channels are shuffled together into a single long channel. So we finally get a channel formatted correctly(r1,g1,b1,r2, …).</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*aiElTzOnCdS6rwlliurpeg.png" /></figure><p>To send our data to board over UDP we use an Execute Dat in conjunction with a UdpOut Dat. At the end of each frame our execute Dat references both our formatted pixel data (‘null2’) and our udpout node. We use the command sendBytes and give it an argument of all our channel data. On the udpout dat be sure and set the network address and port correctly for the wifi board you’ve setup. The board should tell you it’s IP through the serial monitor when it boots up.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*uzgfnYgD-osE0Virf3-DOw.png" /></figure><h3>Conclusion</h3><p>That covers most of the technical details on how to set up communication between touchdesigner and a wifi board using UDP. Hopefully helpful!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=3a702fc3da45" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Research Plan]]></title>
            <link>https://medium.com/advanced-design-for-ai/research-plan-34c359055bd3?source=rss-506382294dc2------2</link>
            <guid isPermaLink="false">https://medium.com/p/34c359055bd3</guid>
            <category><![CDATA[language-learning]]></category>
            <category><![CDATA[research]]></category>
            <category><![CDATA[ai]]></category>
            <dc:creator><![CDATA[Michael Bruner]]></dc:creator>
            <pubDate>Tue, 26 Feb 2019 20:29:06 GMT</pubDate>
            <atom:updated>2019-02-26T20:29:06.243Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/650/1*SZkv9klO_xdO32fs2Wyyaw.png" /></figure><p><em>Team: Michael Bruner, Sarah Lee, Eric Chi, Louis An, Emily Whiteside</em></p><h3>Research Objectives</h3><p><em>What we want to learn</em></p><p>In our research we have a few primary objectives. These objectives involve understanding our potential users, gaining insight into current learning tools, and grasping the potential of state of the art AI.</p><p>With our Users we want to understand their language learning journey so far. What have been their greatest pain points and how have they addressed them with current technology and conventional learning practices. We want to understand what aspects have been difficult for them and which have been fairly easy. Which parts of language learning would be most beneficial to address? Specific subjects, accent help, simulated conversations.</p><p>With secondary research and interviews we want to discover any tools and applications that have been helpful in learning a new language and what specifically made them helpful.</p><p>Understanding the current state of AI natural language processing through secondary research, will allow us to make informed design decisions on how AI can leveraged to solve problems for our users.</p><h3>Frames</h3><h4>User Frame</h4><p><em>Our target users are international students, exchange students, and study abroad students.</em></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/683/0*qEHTvCN1VxT6Wnrq" /></figure><h4>Business Frame</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/705/0*kOSExc-h1DnXrkm-" /></figure><h4>Problem Frame</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/807/0*dEiFfq6DnoZpIXwB" /></figure><h3>Prove Me Wrong</h3><p><em>What “prove me wrong data” will you be looking for</em></p><p>When conducting our research we want to be sure and look for data that will challenge our assumptions. For us this involves asking our users if they perceive any additional need for digital language learning tools. Are digital language learning tools even helpful? How often do people use the language learning technology they already have? Is AI able to address the problems our users are facing? Are they comfortable using AI to help them learn languages and listen to their conversations.</p><p>We want to be sure and ask these challenging questions so we can get a clear idea of what users are actually looking for.</p><h3>Secondary Research</h3><p><em>The secondary research resources we will be investigating include:</em></p><h4>AI Natural Language</h4><p><a href="http://ruder.io/tracking-progress-nlp/">Natural Language Progress</a>, <a href="https://medium.com/dair-ai/deep-learning-for-nlp-an-overview-of-recent-trends-d0d8f40a776d">Medium Article: Different model types</a></p><p><a href="https://blog.openai.com/better-language-models/#sample5">Open AI GPT-2 model</a></p><p><a href="https://blog.openai.com/language-unsupervised/">Open AI Unsupervised Language Understanding</a></p><p><a href="https://www.engadget.com/2018/11/01/ai-can-learn-language-the-way-children-do/">AI Learning</a></p><h4><strong>Teaching</strong></h4><p><a href="https://www.independent.co.uk/life-style/learn-language-scientists-second-education-best-age-a8330911.html">Natural Age Cut off For Learning A New Language</a></p><p><a href="https://www.goethe.de/en/spr/mag/dsk/21290629.html">Creativity</a></p><h4>Language Learning Products</h4><p><a href="https://lifehacker.com/five-best-language-learning-tools-1448103513">Learning Tools</a></p><h4>Popular Languages to Learn</h4><p><a href="https://www.babbel.com/en/magazine/the-10-most-spoken-languages-in-the-world/">Most Common Languages</a></p><h4>Individual Learning Styles</h4><p><a href="https://www.businessinsider.com/fun-ways-learn-new-language-rosetta-babbel-duolingo">Popular ways to learn for adults</a></p><p><a href="https://deepenglish.com/article/8-ways-internet-changed-learning-language/">How The Internet Had Changed Language Learning</a></p><p><a href="http://www.melta.org.my/journals/index.php/tet/article/view/290">Vocabulary learning strategies of adult <strong>ESL learners</strong></a></p><p><a href="https://academic.oup.com/eltj/article-abstract/69/3/308/465365">Developments in <strong>online language </strong>learning</a></p><p><a href="https://riunet.upv.es/handle/10251/80112">A <strong>review </strong>of mobile <strong>language learning </strong>applications: Trends, challenges, and opportunities</a></p><p><a href="https://onlinelibrary.wiley.com/doi/abs/10.1111/bjet.12580">Mobile collaborative <strong>language learning</strong>: State of the art</a></p><h3>Primary Research Method</h3><p>Our Primary Research method will be to conduct in-person interviews with international students at UT to learn about their experiences with learning languages.</p><p>We would also like to interview professors who teach language and education.</p><h3>Schedule</h3><p><em>Our Three week schedule</em></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/758/0*y68Dg2e5865e-JSh" /></figure><h3>Data Collection Method</h3><p>We plan to video record our individual interviews and summarize our findings afterwards to share them with the group. For our secondary research, we are going to add our useful findings to a shared Google Doc.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=34c359055bd3" width="1" height="1" alt=""><hr><p><a href="https://medium.com/advanced-design-for-ai/research-plan-34c359055bd3">Research Plan</a> was originally published in <a href="https://medium.com/advanced-design-for-ai">Advanced Design for Artificial Intelligence</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Bias in the System]]></title>
            <link>https://medium.com/advanced-design-for-ai/bias-in-the-system-eb03f52fc4f6?source=rss-506382294dc2------2</link>
            <guid isPermaLink="false">https://medium.com/p/eb03f52fc4f6</guid>
            <category><![CDATA[bias]]></category>
            <category><![CDATA[language-learning]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <dc:creator><![CDATA[Michael Bruner]]></dc:creator>
            <pubDate>Tue, 26 Feb 2019 15:37:01 GMT</pubDate>
            <atom:updated>2019-02-26T15:37:01.596Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Jq1BwkbNrEg3UPqlkOshgQ.jpeg" /><figcaption><a href="https://www.flickr.com/photos/engrave/4385397139/in/photostream/">“Grid distortion</a>” by Marius Watz*</figcaption></figure><p>These last few weeks we’ve been in the process of designing an AI application that will help people learn new languages faster and more fluently. Our current concept focuses around an app where real people of different languages talk with one another and an AI facilitates the conversation and ensures that both sides are understanding one another. Users would be matched up based on what language they are wanting to learn and by conversing together both sides would gain practical conversation skills. The conversations happen asynchronously so that the AI assistant can work with each side, helping them both understand the spoken word as well as craft a coherent reply. Conversation partners would periodically switch what language is being spoken, allowing for each person to learn and teach.</p><p>For this application to be widely useful some bias and accessibility areas need to be addressed. The main feature of this application is a system that can understand spoken languages and convey that meaning to a user of another language. Our main concern here is that the model will only be well trained on major language groups, and so the application won’t be useful for people from minority language groups. Another important aspect is being able to understand differing accents and give users feedback on correction pronunciation. This requires, training the model on many different accent variations, and again some groups may not be well represented so the application will struggle with these cases. Each user, depending on their background, learning style, and culture will progress differently and the system needs to be able to teach all of them effectively. If the system is not trained on a large set of people from varying ages, languages, cultures, and socio-economic backgrounds it won’t be able to effectively teach or engage the user.</p><p>Most of these challenges can be resolved by gathering the right data and being intentional throughout the design process. If, from the beginning we set out to gather a diverse set of languages and include as many accent variations as possible our model with have a strong base. On a personal level this means that when we conduct research we need to go out and speak with the people that we have trouble understanding. These are the individuals that the system is most likely to have trouble understanding and are therefore the people who most benefit from having a tool that helps them be understood.</p><p>To improve our application continually as it is used data from our users conversations could be used to further train the model. If the system is ever unsure of what a user has said it can ask for a written translation of the words, so that it can add to it’s understanding. By gathering a diverse set of languages and accents and a system that understands how they are used in context we can improve natural language processing algorithms as a whole.</p><p>It is impossible to create a system which works perfectly for all users, there will always be bias in the system. Our goal is to reduce this as much as possible so that many people can become fluent in new languages and connect with one another. By helping people across the world understand one another we can minimize the bias in our selves and improve the world!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=eb03f52fc4f6" width="1" height="1" alt=""><hr><p><a href="https://medium.com/advanced-design-for-ai/bias-in-the-system-eb03f52fc4f6">Bias in the System</a> was originally published in <a href="https://medium.com/advanced-design-for-ai">Advanced Design for Artificial Intelligence</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[AI in live Performance]]></title>
            <link>https://medium.com/advanced-design-for-ai/ai-in-live-performance-f2fad2ac747b?source=rss-506382294dc2------2</link>
            <guid isPermaLink="false">https://medium.com/p/f2fad2ac747b</guid>
            <category><![CDATA[generative-art]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[performance]]></category>
            <category><![CDATA[ai]]></category>
            <dc:creator><![CDATA[Michael Bruner]]></dc:creator>
            <pubDate>Tue, 26 Feb 2019 15:00:45 GMT</pubDate>
            <atom:updated>2019-02-26T15:00:45.634Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*sCeOOQhS3_ei7dLJ.jpg" /><figcaption>New Nature by <a href="https://marpi.pl/work/new-nature/">Marpi Studios</a></figcaption></figure><p>How can AI be used in live performance both as a tool for generating content and as an active system that interacts with performers and viewers? My interest is in integrating technology into performances and spaces to create meaningful and magical experiences. So many experiences are possible now through computer graphics and interactive systems that were science fiction only a few decades ago. My belief is that joining AI and live performance will allow for a new dimension of interaction and generative systems that continually evolve.</p><p>The combination of these two fields is still very young and what is possible is still largely unexplored. Some artists have started to explore the space and have created some compelling work. Marpi Studios had an exhibition called <em>New Nature</em> where digital creatures interacted with attendees and throughout the exhibition the creatures continually evolved and morphed. <em>Sight Machine</em>, a piece by Trevor Paglen in collaboration with Obscura Digital displayed the output of computer vision algorithms during a performance by the Kronos Quartet. The intention of this piece was to bring to the foreground all of the algorithms that are apart of our daily life but we fail to notice.</p><ul><li><a href="https://marpi.pl/work/new-nature/">New Nature</a></li><li><a href="https://www.wired.com/2017/04/unsettling-performance-showed-world-ais-eyes/">The Unsettling Performance That Showed the World Through AI&#39;s Eyes</a></li></ul><p>There are two major reasons that I want to integrate AI into live performances. First, these tools allow for a whole new paradigm of generative art and intelligent interactions. The possibility of creating installation art where each person has a unique meaningful experience because an AI is able to adapt to the individual is very exciting. No longer is the piece static it is dynamic and adaptive based on context. Also, from a content creation point of view, creating images and 3D models with a certain look has the potential to be vastly sped up using AI.</p><p>The other major reason AI needs to be explored in the context of live performance is to create conversation and debate about the use of AI in our lives. Tech corporations often release intelligent features into our lives without us being aware and most of us have no idea how they work or what they are doing. A performance can bring these to the forefront and allow people to experience more directly how AI is used and how they feel about it.</p><p><a href="https://unity3d.com/machine-learning">Machine Learning - Unity</a></p><p>My skills and background in computer graphics and interactive systems prepare me well for bringing AI into these settings. Creating custom ML agents is an area I’m very interested in exploring and new tools continue to make it easier to get going. The possibilities in this field are wide open and that is exhilarating and just a bit daunting.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=f2fad2ac747b" width="1" height="1" alt=""><hr><p><a href="https://medium.com/advanced-design-for-ai/ai-in-live-performance-f2fad2ac747b">AI in live Performance</a> was originally published in <a href="https://medium.com/advanced-design-for-ai">Advanced Design for Artificial Intelligence</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[A Ideas]]></title>
            <link>https://medium.com/advanced-design-for-ai/a-ideas-5e2ada24bb90?source=rss-506382294dc2------2</link>
            <guid isPermaLink="false">https://medium.com/p/5e2ada24bb90</guid>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[style-transfer]]></category>
            <category><![CDATA[nvidia]]></category>
            <dc:creator><![CDATA[Michael Bruner]]></dc:creator>
            <pubDate>Mon, 04 Feb 2019 00:56:33 GMT</pubDate>
            <atom:updated>2019-02-26T14:42:50.835Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*TasY0XAH6r26HO6lpaqdBA.png" /><figcaption>Style by <a href="http://michellechrisman.com/about">Michelle Chrisman</a> Photo by me</figcaption></figure><p>This week I’ve been experimenting with Style transfer, “which is the technique of recomposing images in the style of other images. These were mostly created using <a href="https://github.com/jcjohnson/neural-style">Justin Johnson’s code</a> based on the <a href="https://arxiv.org/abs/1508.06576">paper by Gatys, Ecker, and Bethge</a> demonstrating a method for restyling images using convolutional neural networks.”</p><p><a href="http://genekogan.com/works/style-transfer/">Experiments with style transfer</a></p><p>The results from these networks are particularly impressive for their ability to apply the texture and colors that make an artwork unique. Something I love about style transfer is how in creating an image you end up discovering and looking at art from throughout history. However, interesting questions arise concerning who the artist of the final piece is. Is it me, the AI, the style reference Artist, or the writer of the algorithm?</p><p>There are many potential applications and uses for this technique today and for the future. In the near future an AI could be trained to choose a style it thinks goes well with an image that it is sent as input. The AI could even be the one that choose the input image based on what it perceives your preference to be, at which point it would have full control over the final images.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*baQLerilFij0aOTfGHx-jA.png" /><figcaption>Style transfer with two original works by my friend Will Tallent</figcaption></figure><p>Some exciting applications that are possible now is using real-time style transfer to transform video and images. A public installation with this effect would allow people to experience an artist’s style interactively from a new perspective. Soon it could be possible to run these networks on mobile devices which would open up a new way of creating compelling stylized photography on the go.</p><p><a href="https://github.com/lengstrom/fast-style-transfer">lengstrom/fast-style-transfer</a></p><p>Nvidia has developed an impressive algorithm that can generate hyper-realistic faces as well as other images it has been trained on. The model takes in 3 reference images and generates an entirely new face from them. These networks have a lot of potential applications but also can be misused in many ways.</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2FkSLJriaOumA%3Ffeature%3Doembed&amp;display_name=YouTube&amp;url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DkSLJriaOumA&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2FkSLJriaOumA%2Fhqdefault.jpg&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/ac24ee2337f662ff8244cd0e33cf41b6/href">https://medium.com/media/ac24ee2337f662ff8244cd0e33cf41b6/href</a></iframe><p>A compelling use for this technology when it is possible to run the network in real-time would be an installation art piece where viewers see a mirror of their transformed face. There could be three stations where each person sees a unique face that is the composite of all the participants. The resulting face would be familiar and follow your facial expressions. The age, race, and gender could all be changed, allowing participants to reflect on their identity and how all people are actually very similar and connected.</p><p><a href="https://github.com/NVIDIA/vid2vid">NVIDIA/vid2vid</a></p><p>On a less altruistic note, this same technology can be used to create convincing but fake images and videos that can be used to manipulate perceptions. A few examples include, fraudulent ID’s, fake news and impersonation. As these technologies advance it becomes important for new ways of detecting truth from fabrication to also be developed.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=5e2ada24bb90" width="1" height="1" alt=""><hr><p><a href="https://medium.com/advanced-design-for-ai/a-ideas-5e2ada24bb90">A Ideas</a> was originally published in <a href="https://medium.com/advanced-design-for-ai">Advanced Design for Artificial Intelligence</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[AI 2119]]></title>
            <link>https://medium.com/advanced-design-for-ai/ai-2119-faadfb50f7d2?source=rss-506382294dc2------2</link>
            <guid isPermaLink="false">https://medium.com/p/faadfb50f7d2</guid>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[future]]></category>
            <category><![CDATA[quantum-computing]]></category>
            <category><![CDATA[speculation]]></category>
            <dc:creator><![CDATA[Michael Bruner]]></dc:creator>
            <pubDate>Mon, 28 Jan 2019 02:40:03 GMT</pubDate>
            <atom:updated>2019-02-26T15:34:20.763Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*BT5rs4M7oEI8qpLOLApXFA.jpeg" /><figcaption>Image by: <a href="https://www.flickr.com/photos/markjsebastian/">mark sebastian</a></figcaption></figure><p>Where will AI be 100 years from today? In what ways will it change the world and how do we humans fit into that future? There are of course a multitude of possible futures that run the gamut from Utopian paradise to Orwellian dystopia. All technology is a doubled edged sword and because we lack the expertise in wielding these intelligent tools we are likely to feel both sides of the blade.</p><p>To understand how much AI and technology can advance in the span of 100 years it is helpful to look back on the last 100 and see how far things have already progressed. In 1919, IBM (known as CTR at the time) was only 8 years old and was focused on creating Tabulating Machines. These helpful machines helped keep track of census data and train passengers.</p><p><a href="https://www.ibm.com/ibm/history/ibm100/us/en/icons/founded/">IBM100 - IBM Is Founded</a></p><p>At the start of this year, 2019, IBM has unveiled their first commercial quantum computer. These 20 qubit computers will be available for members of the IBM Q Network to access through the cloud and start developing with. Quantum computing has the potential to radically increase the speed of AI processing and learning, especially if quantum chip development follows the same trajectory as classical computers.</p><p><a href="https://medium.com/thenextweb/heres-why-100-qubit-quantum-computers-could-change-everything-89b6321b448f">Here’s why 100 qubit quantum computers could change everything</a></p><p>With all of this in mind lets consider where the world could be in 2119. One of the most exciting possibilities is sentient artificial intelligence, which would mark an entirely new paradigm in evolution. These digital beings could possess an awareness and breadth of knowledge that we can hardly imagine. The entire world could become a coherent mind, with data centers, sensors, and satellites becoming neurons in a vast interconnected network. Globally conscious AI’s would be far more capable of acting collectively to accomplish challenging tasks and understanding our impact on the environment.</p><p>However, it is entirely possible that AI will not develop consciousness and the most major change will be augmenting our own intelligence with increasingly clever algorithms. Brain-Computer interfaces are the most extreme example of this, where AI would decode and encode digital information into understandable brain impulses. This would allow for the mind to expand in many ways. For example, direct mind communication, first person experience sharing, neuro-prosthesis control, and instant knowledge of any domain could all become possible.</p><p><a href="https://medium.com/predict/the-future-of-brain-computer-interfaces-and-the-human-machine-fb3a939e784f">The Future of Brain-Computer Interfaces and the Human Machine</a></p><p>With all of these dazzling possibilities there is also many manipulative futures where our ability to think freely is taken from us. AI’s controlled by governments and corporations could have an unprecedented amount of information about individuals. With brain interfaces it becomes entirely possible for the mind to be manipulated into believing whatever narrative those in power wish to establish.</p><p>At this point in history it is impossible to say how the future will unfold. Only by staying informed about current and future developments can we work towards shaping a positive future.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=faadfb50f7d2" width="1" height="1" alt=""><hr><p><a href="https://medium.com/advanced-design-for-ai/ai-2119-faadfb50f7d2">AI 2119</a> was originally published in <a href="https://medium.com/advanced-design-for-ai">Advanced Design for Artificial Intelligence</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
    </channel>
</rss>