Anant Agarwal: Why massively open online courses (still) matter

Speaker

Anant Agarwal is president of edX, a partnership of MIT and Harvard to provide MOOCs. He is also a professor of MIT with background in electrical engineering and computer science.

Summary

Massively Open Online Courses (MOOCs) are taking off, able to teach millions of people at a time using online interactive technologies. 155,000 people enrolled in edX’s first MOOC, with 7,157 students passing. But in this course, Anant looks at what we can learn from MOOCs to apply in more conventional university courses.

University lectures have not changed significantly in the last 50 years. Anant claims the last big leap in education was textbooks and the printing press, hundreds of years ago. Modern teenagers interact and learn very differently to how they learnt decades ago – they are more comfortable in the online world. Education needs to embrace technology to better engage students.

  • Online Lecture, blended physical approach – let them watch videos at their own pace, wherever they are most comfortable. They can pause, rewind, skip, mute – letting them control the information flow better. Then they can meet physically to interact, discuss their learning, and do practical exercises in a classroom with other students. By adopting a blended approach in this way, San Jose university’s circuits and electronics course went from a 41% fail rate to 9%.
  • Active Learning – instead of traditional lectures, they are replaced with online ‘lessons’. A lesson includes an 8 minute video and then some exercises – and this lets people interact with the material more.
  • Instant feedback – exercises can be marked instantly, rather than submitting work and getting feedback weeks later. This lets people instantly know whether they are correct.
  • Gamification – Use computer simulations to build online labs – for example building a circuit using an online lab.
  • Peer learning – using forums and discussions to let students talk about the subject matter. Originally Anant was going to answer all questions himself, but after a while he realised that students would respond to each other. They could talk and get to the answer without any input from himself – the students were learning by teaching each other.

 

Advertisements

Johnny Lee: Wii Remote hacks

Speaker

Johnny Chung Lee is a Human-Computer Interaction researcher currently working at Google. Lee is best known for his work on Kinect development.

Summary

Researchers sometimes use vast resources to achieve their goal. Sometimes the best solution uses cheap, off-the-shelf products to do somethine similar. Johnny shows off some hacks taking advantage of the Wii-mote – a cheap game controller (~$40) with its motion sensing capabilities and infra-red camera.

  • With the right software, and a cheap infra-red pen (can be built for $5), the Wii-mote’s IR camera can simulate a digital white board (worth $3,000). This software has been downloaded half a million times.
  • By putting the Wii-mote’s IR camera under a TV screen, and fitting a (cheap) IR sensor to the viewer’s head, the screen will track them. So as they move around the room, the 3D interface on screen will show them a different perspective. This could allow a new type of video games.

Johnny is also amazed at how these hacks have gained popularity – using Youtube. Within days of posting videos of these hacks, he was already seeing others imitating the work. He hopes in future Youtube will be used to its full potential by researchers to publish results.

Randall Munroe: Comics that ask “what if?”

Speaker

Randall Monroe – former engineer at NASA and most known for his webcomic at xkcd.com

Summary

Every week on his website xkcd.com, Randall writes a feature called ‘What If’ – where he answers questions using maths and displays them as an explanation interspersed with comics. For example, when asked what would happen when a baseball is thrown at the speed of light, he talks about the interactions with the air – superheating it to plasma and creating a mushroom cloud that would strike the batter. In his opinion, this would constitute a ‘hit by pitch’, and allow the batter to walk to first base (if it still existed).

Randall was asked how big Google’s data centres would be if all their data was stored on punch cards. Noone knows how much data Google holds, but Randall predicted it using estimates of the money Google has, or how many hard drives they use, or how big their data centres are, or how much electricity they used. He can use what he does know to improve the quality of guessing what he doesn’t. He predicted 10 exabytes of data, and perhaps another 5 exabytes of offline data stored in tape drives. A punch card holds about 80 characters, with about 2,000 cards in a box. 15 exabytes would cover the entire area of New England to a depth of 5km. A few weeks later he got a message from Google, consisting of punch cards. This is a puzzle, including a code, which gives them some equations, which Randall eventually cracks to get a message. The message was “No Comment”.

Randall enjoys maths, but not for it’s own sake. He likes using maths to take things he knows and then use them to discover what he could never know.

But sometimes maths cannot help. One of his user questions just contained the subject line “Urgent’ and the question “If people had wheels and could fly, how would we differentiate them from airplanes?”.

David Epstein: Are athletes really getting faster, better, stronger?

Speaker

David Epstein is a senior writer at Sports Illustrated and the author of The Sports Gene: Inside the Science of Extraordinary Athletic Performance.

Summary

The Olympic motto is “Faster, Higher, Stronger”, and over time our Olympic records are getting better. However, David is investigating why: the technological and training drivers.

100m sprint: Jesse Owens won the 1936 Olympics in 10.2 seconds, but would have came last in the 2013 race where Usain Bolt set his record of 9.77seconds. The human race has not evolved over this time, but technology is supporting them: for example Owens ran on a soft surface of wood ash (cinders), which would have made him lose a lot more energy than Usain’s engineered carpet. Likewise, Usain ran from starting blocks, while Owens used a trowel to dig small holes in the surface. By analysing Owens’ movements, it is believed if he competed directly against Usain Bolt on the same surface, he would have come within a stride of victory, instead of being 14ft slower.

4 minute mile: The first man to run a 4 minute mile was in 1954, also on cinders (soft wood ash). Since then 1,314 have run a 4 minute mile, but the cinders are 1.5% slower than synthetic tracks. If you apply this conversion, only 513 men have run 4 minute miles.

100m freestyle: the record is decreasing over time, but there were sudden drops in time caused by introduction of flip turns, gutters around the pool to absorb ripples, and the introduction of full body swim suits.

Longest distance cycled in an hour: increased from 30miles, 3774 feet in 1972 more than 35miles in 1996. the 1996 bike was aerodynamic and much better engineered. When the rules were changed to force everyone to use a similar bike to what was used in 1972, the new record stands at 30miles 4657ft – not much further than 1972. Essentially the whole gain was from technological gain.

Selection of athletes: In the early 20th century it was believed that the most normal, average body type was best suited for all sports. Since then sports scientists revealed different body shapes were stronger in different sports, resulting in each sport having a certain type of people competing. This coincided with more people wanting to join in on the sports, making a wider range of people available to choose from, and therefore more people able to fit into the perfect body for the sport. This has been called the ‘Big Bang of Bodytypes’. Specialisation has concentrated people into these sports for example 1 in 6 men taller than 7ft are in the NBA. In sports where large bodies are prized, the athletes got bigger. Likewise sports preferring smaller bodies got them.

Ultra-endurance: There was a time when it was believed that ultra-endurance was harmful to our health. However, as we analysed our bodies we found they were hairless and easily cooled, and muscle structures made us well adapted to long runs. Recently Kilian Jornet ran up and down the Matterhorn (8,000ft) in under 3hrs, but he is not a freak. More people will do the same in the future, now that we have changed our mindset about what is possible.

With new technology, body type adaptation, mindset, imagination, and understanding of what the human body is capable of, athletes have been getting faster, higher, stronger.

Jeremy Kasdin: The flower-shaped starshade that might help us detect Earth-like planets

Speaker

Jeremy Kasdin: Aerospace engineer at Princeton University.

Summary

In the next decade, Jeremy wants us to build a space telescope that can take pictures of Earth-like planets in distant galaxies. Astronomers now believe that every star in the galaxy has a planet, and up to a fifth might have an earth-like planet that can sustain life. The ‘pale blue dot’ picture of Earth is difficult to take from a long way away, because the nearby beaming star overwhelms the telescope making it impossible to see the planets. Jeremy’s colleagues are working on technology to block out the extreme light of the sun and instead focus on the planet.

One idea is similar to the concept of an eclipse – where a closer object blocks the star, reducing its interference to a ring or corona effect. This is similar to putting a hand over a spotlight so you can see more clearly. Translating it into space: we’d build a large portable screen, open it between a telescope and a star, then take a picture. However, the corona of a circular sunshade still obscures the planet. Instead the sun shade is designed with a flower like pattern, to control the diffraction of light and prevent it washing out the picture. This should allow clearer pictures to be taken of distant planets.

The star shade is as big as half a football field, and has to be flown 50,000 km away from the telescope and then held right in the shadow. His engineers have been designing the system for unfurling and moving the shade.

His hope is that once completed we can take pictures of the planets around nearby stars, then use the information to analyse them and investigate further. By building a giant flower-shaped star-shade and seeing other stars’ pale blue dot, our understanding of the world will change.

Pattie Maes (and Pranav Mistry): Unveiling game-changing wearable tech

Speaker

Pattie Maes is a professor in MIT’s Program in Media Arts and Sciences. She founded and directed the MIT Media Lab’s Fluid Interfaces group. Previously, she founded and ran the Software Agents group. She currently acts as the associate Department Head for the Media, Arts and Sciences Department. Prior to joining the Media Lab, Maes was a visiting professor and a research scientist at the MIT Artificial Intelligence Lab. She holds bachelor’s and PhD degrees in computer science from the Vrije Universiteit Brussel in Belgium.

Summary

Pattie talked about the limitations of current computer technology – that it is impractical to Google a person or product every time you see it. Wearable technology can do this for you. She showed off an array of items that cost less than $350, including a camera, wearable projector, mirror, phone, and colourful caps on fingers. Using this, you can walk up to any wall and begin using a computer – with the camera tracking any gestures they make by recognising their fingernails. Even without a wall, you can project onto your hand – eg as a simple dialling pad for a phone.

Pattie shows some examples of

  • shopping for paper towels, with reviews and product information projected to each one.
  • Buying a book, you can see a total star review on the front cover. When turning the page you can see user comments, when turning again you can see annotations from professionals.
  • When talking to someone, a ‘tag cloud’ can be projected onto their Tshirt. This can show traits that person is known for, or information from blogs associated with them.
  • by looking at your boarding pass you can see if the flight is delayed or gate changed
  • draw on your wrist to project the time

Pranav Mistry: The thrilling potential of SixthSense technology

Speaker

Pranav Mistry is a PhD student in the Fluid Interfaces Group at MIT’s Media Lab. Before his studies at MIT, he worked with Microsoft as a UX researcher; he’s a graduate of IIT. Mistry is passionate about integrating the digital informational experience with our real-world interactions.

Summary

Gestures are everything, and come naturally to us. Pranav asks why we can’t interact with computers in the same we we interact. He experimented with different input systems for computers

  • a hacked mouse was turned into a glove – allowing the computer read hand movements
  • sticky notes that could be written and read by the computer, then either sent on as sms or treated as an input to the computer.
  • A pen that can draw in 3 dimensions
  • A computer map built into a table

People are interested in information, not necessarily the computers or pixels that show them. His next step was to try to eliminate the computer. SixthSense is a helmet mounted computer projected to a wall, that tracks your fingers using a camera. You can make gestures at any wall to use the computer. One gesture immediately takes a photo, another allows sending it as an email. Some extra features, acting as an interface between physical and digital world are

  • can recognise an object such as a book, and project a review onto it (3star etc)
  • can project videos or images onto newspapers
  • looking at a boarding pass and seeing if the flight is delayed
  • playing pong on the ground with your feet
  • can project onto a piece of paper and use it as a touch screen (play racing games, draw with finger, browse web)
  • ‘copy and paste’ from physical world onto the paper screen

SixthSense has the potential to keep us more connected to the physical world, and keep us human rather than a machine in front of another machine.

At the end of the video, he announced the software will be made open source for others to experiment with. The hardware is relatively cheap at ~$300.