We have a look at visualizing algorithms, Cortana’s prediction for the outcome of Argentina-Belgium this Saturday (#booh), automating your life with Saga IFTTT recipes and two great lamps, and highlight some great readings on affective sensing, superintelligence and the ‘Facebook influencing mood’ discussion.
This Week in Context
Your Weekly Update on All Things Context, July 4 2014
Mike Bostock, Editor at the New York Time’s graphics department and creator of visualisation library D3.js, published an in-depth article on visualising algorithms. A must-read, if not only for its great visual explanation of sampling, shuffling, sorting, making mazes, and the different types of randomness. The value in visualizing algorithms? This can be used to explain the algorithms, but just as well for debugging or reframing them.
“The goal here is to study the behavior of an algorithm rather than a specific dataset. Yet there is still data, necessarily — the data is derived from the execution of the algorithm. And this means we can use the type of derived data to classify algorithm visualizations.”
– Mike Bostock, creator 3D.js
Machines finally managed to match monkeys a key image-recognition test thanks to deep learning techniques. Machines are great at spotting identical images, but until recently, once subtle variations between images were introduced, monkeys still had the upper hand/eye as it came to visual object recognition. More on Wired.
Alternatively, you could use machine learning and algorithms to automatically cut out the ‘boring’ parts of your family videos. The LiveLight method constantly evaluates action in the video, looking for visual novelty and ignoring repetitive or eventless sequences, to create a summary that enables a viewer to get the gist of what happened.
We’re excited Microsoft’s Cortana got it right when she predicted Belgium would probably beat the USA. (Double triumph: data and our team.) To look at the WK2014 outcomes, the Bing prediction engine use models based on previous win/loss/tie records, home field advantage, surface, weather conditions, … and the ‘wisdoms of crowds’. (Alas, Cortana now predicts Belgium won’t make it past Argentina. So let’s hope she’s not always right.)
Robots will become more useful once they learn to properly converse with and talk like us. Scientists are working on not only getting robots to better understand what we want from them, but with inverse semantics, also on getting them to more clearly communicate back on where they are stuck. Neat thing is that the ‘teaching’ of the robot algorithms is being crowd sourced.
Ray Kurzweil says he’s breathing intelligence into Google Search. Inventor and futurist Ray Kurzweil says work is under way at Google to apply his theory of intelligence to understanding online information.
Google has acquired ‘contextual expert-curated’ streaming service Songza. The music streaming service is known for its human-curated playlists , and the Concierge feature, which uses date, time and user-set activity to recommend the appropriate playlist. Google Play announced the acquisition on Google Plus.
CONNECTABLES & WEARABLES
Spire, a wearable that tracks breathing patterns – through reparatory movement of the body – to give feedback about activity, stress levels and state of mind. – $119, ships September 2014
With ‘Automate Your Life’, Saga announced they now have a proper channel on If This Then That (IFTTT). For now these are mostly recipes based on the places you go. Saga says it will add more triggers to IFTTT, including arriving near a place of a certain category, leaving a place of a specified category, and arriving at or leaving your home or workplace.
This driverless car looks more like any “ordinary” car than a cyber-physical system. Yet it has top-of-the-line radar, cameras, sensors and other technologies. All are built into the body of the vehicle. Have a look as it takes a spin around Washington D.C.
Self-driving cars are a few years away still, and until then, designing in-vehicle entertainment systems means designing for minimal – or at least the right balance – of distraction. Anticipating what information you’ll need, is of course a great way to minimise dangerous distractions.
MOOD & SENSING
Design and the (Ir)Rational Mind: The Rise of Affective Sensing. This is part of the larger series ‘Inside Sensing’: “At frog we define sensing as the ability to harness these real-time data streams to identify patterns, generate insights, and design better experiences for people.” A great read passed on by the @Winflotte team.
Quarts joins in on the Mood-tracking – and more importantly, influencing – discussion with The US Military Is Already Tracking Your Mood. (via Marco De Ruiter) For the discussion itself, Niemanlab hosts a splendid overview.
Two great lamps for when you want to set the mood. This IKEA Mood Lamp hack makes for a lamp that changes its colour based on how you feel. The setup requires a webcam, facial recognition & analysis software (Open CV), and some Arduino skills. Personally, I prefer this Internet-connected thunderstorm for the modern home (aptly named The Cloud). Perfect for stormy nights.
Papers, Talks & Research
- Deep Neural Networks Rival the Representation of Primate IT Cortex for Core Visual Object Recognition (paper)
- Mood Lamp – A natural interaction system based on facial expression recognition (paper)
- The new role of radio and its public in the age of social network sites (paper)
- Failures of an embodied AIXI (analysis)
- Superintelligence: Paths, Dangers, Strategies (book)
Do you have research or an interesting article on machine learning, AI, sensing and empathic devices, or a review copy of ‘Superintelligence: Paths, Dangers & Strategies’? Let us know by tweet, or just hit reply.
Enjoy your weekend!
If you’d like to get the highlights of what is moving and shaking in our industry delivered to your mailbox once a week, sign up here.