This week, we have a look at the job market ten years from now – after the intelligent agents & robots happen – through Pew Research’s report on AI, Robotics and the Future of Jobs, we learn about IBM’s SyNpase chip and their investments in ‘non-silicon’ technology, and consider the economical costs of data protectionism. As usual, we also have some research highlights: low-power activity recognition and (my fav) spam filtering inspired by the human immune system.
This Week in Context
Your Weekly Update on All Things Context, August 8 2014
Got any friends and colleagues you think might be interested in this weekly digest on everything hot and brewing for context-aware and intelligent technologies? Point them here to subscribe to the Week in Context! Muchos gracias, herzlichen Dank!
“It is not the large things that will make AI acceptable it will be the small things—portable devices that can aid a person or organization in accomplishing desired outcomes well. AI embedded into everyday technology that proves to save time, energy, and stress that will push consumer demand for it.”
– Lillie Coney, PewResearch Report
‘AI, Robotics, and the Future of Jobs‘
A new report from Pew Research and Elon University tries to tackle the role of algorithms, intelligent digital agents, robots, and self-driving cars in the job market of 2025. They’ve asked almost two thousand experts if “networked, automated, artificial intelligence (AI) applications and robotic devices [will] have displaced more jobs than they have created by 2025?”
While the tally does not tell us much – 52% expect that technology will not displace more jobs than it creates, and 48% envisions it will – the report is a must-read if you want to know how the likes of Vint Cerf, John Markoff and Stowe Boyd think about the future of employment for robots and humans alike.
IBM chip processes data similar to the way your brain does
SyNapse, unveiled by IBM, is a chip that processes information using a million digital neurons, which communicate through electrical spikes, just as actual neurons do. The chip’s transistors are configured to mimic the behavior of our brain’s neurons and synapses, i.e. the connections between neurons. The chip is more battery efficient than conventional chips, and does not require the shuttling of information back and forth between memory and processing blocks. However, it will require an entirely new approach to programming. (via Roel)
Big Blue also unveiled a $3 billion research and development round in ‘non-silicon’ technologies. This will include further research into quantum computers which don’t know binary code, neurosynaptic chips like SyNapse, carbon nanotubes, graphene tools and a variety of other technologies.
(And, as the human brain is an inspiration for computing, so is the human immune system for spam filtering.)
No-Power Wi-Fi Connectivity Could Fuel Internet of Things Reality
Wi-Fi backscatter is a new communication system that uses radio frequency signals as a power source and reuses existing Wi-Fi infrastructure to provide Internet connectivity to Internet of Things-linked devices. The new technology is the first that can connect battery-free devices to Wi-Fi infrastructure.
The Economic Price of Data Protectionism
The Financial Times reports on liberal think thank ECIPE’s computations regarding the economic price of data protectionism; the cost of data localisation requirements and related data privacy and security laws for Brazil, the EU, China, India, Indonesia, Korea and Vietnam. Based on current and upcoming legislations, the required data localisation could mean a 0,1 to 3,1 GDP contraction. For Europe, this would be 0,4 to 1,1.
Prototype app would let ALS patients mind control their devices
Patients with ALS (amyotrophic lateral sclerosis) have gradually diminished muscle control, and can eventually become completely paralyzed without losing any other brain function. Philips and Accenture have developed a proof of concept app that would allow such a patient, equipped with an Emotiv sensor, to control Philips devices like the Philips Lifeline Emergency Alert system using only their minds. (Hat tip to Marnix!)
Smartphones as a Health Tool for Older Adults
Researchers are working on an app that will promote an active lifestyle and encourage habits that improve our quality of life, particularly in older adults. The team is currently working on validating the algorithms that measure heart rate variability (using the smartphone’s camera and flash) and physical activity (using the accelerometer, GPS, gyroscope and a few other smartphone sensors).
Also on mHealth, MedCity News reports Surprise: Pilot project finds mobile engagement with Medicaid patients ‘most cost-effective’. (NSS?!) People participating in the pilot programme who where responsive to mobile messaging, were also more likely to “take desired behaviour-changing action.”
ChoiceStream CEO on Pivot Decision And Embrace Of Programmatic Ads
In 2011, ChoiceStream switched from product recommendation software to programmatic advertising services. They use a combination of data mining & machine learning, with the more traditional survey-based methodology. AdExchange interviews CEO Eric Bosco. (via Vincent J)
Robots helped inspire deep learning and might become its killer app
This great article by Derrick Harris covers Andrew Ng’s talk at the Robotics: Science and Systems conference, the deep learning center of gravity shift to unsupervised learning, and the lack of images labeled ‘coffee mug’. 😉
“For a lot of applications,” Ng explained, “we’re starting to run out of labeled data.” As researchers try scaling training datasets from 50,000 to millions in order to improve accuracy, Ng noted, “there really aren’t that many coffee mugs in the world.” If there are that many images, most of them won’t be labeled. Computers will need to learn the concept of coffee mugs by themselves and then be told what they’ve discovered, because no one can afford to spend the amount it would take to label them.
Papers, Talks & Research
- AI, Robotics and the Future of Jobs (economics, research report)
- A spam filtering model based on immune mechanism (artificial immune mechanism, paper)
- Continuous context inference on mobile platforms (ubiquitous computing, activity recognition, PhD thesis)
- Developing a Body Sensor Network to Detect Emotions During Driving (automotive, emotion detection, abstract)
- Measures of Emotion in Interaction for Health Smart Home (human-computer interaction, emotion detection, paper)
- Smartphone Based Recognition of States and State Changes in Bipolar Disorder Patients (mood recognition, activity recognition, paper)
- Big Data Small Footprint: The Design of A Low-Power Classifier for Detecting Transportation Modes (activity recognition, mobile sensors, paper)
Enjoy the weekend, see you next week.