Category Archives: robots

Artificial intelligence and metaphors

This is a different approach to artificial intelligence. From a June 27, 2017 news item on ScienceDaily,

Ask Siri to find a math tutor to help you “grasp” calculus and she’s likely to respond that your request is beyond her abilities. That’s because metaphors like “grasp” are difficult for Apple’s voice-controlled personal assistant to, well, grasp.

But new UC Berkeley research suggests that Siri and other digital helpers could someday learn the algorithms that humans have used for centuries to create and understand metaphorical language.

Mapping 1,100 years of metaphoric English language, researchers at UC Berkeley and Lehigh University in Pennsylvania have detected patterns in how English speakers have added figurative word meanings to their vocabulary.

The results, published in the journal Cognitive Psychology, demonstrate how throughout history humans have used language that originally described palpable experiences such as “grasping an object” to describe more intangible concepts such as “grasping an idea.”

Unfortunately, this image is not the best quality,

Scientists have created historical maps showing the evolution of metaphoric language. (Image courtesy of Mahesh Srinivasan)

A June 27, 2017 University of California at Berkeley (or UC Berkeley) news release by Yasmin Anwar, which originated the news item,

“The use of concrete language to talk about abstract ideas may unlock mysteries about how we are able to communicate and conceptualize things we can never see or touch,” said study senior author Mahesh Srinivasan, an assistant professor of psychology at UC Berkeley. “Our results may also pave the way for future advances in artificial intelligence.”

The findings provide the first large-scale evidence that the creation of new metaphorical word meanings is systematic, researchers said. They can also inform efforts to design natural language processing systems like Siri to help them understand creativity in human language.

“Although such systems are capable of understanding many words, they are often tripped up by creative uses of words that go beyond their existing, pre-programmed vocabularies,” said study lead author Yang Xu, a postdoctoral researcher in linguistics and cognitive science at UC Berkeley.

“This work brings opportunities toward modeling metaphorical words at a broad scale, ultimately allowing the construction of artificial intelligence systems that are capable of creating and comprehending metaphorical language,” he added.

Srinivasan and Xu conducted the study with Lehigh University psychology professor Barbara Malt.

Using the Metaphor Map of English database, researchers examined more than 5,000 examples from the past millennium in which word meanings from one semantic domain, such as “water,” were extended to another semantic domain, such as “mind.”

Researchers called the original semantic domain the “source domain” and the domain that the metaphorical meaning was extended to, the “target domain.”

More than 1,400 online participants were recruited to rate semantic domains such as “water” or “mind” according to the degree to which they were related to the external world (light, plants), animate things (humans, animals), or intense emotions (excitement, fear).

These ratings were fed into computational models that the researchers had developed to predict which semantic domains had been the sources or targets of metaphorical extension.

In comparing their computational predictions against the actual historical record provided by the Metaphor Map of English, researchers found that their models correctly forecast about 75 percent of recorded metaphorical language mappings over the past millennium.

Furthermore, they found that the degree to which a domain is tied to experience in the external world, such as “grasping a rope,” was the primary predictor of how a word would take on a new metaphorical meaning such as “grasping an idea.”

For example, time and again, researchers found that words associated with textiles, digestive organs, wetness, solidity and plants were more likely to provide sources for metaphorical extension, while mental and emotional states, such as excitement, pride and fear were more likely to be the targets of metaphorical extension.

Scientists have created historical maps showing the evolution of metaphoric language. (Image courtesy of Mahesh Srinivasan)

Here’s a link to and a citation for the paper,

Evolution of word meanings through metaphorical mapping: Systematicity over the past millennium by Yang Xu, Barbara C. Malt, Mahesh Srinivasan. Cognitive Psychology Volume 96, August 2017, Pages 41–53 DOI: https://doi.org/10.1016/j.cogpsych.2017.05.005

The early web version of this paper is behind a paywall.

For anyone interested in the ‘Metaphor Map of English’ database mentioned in the news release, you find it here on the University of Glasgow website. By the way, it also seems to be known as ‘Mapping Metaphor with the Historical Thesaurus‘.

IBM to build brain-inspired AI supercomputing system equal to 64 million neurons for US Air Force

This is the second IBM computer announcement I’ve stumbled onto within the last 4 weeks or so,  which seems like a veritable deluge given the last time I wrote about IBM’s computing efforts was in an Oct. 8, 2015 posting about carbon nanotubes,. I believe that up until now that was my  most recent posting about IBM and computers.

Moving onto the news, here’s more from a June 23, 3017 news item on Nanotechnology Now,

IBM (NYSE: IBM) and the U.S. Air Force Research Laboratory (AFRL) today [June 23, 2017] announced they are collaborating on a first-of-a-kind brain-inspired supercomputing system powered by a 64-chip array of the IBM TrueNorth Neurosynaptic System. The scalable platform IBM is building for AFRL will feature an end-to-end software ecosystem designed to enable deep neural-network learning and information discovery. The system’s advanced pattern recognition and sensory processing power will be the equivalent of 64 million neurons and 16 billion synapses, while the processor component will consume the energy equivalent of a dim light bulb – a mere 10 watts to power.

A June 23, 2017 IBM news release, which originated the news item, describes the proposed collaboration, which is based on IBM’s TrueNorth brain-inspired chip architecture (see my Aug. 8, 2014 posting for more about TrueNorth),

IBM researchers believe the brain-inspired, neural network design of TrueNorth will be far more efficient for pattern recognition and integrated sensory processing than systems powered by conventional chips. AFRL is investigating applications of the system in embedded, mobile, autonomous settings where, today, size, weight and power (SWaP) are key limiting factors.

The IBM TrueNorth Neurosynaptic System can efficiently convert data (such as images, video, audio and text) from multiple, distributed sensors into symbols in real time. AFRL will combine this “right-brain” perception capability of the system with the “left-brain” symbol processing capabilities of conventional computer systems. The large scale of the system will enable both “data parallelism” where multiple data sources can be run in parallel against the same neural network and “model parallelism” where independent neural networks form an ensemble that can be run in parallel on the same data.

“AFRL was the earliest adopter of TrueNorth for converting data into decisions,” said Daniel S. Goddard, director, information directorate, U.S. Air Force Research Lab. “The new neurosynaptic system will be used to enable new computing capabilities important to AFRL’s mission to explore, prototype and demonstrate high-impact, game-changing technologies that enable the Air Force and the nation to maintain its superior technical advantage.”

“The evolution of the IBM TrueNorth Neurosynaptic System is a solid proof point in our quest to lead the industry in AI hardware innovation,” said Dharmendra S. Modha, IBM Fellow, chief scientist, brain-inspired computing, IBM Research – Almaden. “Over the last six years, IBM has expanded the number of neurons per system from 256 to more than 64 million – an 800 percent annual increase over six years.’’

The system fits in a 4U-high (7”) space in a standard server rack and eight such systems will enable the unprecedented scale of 512 million neurons per rack. A single processor in the system consists of 5.4 billion transistors organized into 4,096 neural cores creating an array of 1 million digital neurons that communicate with one another via 256 million electrical synapses.    For CIFAR-100 dataset, TrueNorth achieves near state-of-the-art accuracy, while running at >1,500 frames/s and using 200 mW (effectively >7,000 frames/s per Watt) – orders of magnitude lower speed and energy than a conventional computer running inference on the same neural network.

The IBM TrueNorth Neurosynaptic System was originally developed under the auspices of Defense Advanced Research Projects Agency’s (DARPA) Systems of Neuromorphic Adaptive Plastic Scalable Electronics (SyNAPSE) program in collaboration with Cornell University. In 2016, the TrueNorth Team received the inaugural Misha Mahowald Prize for Neuromorphic Engineering and TrueNorth was accepted into the Computer History Museum.  Research with TrueNorth is currently being performed by more than 40 universities, government labs, and industrial partners on five continents.

There is an IBM video accompanying this news release, which seems more promotional than informational,

The IBM scientist featured in the video has a Dec. 19, 2016 posting on an IBM research blog which provides context for this collaboration with AFRL,

2016 was a big year for brain-inspired computing. My team and I proved in our paper “Convolutional networks for fast, energy-efficient neuromorphic computing” that the value of this breakthrough is that it can perform neural network inference at unprecedented ultra-low energy consumption. Simply stated, our TrueNorth chip’s non-von Neumann architecture mimics the brain’s neural architecture — giving it unprecedented efficiency and scalability over today’s computers.

The brain-inspired TrueNorth processor [is] a 70mW reconfigurable silicon chip with 1 million neurons, 256 million synapses, and 4096 parallel and distributed neural cores. For systems, we present a scale-out system loosely coupling 16 single-chip boards and a scale-up system tightly integrating 16 chips in a 4´4 configuration by exploiting TrueNorth’s native tiling.

For the scale-up systems we summarize our approach to physical placement of neural network, to reduce intra- and inter-chip network traffic. The ecosystem is in use at over 30 universities and government / corporate labs. Our platform is a substrate for a spectrum of applications from mobile and embedded computing to cloud and supercomputers.
TrueNorth Ecosystem for Brain-Inspired Computing: Scalable Systems, Software, and Applications

TrueNorth, once loaded with a neural network model, can be used in real-time as a sensory streaming inference engine, performing rapid and accurate classifications while using minimal energy. TrueNorth’s 1 million neurons consume only 70 mW, which is like having a neurosynaptic supercomputer the size of a postage stamp that can run on a smartphone battery for a week.

Recently, in collaboration with Lawrence Livermore National Laboratory, U.S. Air Force Research Laboratory, and U.S. Army Research Laboratory, we published our fifth paper at IEEE’s prestigious Supercomputing 2016 conference that summarizes the results of the team’s 12.5-year journey (see the associated graphic) to unlock this value proposition. [keep scrolling for the graphic]

Applying the mind of a chip

Three of our partners, U.S. Army Research Lab, U.S. Air Force Research Lab and Lawrence Livermore National Lab, contributed sections to the Supercomputing paper each showcasing a different TrueNorth system, as summarized by my colleagues Jun Sawada, Brian Taba, Pallab Datta, and Ben Shaw:

U.S. Army Research Lab (ARL) prototyped a computational offloading scheme to illustrate how TrueNorth’s low power profile enables computation at the point of data collection. Using the single-chip NS1e board and an Android tablet, ARL researchers created a demonstration system that allows visitors to their lab to hand write arithmetic expressions on the tablet, with handwriting streamed to the NS1e for character recognition, and recognized characters sent back to the tablet for arithmetic calculation.

Of course, the point here is not to make a handwriting calculator, it is to show how TrueNorth’s low power and real time pattern recognition might be deployed at the point of data collection to reduce latency, complexity and transmission bandwidth, as well as back-end data storage requirements in distributed systems.

U.S. Air Force Research Lab (AFRL) contributed another prototype application utilizing a TrueNorth scale-out system to perform a data-parallel text extraction and recognition task. In this application, an image of a document is segmented into individual characters that are streamed to AFRL’s NS1e16 TrueNorth system for parallel character recognition. Classification results are then sent to an inference-based natural language model to reconstruct words and sentences. This system can process 16,000 characters per second! AFRL plans to implement the word and sentence inference algorithms on TrueNorth, as well.

Lawrence Livermore National Lab (LLNL) has a 16-chip NS16e scale-up system to explore the potential of post-von Neumann computation through larger neural models and more complex algorithms, enabled by the native tiling characteristics of the TrueNorth chip. For the Supercomputing paper, they contributed a single-chip application performing in-situ process monitoring in an additive manufacturing process. LLNL trained a TrueNorth network to recognize seven classes related to track weld quality in welds produced by a selective laser melting machine. Real-time weld quality determination allows for closed-loop process improvement and immediate rejection of defective parts. This is one of several applications LLNL is developing to showcase TrueNorth as a scalable platform for low-power, real-time inference.

[downloaded from https://www.ibm.com/blogs/research/2016/12/the-brains-architecture-efficiency-on-a-chip/] Courtesy: IBM

I gather this 2017 announcement is the latest milestone on the TrueNorth journey.

Robot artists—should they get copyright protection

Clearly a lawyer wrote this June 26, 2017 essay on theconversation.com (Note: A link has been removed),

When a group of museums and researchers in the Netherlands unveiled a portrait entitled The Next Rembrandt, it was something of a tease to the art world. It wasn’t a long lost painting but a new artwork generated by a computer that had analysed thousands of works by the 17th-century Dutch artist Rembrandt Harmenszoon van Rijn.

The computer used something called machine learning [emphasis mine] to analyse and reproduce technical and aesthetic elements in Rembrandt’s works, including lighting, colour, brush-strokes and geometric patterns. The result is a portrait produced based on the styles and motifs found in Rembrandt’s art but produced by algorithms.

But who owns creative works generated by artificial intelligence? This isn’t just an academic question. AI is already being used to generate works in music, journalism and gaming, and these works could in theory be deemed free of copyright because they are not created by a human author.

This would mean they could be freely used and reused by anyone and that would be bad news for the companies selling them. Imagine you invest millions in a system that generates music for video games, only to find that music isn’t protected by law and can be used without payment by anyone in the world.

Unlike with earlier computer-generated works of art, machine learning software generates truly creative works without human input or intervention. AI is not just a tool. While humans program the algorithms, the decision making – the creative spark – comes almost entirely from the machine.

It could have been someone involved in the technology but nobody with that background would write “… something called machine learning … .”  Andres Guadamuz, lecturer in Intellectual Property Law at the University of Sussex, goes on to say (Note: Links have been removed),

Unlike with earlier computer-generated works of art, machine learning software generates truly creative works without human input or intervention. AI is not just a tool. While humans program the algorithms, the decision making – the creative spark – comes almost entirely from the machine.

That doesn’t mean that copyright should be awarded to the computer, however. Machines don’t (yet) have the rights and status of people under the law. But that doesn’t necessarily mean there shouldn’t be any copyright either. Not all copyright is owned by individuals, after all.

Companies are recognised as legal people and are often awarded copyright for works they don’t directly create. This occurs, for example, when a film studio hires a team to make a movie, or a website commissions a journalist to write an article. So it’s possible copyright could be awarded to the person (company or human) that has effectively commissioned the AI to produce work for it.

 

Things are likely to become yet more complex as AI tools are more commonly used by artists and as the machines get better at reproducing creativity, making it harder to discern if an artwork is made by a human or a computer. Monumental advances in computing and the sheer amount of computational power becoming available may well make the distinction moot. At that point, we will have to decide what type of protection, if any, we should give to emergent works created by intelligent algorithms with little or no human intervention.

The most sensible move seems to follow those countries that grant copyright to the person who made the AI’s operation possible, with the UK’s model looking like the most efficient. This will ensure companies keep investing in the technology, safe in the knowledge they will reap the benefits. What happens when we start seriously debating whether computers should be given the status and rights of people is a whole other story.

The team that developed a ‘new’ Rembrandt produced a video about the process,

Mark Brown’s April 5, 2016 article abut this project (which was unveiled on April 5, 2017 in Amsterdam, Netherlands) for the Guardian newspaper provides more detail such as this,

It [Next Rembrandt project] is the result of an 18-month project which asks whether new technology and data can bring back to life one of the greatest, most innovative painters of all time.

Advertising executive [Bas] Korsten, whose brainchild the project was, admitted that there were many doubters. “The idea was greeted with a lot of disbelief and scepticism,” he said. “Also coming up with the idea is one thing, bringing it to life is another.”

The project has involved data scientists, developers, engineers and art historians from organisations including Microsoft, Delft University of Technology, the Mauritshuis in The Hague and the Rembrandt House Museum in Amsterdam.

The final 3D printed painting consists of more than 148 million pixels and is based on 168,263 Rembrandt painting fragments.

Some of the challenges have been in designing a software system that could understand Rembrandt based on his use of geometry, composition and painting materials. A facial recognition algorithm was then used to identify and classify the most typical geometric patterns used to paint human features.

It sounds like it was a fascinating project but I don’t believe ‘The Next Rembrandt’ is an example of AI creativity or an example of the ‘creative spark’ Guadamuz discusses. This seems more like the kind of work  that could be done by a talented forger or fraudster. As I understand it, even when a human creates this type of artwork (a newly discovered and unknown xxx masterpiece), the piece is not considered a creative work in its own right. Some pieces are outright fraudulent and others which are described as “in the manner of xxx.”

Taking a somewhat different approach to mine, Timothy Geigner at Techdirt has also commented on the question of copyright and AI in relation to Guadamuz’s essay in a July 7, 2017 posting,

Unlike with earlier computer-generated works of art, machine learning software generates truly creative works without human input or intervention. AI is not just a tool. While humans program the algorithms, the decision making – the creative spark – comes almost entirely from the machine.

Let’s get the easy part out of the way: the culminating sentence in the quote above is not true. The creative spark is not the artistic output. Rather, the creative spark has always been known as the need to create in the first place. This isn’t a trivial quibble, either, as it factors into the simple but important reasoning for why AI and machines should certainly not receive copyright rights on their output.

That reasoning is the purpose of copyright law itself. Far too many see copyright as a reward system for those that create art rather than what it actually was meant to be: a boon to an artist to compensate for that artist to create more art for the benefit of the public as a whole. Artificial intelligence, however far progressed, desires only what it is programmed to desire. In whatever hierarchy of needs an AI might have, profit via copyright would factor either laughably low or not at all into its future actions. Future actions of the artist, conversely, are the only item on the agenda for copyright’s purpose. If receiving a copyright wouldn’t spur AI to create more art beneficial to the public, then copyright ought not to be granted.

Geigner goes on (July 7, 2017 posting) to elucidate other issues with the ideas expressed in the general debates of AI and ‘rights’ and the EU’s solution.

Robots and a new perspective on disability

I’ve long wondered about how disabilities would be viewed in a future (h/t May 4, 2017 news item on phys.org) where technology could render them largely irrelevant. A May 4, 2017 essay by Thusha (Gnanthusharan) Rajendran of Heriot-Watt University on TheConversation.com provides a perspective on the possibilities (Note: Links have been removed),

When dealing with the otherness of disability, the Victorians in their shame built huge out-of-sight asylums, and their legacy of “them” and “us” continues to this day. Two hundred years later, technologies offer us an alternative view. The digital age is shattering barriers, and what used to the norm is now being challenged.

What if we could change the environment, rather than the person? What if a virtual assistant could help a visually impaired person with their online shopping? And what if a robot “buddy” could help a person with autism navigate the nuances of workplace politics? These are just some of the questions that are being asked and which need answers as the digital age challenges our perceptions of normality.

The treatment of people with developmental conditions has a chequered history. In towns and cities across Britain, you will still see large Victorian buildings that were once places to “look after” people with disabilities, that is, remove them from society. Things became worse still during the time of the Nazis with an idealisation of the perfect and rejection of Darwin’s idea of natural diversity.

Today we face similar challenges about differences versus abnormalities. Arguably, current diagnostic systems do not help, because they diagnose the person and not “the system”. So, a child has challenging behaviour, rather than being in distress; the person with autism has a communication disorder rather than simply not being understood.

Natural-born cyborgs

In contrast, the digital world is all about systems. The field of human-computer interaction is about how things work between humans and computers or robots. Philosopher Andy Clark argues that humans have always been natural-born cyborgs – that is, we have always used technology (in its broadest sense) to improve ourselves.

The most obvious example is language itself. In the digital age we can become truly digitally enhanced. How many of us Google something rather than remembering it? How do you feel when you have no access to wi-fi? How much do we favour texting, tweeting and Facebook over face-to-face conversations? How much do we love and need our smartphones?

In the new field of social robotics, my colleagues and I are developing a robot buddy to help adults with autism to understand, for example, if their boss is pleased or displeased with their work. For many adults with autism, it is not the work itself that stops from them from having successful careers, it is the social environment surrounding work. From the stress-inducing interview to workplace politics, the modern world of work is a social minefield. It is not easy, at times, for us neurotypticals, but for a person with autism it is a world full contradictions and implied meaning.

Rajendra goes on to highlight efforts with autistic individuals; he also includes this video of his December 14, 2016 TEDx Heriot-Watt University talk, which largely focuses on his work with robots and autism  (Note: This runs approximately 15 mins.),

The talk reminded me of a Feb. 6, 2017 posting (scroll down about 33% of the way) where I discussed a recent book about science communication and its failure to recognize the importance of pop culture in that endeavour. As an example, I used a then recent announcement from MIT (Massachusetts Institute of Technology) about their emotion detection wireless application and the almost simultaneous appearance of that application in a Feb. 2, 2017 episode of The Big Bang Theory (a popular US television comedy) featuring a character who could be seen as autistic making use of the emotion detection device.

In any event, the work described in the MIT news release is very similar to Rajendra’s albeit the communication is delivered to the public through entirely different channels: TEDx Talk and TheConversation.com (channels aimed at academics and those with academic interests) and a pop culture television comedy with broad appeal.

Artificial intelligence (AI) company (in Montréal, Canada) attracts $135M in funding from Microsoft, Intel, Nvidia and others

It seems there’s a push on to establish Canada as a centre for artificial intelligence research and, if the federal and provincial governments have their way, for commercialization of said research. As always, there seems to be a bit of competition between Toronto (Ontario) and Montréal (Québec) as to which will be the dominant hub for the Canadian effort if one is to take Braga’s word for the situation.

In any event, Toronto seemed to have a mild advantage over Montréal initially with the 2017 Canadian federal government  budget announcement that the Canadian Institute for Advanced Research (CIFAR), based in Toronto, would launch a Pan-Canadian Artificial Intelligence Strategy and with an announcement from the University of Toronto shortly after (from my March 31, 2017 posting),

On the heels of the March 22, 2017 federal budget announcement of $125M for a Pan-Canadian Artificial Intelligence Strategy, the University of Toronto (U of T) has announced the inception of the Vector Institute for Artificial Intelligence in a March 28, 2017 news release by Jennifer Robinson (Note: Links have been removed),

A team of globally renowned researchers at the University of Toronto is driving the planning of a new institute staking Toronto’s and Canada’s claim as the global leader in AI.

Geoffrey Hinton, a University Professor Emeritus in computer science at U of T and vice-president engineering fellow at Google, will serve as the chief scientific adviser of the newly created Vector Institute based in downtown Toronto.

“The University of Toronto has long been considered a global leader in artificial intelligence research,” said U of T President Meric Gertler. “It’s wonderful to see that expertise act as an anchor to bring together researchers, government and private sector actors through the Vector Institute, enabling them to aim even higher in leading advancements in this fast-growing, critical field.”

As part of the Government of Canada’s Pan-Canadian Artificial Intelligence Strategy, Vector will share $125 million in federal funding with fellow institutes in Montreal and Edmonton. All three will conduct research and secure talent to cement Canada’s position as a world leader in AI.

However, Montréal and the province of Québec are no slouches when it comes to supporting to technology. From a June 14, 2017 article by Matthew Braga for CBC (Canadian Broadcasting Corporation) news online (Note: Links have been removed),

One of the most promising new hubs for artificial intelligence research in Canada is going international, thanks to a $135 million investment with contributions from some of the biggest names in tech.

The company, Montreal-based Element AI, was founded last October [2016] to help companies that might not have much experience in artificial intelligence start using the technology to change the way they do business.

It’s equal parts general research lab and startup incubator, with employees working to develop new and improved techniques in artificial intelligence that might not be fully realized for years, while also commercializing products and services that can be sold to clients today.

It was co-founded by Yoshua Bengio — one of the pioneers of a type of AI research called machine learning — along with entrepreneurs Jean-François Gagné and Nicolas Chapados, and the Canadian venture capital fund Real Ventures.

In an interview, Bengio and Gagné said the money from the company’s funding round will be used to hire 250 new employees by next January. A hundred will be based in Montreal, but an additional 100 employees will be hired for a new office in Toronto, and the remaining 50 for an Element AI office in Asia — its first international outpost.

They will join more than 100 employees who work for Element AI today, having left jobs at Amazon, Uber and Google, among others, to work at the company’s headquarters in Montreal.

The expansion is a big vote of confidence in Element AI’s strategy from some of the world’s biggest technology companies. Microsoft, Intel and Nvidia all contributed to the round, and each is a key player in AI research and development.

The company has some not unexpected plans and partners (from the Braga, article, Note: A link has been removed),

The Series A round was led by Data Collective, a Silicon Valley-based venture capital firm, and included participation by Fidelity Investments Canada, National Bank of Canada, and Real Ventures.

What will it help the company do? Scale, its founders say.

“We’re looking at domain experts, artificial intelligence experts,” Gagné said. “We already have quite a few, but we’re looking at people that are at the top of their game in their domains.

“And at this point, it’s no longer just pure artificial intelligence, but people who understand, extremely well, robotics, industrial manufacturing, cybersecurity, and financial services in general, which are all the areas we’re going after.”

Gagné says that Element AI has already delivered 10 projects to clients in those areas, and have many more in development. In one case, Element AI has been helping a Japanese semiconductor company better analyze the data collected by the assembly robots on its factory floor, in a bid to reduce manufacturing errors and improve the quality of the company’s products.

There’s more to investment in Québec’s AI sector than Element AI (from the Braga article; Note: Links have been removed),

Element AI isn’t the only organization in Canada that investors are interested in.

In September, the Canadian government announced $213 million in funding for a handful of Montreal universities, while both Google and Microsoft announced expansions of their Montreal AI research groups in recent months alongside investments in local initiatives. The province of Quebec has pledged $100 million for AI initiatives by 2022.

Braga goes on to note some other initiatives but at that point the article’s focus is exclusively Toronto.

For more insight into the AI situation in Québec, there’s Dan Delmar’s May 23, 2017 article for the Montreal Express (Note: Links have been removed),

Advocating for massive government spending with little restraint admittedly deviates from the tenor of these columns, but the AI business is unlike any other before it. [emphasis misn] Having leaders acting as fervent advocates for the industry is crucial; resisting the coming technological tide is, as the Borg would say, futile.

The roughly 250 AI researchers who call Montreal home are not simply part of a niche industry. Quebec’s francophone character and Montreal’s multilingual citizenry are certainly factors favouring the development of language technology, but there’s ample opportunity for more ambitious endeavours with broader applications.

AI isn’t simply a technological breakthrough; it is the technological revolution. [emphasis mine] In the coming decades, modern computing will transform all industries, eliminating human inefficiencies and maximizing opportunities for innovation and growth — regardless of the ethical dilemmas that will inevitably arise.

“By 2020, we’ll have computers that are powerful enough to simulate the human brain,” said (in 2009) futurist Ray Kurzweil, author of The Singularity Is Near, a seminal 2006 book that has inspired a generation of AI technologists. Kurzweil’s projections are not science fiction but perhaps conservative, as some forms of AI already effectively replace many human cognitive functions. “By 2045, we’ll have expanded the intelligence of our human-machine civilization a billion-fold. That will be the singularity.”

The singularity concept, borrowed from physicists describing event horizons bordering matter-swallowing black holes in the cosmos, is the point of no return where human and machine intelligence will have completed their convergence. That’s when the machines “take over,” so to speak, and accelerate the development of civilization beyond traditional human understanding and capability.

The claims I’ve highlighted in Delmar’s article have been made before for other technologies, “xxx is like no other business before’ and “it is a technological revolution.”  Also if you keep scrolling down to the bottom of the article, you’ll find Delmar is a ‘public relations consultant’ which, if you look at his LinkedIn profile, you’ll find means he’s a managing partner in a PR firm known as Provocateur.

Bertrand Marotte’s May 20, 2017 article for the Montreal Gazette offers less hyperbole along with additional detail about the Montréal scene (Note: Links have been removed),

It might seem like an ambitious goal, but key players in Montreal’s rapidly growing artificial-intelligence sector are intent on transforming the city into a Silicon Valley of AI.

Certainly, the flurry of activity these days indicates that AI in the city is on a roll. Impressive amounts of cash have been flowing into academia, public-private partnerships, research labs and startups active in AI in the Montreal area.

…, researchers at Microsoft Corp. have successfully developed a computing system able to decipher conversational speech as accurately as humans do. The technology makes the same, or fewer, errors than professional transcribers and could be a huge boon to major users of transcription services like law firms and the courts.

Setting the goal of attaining the critical mass of a Silicon Valley is “a nice point of reference,” said tech entrepreneur Jean-François Gagné, co-founder and chief executive officer of Element AI, an artificial intelligence startup factory launched last year.

The idea is to create a “fluid, dynamic ecosystem” in Montreal where AI research, startup, investment and commercialization activities all mesh productively together, said Gagné, who founded Element with researcher Nicolas Chapados and Université de Montréal deep learning pioneer Yoshua Bengio.

“Artificial intelligence is seen now as a strategic asset to governments and to corporations. The fight for resources is global,” he said.

The rise of Montreal — and rival Toronto — as AI hubs owes a lot to provincial and federal government funding.

Ottawa promised $213 million last September to fund AI and big data research at four Montreal post-secondary institutions. Quebec has earmarked $100 million over the next five years for the development of an AI “super-cluster” in the Montreal region.

The provincial government also created a 12-member blue-chip committee to develop a strategic plan to make Quebec an AI hub, co-chaired by Claridge Investments Ltd. CEO Pierre Boivin and Université de Montréal rector Guy Breton.

But private-sector money has also been flowing in, particularly from some of the established tech giants competing in an intense AI race for innovative breakthroughs and the best brains in the business.

Montreal’s rich talent pool is a major reason Waterloo, Ont.-based language-recognition startup Maluuba decided to open a research lab in the city, said the company’s vice-president of product development, Mohamed Musbah.

“It’s been incredible so far. The work being done in this space is putting Montreal on a pedestal around the world,” he said.

Microsoft struck a deal this year to acquire Maluuba, which is working to crack one of the holy grails of deep learning: teaching machines to read like the human brain does. Among the company’s software developments are voice assistants for smartphones.

Maluuba has also partnered with an undisclosed auto manufacturer to develop speech recognition applications for vehicles. Voice recognition applied to cars can include such things as asking for a weather report or making remote requests for the vehicle to unlock itself.

Marotte’s Twitter profile describes him as a freelance writer, editor, and translator.

Meet Pepper, a robot for health care clinical settings

A Canadian project to introduce robots like Pepper into clinical settings (aside: can seniors’ facilities be far behind?) is the subject of a June 23, 2017 news item on phys.org,

McMaster and Ryerson universities today announced the Smart Robots for Health Communication project, a joint research initiative designed to introduce social robotics and artificial intelligence into clinical health care.

A June 22, 2017 McMaster University news release, which originated the news item, provides more detail,

With the help of Softbank’s humanoid robot Pepper and IBM Bluemix Watson Cognitive Services, the researchers will study health information exchange through a state-of-the-art human-robot interaction system. The project is a collaboration between David Harris Smith, professor in the Department of Communication Studies and Multimedia at McMaster University, Frauke Zeller, professor in the School of Professional Communication at Ryerson University and Hermenio Lima, a dermatologist and professor of medicine at McMaster’s Michael G. DeGroote School of Medicine. His main research interests are in the area of immunodermatology and technology applied to human health.

The research project involves the development and analysis of physical and virtual human-robot interactions, and has the capability to improve healthcare outcomes by helping healthcare professionals better understand patients’ behaviour.

Zeller and Harris Smith have previously worked together on hitchBOT, the friendly hitchhiking robot that travelled across Canada and has since found its new home in the [Canada] Science and Technology Museum in Ottawa.

“Pepper will help us highlight some very important aspects and motives of human behaviour and communication,” said Zeller.

Designed to be used in professional environments, Pepper is a humanoid robot that can interact with people, ‘read’ emotions, learn, move and adapt to its environment, and even recharge on its own. Pepper is able to perform facial recognition and develop individualized relationships when it interacts with people.

Lima, the clinic director, said: “We are excited to have the opportunity to potentially transform patient engagement in a clinical setting, and ultimately improve healthcare outcomes by adapting to clients’ communications needs.”

At Ryerson, Pepper was funded by the Co-lab in the Faculty of Communication and Design. FCAD’s Co-lab provides strategic leadership, technological support and acquisitions of technologies that are shaping the future of communications.

“This partnership is a testament to the collaborative nature of innovation,” said dean of FCAD, Charles Falzon. “I’m thrilled to support this multidisciplinary project that pushes the boundaries of research, and allows our faculty and students to find uses for emerging tech inside and outside the classroom.”

“This project exemplifies the value that research in the Humanities can bring to the wider world, in this case building understanding and enhancing communications in critical settings such as health care,” says McMaster’s Dean of Humanities, Ken Cruikshank.

The integration of IBM Watson cognitive computing services with the state-of-the-art social robot Pepper, offers a rich source of research potential for the projects at Ryerson and McMaster. This integration is also supported by IBM Canada and [Southern Ontario Smart Computing Innovation Platform] SOSCIP by providing the project access to high performance research computing resources and staff in Ontario.

“We see this as the initiation of an ongoing collaborative university and industry research program to develop and test applications of embodied AI, a research program that is well-positioned to integrate and apply emerging improvements in machine learning and social robotics innovations,” said Harris Smith.

I just went to a presentation at the facility where my mother lives and it was all about delivering more individualized and better care for residents. Given that most seniors in British Columbia care facilities do not receive the number of service hours per resident recommended by the province due to funding issues, it seemed a well-meaning initiative offered in the face of daunting odds against success. Now with this news, I wonder what impact ‘Pepper’ might ultimately have on seniors and on the people who currently deliver service. Of course, this assumes that researchers will be able to tackle problems with understanding various accents and communication strategies, which are strongly influenced by culture and, over time, the aging process.

After writing that last paragraph I stumbled onto this June 27, 2017 Sage Publications press release on EurekAlert about a related matter,

Existing digital technologies must be exploited to enable a paradigm shift in current healthcare delivery which focuses on tests, treatments and targets rather than the therapeutic benefits of empathy. Writing in the Journal of the Royal Society of Medicine, Dr Jeremy Howick and Dr Sian Rees of the Oxford Empathy Programme, say a new paradigm of empathy-based medicine is needed to improve patient outcomes, reduce practitioner burnout and save money.

Empathy-based medicine, they write, re-establishes relationship as the heart of healthcare. “Time pressure, conflicting priorities and bureaucracy can make practitioners less likely to express empathy. By re-establishing the clinical encounter as the heart of healthcare, and exploiting available technologies, this can change”, said Dr Howick, a Senior Researcher in Oxford University’s Nuffield Department of Primary Care Health Sciences.

Technology is already available that could reduce the burden of practitioner paperwork by gathering basic information prior to consultation, for example via email or a mobile device in the waiting room.

During the consultation, the computer screen could be placed so that both patient and clinician can see it, a help to both if needed, for example, to show infographics on risks and treatment options to aid decision-making and the joint development of a treatment plan.

Dr Howick said: “The spread of alternatives to face-to-face consultations is still in its infancy, as is our understanding of when a machine will do and when a person-to-person relationship is needed.” However, he warned, technology can also get in the way. A computer screen can become a barrier to communication rather than an aid to decision-making. “Patients and carers need to be involved in determining the need for, and designing, new technologies”, he said.

I sincerely hope that the Canadian project has taken into account some of the issues described in the ’empathy’ press release and in the article, which can be found here,

Overthrowing barriers to empathy in healthcare: empathy in the age of the Internet
by J Howick and S Rees. Journaly= of the Royal Society of Medicine Article first published online: June 27, 2017 DOI: https://doi.org/10.1177/0141076817714443

This article is open access.

Hacking the human brain with a junction-based artificial synaptic device

Earlier today I published a piece featuring Dr. Wei Lu’s work on memristors and the movement to create an artificial brain (my June 28, 2017 posting: Dr. Wei Lu and bio-inspired ‘memristor’ chips). For this posting I’m featuring a non-memristor (if I’ve properly understood the technology) type of artificial synapse. From a June 28, 2017 news item on Nanowerk,

One of the greatest challenges facing artificial intelligence development is understanding the human brain and figuring out how to mimic it.

Now, one group reports in ACS Nano (“Emulating Bilingual Synaptic Response Using a Junction-Based Artificial Synaptic Device”) that they have developed an artificial synapse capable of simulating a fundamental function of our nervous system — the release of inhibitory and stimulatory signals from the same “pre-synaptic” terminal.

Unfortunately, the American Chemical Society news release on EurekAlert, which originated the news item, doesn’t provide too much more detail,

The human nervous system is made up of over 100 trillion synapses, structures that allow neurons to pass electrical and chemical signals to one another. In mammals, these synapses can initiate and inhibit biological messages. Many synapses just relay one type of signal, whereas others can convey both types simultaneously or can switch between the two. To develop artificial intelligence systems that better mimic human learning, cognition and image recognition, researchers are imitating synapses in the lab with electronic components. Most current artificial synapses, however, are only capable of delivering one type of signal. So, Han Wang, Jing Guo and colleagues sought to create an artificial synapse that can reconfigurably send stimulatory and inhibitory signals.

The researchers developed a synaptic device that can reconfigure itself based on voltages applied at the input terminal of the device. A junction made of black phosphorus and tin selenide enables switching between the excitatory and inhibitory signals. This new device is flexible and versatile, which is highly desirable in artificial neural networks. In addition, the artificial synapses may simplify the design and functions of nervous system simulations.

Here’s how I concluded that this is not a memristor-type device (from the paper [first paragraph, final sentence]; a link and citation will follow; Note: Links have been removed)),

The conventional memristor-type [emphasis mine](14-20) and transistor-type(21-25) artificial synapses can realize synaptic functions in a single semiconductor device but lacks the ability [emphasis mine] to dynamically reconfigure between excitatory and inhibitory responses without the addition of a modulating terminal.

Here’s a link to and a citation for the paper,

Emulating Bilingual Synaptic Response Using a Junction-Based Artificial Synaptic Device by
He Tian, Xi Cao, Yujun Xie, Xiaodong Yan, Andrew Kostelec, Don DiMarzio, Cheng Chang, Li-Dong Zhao, Wei Wu, Jesse Tice, Judy J. Cha, Jing Guo, and Han Wang. ACS Nano, Article ASAP DOI: 10.1021/acsnano.7b03033 Publication Date (Web): June 28, 2017

Copyright © 2017 American Chemical Society

This paper is behind a paywall.

May/June 2017 scienceish events in Canada (mostly in Vancouver)

I have five* events for this posting

(1) Science and You (Montréal)

The latest iteration of the Science and You conference took place May 4 – 6, 2017 at McGill University (Montréal, Québec). That’s the sad news, the good news is that they have recorded and released the sessions onto YouTube. (This is the first time the conference has been held outside of Europe, in fact, it’s usually held in France.) Here’s why you might be interested (from the 2017 conference page),

The animator of the conference will be Véronique Morin:

Véronique Morin is science journalist and communicator, first president of the World Federation of Science Journalists (WFSJ) and serves as judge for science communication awards. She worked for a science program on Quebec’s public TV network, CBCRadio-Canada, TVOntario, and as a freelancer is also a contributor to -among others-  The Canadian Medical Journal, University Affairs magazine, NewsDeeply, while pursuing documentary projects.

Let’s talk about S …

Holding the attention of an audience full of teenagers may seem impossible… particularly on topics that might be seen as boring, like sciences! Yet, it’s essential to demistify science in order to make it accessible, even appealing in the eyes of futur citizens.
How can we encourage young adults to ask themselves questions about the surrounding world, nature and science? How can we make them discover sciences with and without digital tools?

Find out tips and tricks used by our speakers Kristin Alford and Amanda Tyndall.

Kristin Alford
Dr Kristin Alford is a futurist and the inaugural Director of MOD., a futuristic museum of discovery at the University of South Australia. Her mind is presently occupied by the future of work and provoking young adults to ask questions about the role of science at the intersection of art and innovation.

Internet Website

Amanda Tyndall
Over 20 years of  science communication experience with organisations such as Café Scientifique, The Royal Institution of Great Britain (and Australia’s Science Exchange), the Science Museum in London and now with the Edinburgh International Science Festival. Particularly interested in engaging new audiences through linkages with the arts and digital/creative industries.

Internet Website

A troll in the room

Increasingly used by politicians, social media can reach thousand of people in few seconds. Relayed to infinity, the message seems truthful, but is it really? At a time of fake news and alternative facts, how can we, as a communicator or a journalist, take up the challenge of disinformation?
Discover the traps and tricks of disinformation in the age of digital technologies with our two fact-checking experts, Shawn Otto and Vanessa Schipani, who will offer concrete solutions to unravel the true from the false..

 

Shawn Otto
Shawn Otto was awarded the IEEE-USA (“I-Triple-E”) National Distinguished Public Service Award for his work elevating science in America’s national public dialogue. He is cofounder and producer of the US presidential science debates at ScienceDebate.org. He is also an award-winning screenwriter and novelist, best known for writing and co-producing the Academy Award-nominated movie House of Sand and Fog.

Vanessa Schipani
Vanessa is a science journalist at FactCheck.org, which monitors U.S. politicians’ claims for accuracy. Previously, she wrote for outlets in the U.S., Europe and Japan, covering topics from quantum mechanics to neuroscience. She has bachelor’s degrees in zoology and philosophy and a master’s in the history and philosophy of science.

At 20,000 clicks from the extreme

Sharing living from a space station, ship or submarine. The examples of social media use in extreme conditions are multiplying and the public is asking for more. How to use public tools to highlight practices and discoveries? How to manage the use of social networks of a large organisation? What pitfalls to avoid? What does this mean for citizens and researchers?
Find out with Phillipe Archambault and Leslie Elliott experts in extrem conditions.

Philippe Archambault

Professor Philippe Archambault is a marine ecologist at Laval University, the director of the Notre Golfe network and president of the 4th World Conference on Marine Biodiversity. His research on the influence of global changes on biodiversity and the functioning of ecosystems has led him to work in all four corners of our oceans from the Arctic to the Antarctic, through Papua New Guinea and the French Polynesia.

Website

Leslie Elliott

Leslie Elliott leads a team of communicators at Ocean Networks Canada in Victoria, British Columbia, home to Canada’s world-leading ocean observatories in the Pacific and Arctic Oceans. Audiences can join robots equipped with high definition cameras via #livedive to discover more about our ocean.

Website

Science is not a joke!

Science and humor are two disciplines that might seem incompatible … and yet, like the ig-Nobels, humour can prove to be an excellent way to communicate a scientific message. This, however, can prove to be quite challenging since one needs to ensure they employ the right tone and language to both captivate the audience while simultaneously communicating complex topics.

Patrick Baud and Brian Malow, both well-renowned scientific communicators, will give you with the tools you need to capture your audience and also convey a proper scientific message. You will be surprised how, even in Science, a good dose of humour can make you laugh and think.

Patrick Baud
Patrick Baud is a French author who was born on June 30, 1979, in Avignon. He has been sharing for many years his passion for tales of fantasy, and the marvels and curiosities of the world, through different media: radio, web, novels, comic strips, conferences, and videos. His YouTube channel “Axolot”, was created in 2013, and now has over 420,000 followers.

Internet Website
Youtube

Brian Malow
Brian Malow is Earth’s Premier Science Comedian (self-proclaimed).  Brian has made science videos for Time Magazine and contributed to Neil deGrasse Tyson’s radio show.  He worked in science communications at a museum, blogged for Scientific American, and trains scientists to be better communicators.

Internet Website
YouTube

I don’t think they’ve managed to get everything up on YouTube yet but the material I’ve found has been subtitled (into French or English, depending on which language the speaker used).

Here are the opening day’s talks on YouTube with English subtitles or French subtitles when appropriate. You can also find some abstracts for the panel presentations here. I was particularly in this panel (S3 – The Importance of Reaching Out to Adults in Scientific Culture), Note: I have searched out the French language descriptions for those unavailable in English,

Organized by Coeur des sciences, Université du Québec à Montréal (UQAM)
Animator: Valérie Borde, Freelance Science Journalist

Anouk Gingras, Musée de la civilisation, Québec
Text not available in English

[La science au Musée de la civilisation c’est :
• Une cinquantaine d’expositions et espaces découvertes
• Des thèmes d’actualité, liés à des enjeux sociaux, pour des exposition souvent destinées aux adultes
• Un potentiel de nouveaux publics en lien avec les autres thématiques présentes au Musée (souvent non scientifiques)
L’exposition Nanotechnologies : l’invisible révolution :
• Un thème d’actualité suscitant une réflexion
• Un sujet sensible menant à la création d’un parcours d’exposition polarisé : choix entre « oui » ou « non » au développement des nanotechnologies pour l’avenir
• L’utilisation de divers éléments pour rapprocher le sujet du visiteur

  • Les nanotechnologies dans la science-fiction
  • Les objets du quotidien contenant des nanoparticules
  • Les objets anciens qui utilisant les nanotechnologies
  • Divers microscopes retraçant l’histoire des nanotechnologies

• Une forme d’interaction suscitant la réflexion du visiteur via un objet sympatique : le canard  de plastique jaune, muni d’une puce RFID

  • Sept stations de consultation qui incitent le visiteur à se prononcer et à réfléchir sur des questions éthiques liées au développement des nanotechnologies
  • Une compilation des données en temps réel
  • Une livraison des résultats personnalisée
  • Une mesure des visiteurs dont l’opinion s’est modifiée à la suite de la visite de l’exposition

Résultats de fréquentation :
• Public de jeunes adultes rejoint (51%)
• Plus d’hommes que de femmes ont visité l’exposition
• Parcours avec canard: incite à la réflexion et augmente l’attention
• 3 visiteurs sur 4 prennent le canard; 92% font l’activité en entier]

Marie Lambert-Chan, Québec Science
Capting the attention of adult readership : challenging mission, possible mission
Since 1962, Québec Science Magazine is the only science magazine aimed at an adult readership in Québec. Our mission : covering topical subjects related to science and technology, as well as social issues from a scientific point of view. Each year, we print eight issues, with a circulation of 22,000 copies. Furthermore, the magazine has received several awards and accolades. In 2017, Québec Science Magazine was honored by the Canadian Magazine Awards/Grands Prix du Magazine and was named Best Magazine in Science, Business and Politics category.
Although we have maintained a solid reputation among scientists and the media industry, our magazine is still relatively unknown to the general public. Why is that ? How is it that, through all those years, we haven’t found the right angle to engage a broader readership ?
We are still searching for definitive answers, but here are our observations :
Speaking science to adults is much more challenging than it is with children, who can marvel endlessly at the smallest things. Unfortunately, adults lose this capacity to marvel and wonder for various reasons : they have specific interests, they failed high-school science, they don’t feel competent enough to understand scientific phenomena. How do we bring the wonder back ? This is our mission. Not impossible, and hopefully soon to be accomplished. One noticible example is the number of reknown scientists interviewed during the popular talk-show Tout le monde en parle, leading us to believe the general public may have an interest in science.
However, to accomplish our mission, we have to recount science. According to the Bulgarian writer and blogger Maria Popova, great science writing should explain, elucidate and enchant . To explain : to make the information clear and comprehensible. To elucidate : to reveal all the interconnections between the pieces of information. To enchant : to go beyond the scientific terms and information and tell a story, thus giving a kaleidoscopic vision of the subject. This is how we intend to capture our readership’s attention.
Our team aims to accomplish this challenge. Although, to be perfectly honest, it would be much easier if we had more resources, financial-wise or human-wise. However, we don’t lack ideas. We dream of major scientific investigations, conferences organized around themes from the magazine’s issues, Web documentaries, podcasts… Such initiatives would give us the visibility we desperately crave.
That said, even in the best conditions, would be have more subscribers ? Perhaps. But it isn’t assured. Even if our magazine is aimed at adult readership, we are convinced that childhood and science go hand in hand, and is even decisive for the children’s future. At the moment, school programs are not in place for continuous scientific development. It is possible to develop an interest for scientific culture as adults, but it is much easier to achieve this level of curiosity if it was previously fostered.

Robert Lamontagne, Université de Montréal
Since the beginning of my career as an astrophysicist, I have been interested in scientific communication to non-specialist audiences. I have presented hundreds of lectures describing the phenomena of the cosmos. Initially, these were mainly offered in amateur astronomers’ clubs or in high-schools and Cégeps. Over the last few years, I have migrated to more general adult audiences in the context of cultural activities such as the “Festival des Laurentides”, the Arts, Culture and Society activities in Repentigny and, the Université du troisième âge (UTA) or Senior’s University.
The Quebec branch of the UTA, sponsored by the Université de Sherbrooke (UdeS), exists since 1976. Seniors universities, created in Toulouse, France, are part of a worldwide movement. The UdeS and its senior’s university antennas are members of the International Association of the Universities of the Third Age (AIUTA). The UTA is made up of 28 antennas located in 10 regions and reaches more than 10,000 people per year. Antenna volunteers prepare educational programming by drawing on a catalog of courses, seminars and lectures, covering a diverse range of subjects ranging from history and politics to health, science, or the environment.
The UTA is aimed at people aged 50 and over who wish to continue their training and learn throughout their lives. It is an attentive, inquisitive, educated public and, given the demographics in Canada, its number is growing rapidly. This segment of the population is often well off and very involved in society.
I usually use a two-prong approach.
• While remaining rigorous, the content is articulated around a few ideas, avoiding analytical expressions in favor of a qualitative description.
• The narrative framework, the story, which allows to contextualize the scientific content and to forge links with the audience.

Sophie Malavoy, Coeur des sciences – UQAM

Many obstacles need to be overcome in order to reach out to adults, especially those who aren’t in principle interested in science.
• Competing against cultural activities such as theater, movies, etc.
• The idea that science is complex and dull
• A feeling of incompetence. « I’ve always been bad in math and physics»
• Funding shortfall for activities which target adults
How to reach out to those adults?
• To put science into perspective. To bring its relevance out by making links with current events and big issues (economic, heath, environment, politic). To promote a transdisciplinary approach which includes humanities and social sciences.
• To stake on originality by offering uncommon and ludic experiences (scientific walks in the city, street performances, etc.)
• To bridge between science and popular activities to the public (science/music; science/dance; science/theater; science/sports; science/gastronomy; science/literature)
• To reach people with emotions without sensationalism. To boost their curiosity and ability to wonder.
• To put a human face on science, by insisting not only on the results of a research but on its process. To share the adventure lived by researchers.
• To liven up people’s feeling of competence. To insist on the scientific method.
• To invite non-scientists (citizens groups, communities, consumers, etc.) to the reflections on science issues (debate, etc.).  To move from dissemination of science to dialog

Didier Pourquery, The Conversation France
Text not available in English

[Depuis son lancement en septembre 2015 la plateforme The Conversation France (2 millions de pages vues par mois) n’a cessé de faire progresser son audience. Selon une étude menée un an après le lancement, la structure de lectorat était la suivante
Pour accrocher les adultes et les ainés deux axes sont intéressants ; nous les utilisons autant sur notre site que sur notre newsletter quotidienne – 26.000 abonnés- ou notre page Facebook (11500 suiveurs):
1/ expliquer l’actualité : donner les clefs pour comprendre les débats scientifiques qui animent la société ; mettre de la science dans les discussions (la mission du site est de  « nourrir le débat citoyen avec de l’expertise universitaire et de la recherche »). L’idée est de poser des questions de compréhension simple au moment où elles apparaissent dans le débat (en période électorale par exemple : qu’est-ce que le populisme ? Expliqué par des chercheurs de Sciences Po incontestables.)
Exemples : comprendre les conférences climat -COP21, COP22 – ; comprendre les débats de société (Gestation pour autrui); comprendre l’économie (revenu universel); comprendre les maladies neurodégénératives (Alzheimer) etc.
2/ piquer la curiosité : utiliser les formules classiques (le saviez-vous ?) appliquées à des sujets surprenants (par exemple : «  Que voit un chien quand il regarde la télé ? » a eu 96.000 pages vues) ; puis jouer avec ces articles sur les réseaux sociaux. Poser des questions simples et surprenantes. Par exemple : ressemblez-vous à votre prénom ? Cet article académique très sérieux a comptabilisé 95.000 pages vues en français et 171.000 en anglais.
3/ Susciter l’engagement : faire de la science participative simple et utile. Par exemple : appeler nos lecteurs à surveiller l’invasion de moustiques tigres partout sur le territoire. Cet article a eu 112.000 pages vues et a été republié largement sur d’autres sites. Autre exemple : appeler les lecteurs à photographier les punaises de leur environnement.]

Here are my very brief and very rough translations. (1) Anouk Gingras is focused largely on a nanotechnology exhibit and whether or not visitors went through it and participated in various activities. She doesn’t seem specifically focused on science communication for adults but they are doing some very interesting and related work at Québec’s Museum of Civilization. (2) Didier Pourquery is describing an online initiative known as ‘The Conversation France’ (strange—why not La conversation France?). Moving on, there’s a website with a daily newsletter (blog?) and a Facebook page. They have two main projects, one is a discussion of current science issues in society, which is informed with and by experts but is not exclusive to experts, and more curiosity-based science questions and discussion such as What does a dog see when it watches television?

Serendipity! I hadn’t stumbled across this conference when I posted my May 12, 2017 piece on the ‘insanity’ of science outreach in Canada. It’s good to see I’m not the only one focused on science outreach for adults and that there is some action, although seems to be a Québec-only effort.

(2) Ingenious—a book launch in Vancouver

The book will be launched on Thursday, June 1, 2017 at the Vancouver Public Library’s Central Branch (from the Ingenious: An Evening of Canadian Innovation event page)

Ingenious: An Evening of Canadian Innovation
Thursday, June 1, 2017 (6:30 pm – 8:00 pm)
Central Branch
Description

Gov. Gen. David Johnston and OpenText Corp. chair Tom Jenkins discuss Canadian innovation and their book Ingenious: How Canadian Innovators Made the World Smarter, Smaller, Kinder, Safer, Healthier, Wealthier and Happier.

Books will be available for purchase and signing.

Doors open at 6 p.m.

INGENIOUS : HOW CANADIAN INNOVATORS MADE THE WORLD SMARTER, SMALLER, KINDER, SAFER, HEALTHIER, WEALTHIER, AND HAPPIER

Address:

350 West Georgia St.
VancouverV6B 6B1

Get Directions

  • Phone:

Location Details:

Alice MacKay Room, Lower Level

I do have a few more details about the authors and their book. First, there’s this from the Ottawa Writer’s Festival March 28, 2017 event page,

To celebrate Canada’s 150th birthday, Governor General David Johnston and Tom Jenkins have crafted a richly illustrated volume of brilliant Canadian innovations whose widespread adoption has made the world a better place. From Bovril to BlackBerrys, lightbulbs to liquid helium, peanut butter to Pablum, this is a surprising and incredibly varied collection to make Canadians proud, and to our unique entrepreneurial spirit.

Successful innovation is always inspired by at least one of three forces — insight, necessity, and simple luck. Ingenious moves through history to explore what circumstances, incidents, coincidences, and collaborations motivated each great Canadian idea, and what twist of fate then brought that idea into public acceptance. Above all, the book explores what goes on in the mind of an innovator, and maps the incredible spectrum of personalities that have struggled to improve the lot of their neighbours, their fellow citizens, and their species.

From the marvels of aboriginal invention such as the canoe, snowshoe, igloo, dogsled, lifejacket, and bunk bed to the latest pioneering advances in medicine, education, philanthropy, science, engineering, community development, business, the arts, and the media, Canadians have improvised and collaborated their way to international admiration. …

Then, there’s this April 5, 2017 item on Canadian Broadcasting Corporation’s (CBC) news online,

From peanut butter to the electric wheelchair, the stories behind numerous life-changing Canadian innovations are detailed in a new book.

Gov. Gen. David Johnston and Tom Jenkins, chair of the National Research Council and former CEO of OpenText, are the authors of Ingenious: How Canadian Innovators Made the World Smarter, Smaller, Kinder, Safer, Healthier, Wealthier and Happier. The authors hope their book reinforces and extends the culture of innovation in Canada.

“We started wanting to tell 50 stories of Canadian innovators, and what has amazed Tom and myself is how many there are,” Johnston told The Homestretch on Wednesday. The duo ultimately chronicled 297 innovations in the book, including the pacemaker, life jacket and chocolate bars.

“Innovations are not just technological, not just business, but they’re social innovations as well,” Johnston said.

Many of those innovations, and the stories behind them, are not well known.

“We’re sort of a humble people,” Jenkins said. “We’re pretty quiet. We don’t brag, we don’t talk about ourselves very much, and so we then lead ourselves to believe as a culture that we’re not really good inventors, the Americans are. And yet we knew that Canadians were actually great inventors and innovators.”

‘Opportunities and challenges’

For Johnston, his favourite story in the book is on the light bulb.

“It’s such a symbol of both our opportunities and challenges,” he said. “The light bulb was invented in Canada, not the United States. It was two inventors back in the 1870s that realized that if you passed an electric current through a resistant metal it would glow, and they patented that, but then they didn’t have the money to commercialize it.”

American inventor Thomas Edison went on to purchase that patent and made changes to the original design.

Johnston and Jenkins are also inviting readers to share their own innovation stories, on the book’s website.

I’m looking forward to the talk and wondering if they’ve included the botox and cellulose nanocrystal (CNC) stories to the book. BTW, Tom Jenkins was the chair of a panel examining Canadian research and development and lead author of the panel’s report (Innovation Canada: A Call to Action) for the then Conservative government (it’s also known as the Jenkins report). You can find out more about in my Oct. 21, 2011 posting.

(3) Made in Canada (Vancouver)

This is either fortuitous or there’s some very high level planning involved in the ‘Made in Canada; Inspiring Creativity and Innovation’ show which runs from April 21 – Sept. 4, 2017 at Vancouver’s Science World (also known as the Telus World of Science). From the Made in Canada; Inspiring Creativity and Innovation exhibition page,

Celebrate Canadian creativity and innovation, with Science World’s original exhibition, Made in Canada, presented by YVR [Vancouver International Airport] — where you drive the creative process! Get hands-on and build the fastest bobsled, construct a stunning piece of Vancouver architecture and create your own Canadian sound mashup, to share with friends.

Vote for your favourite Canadian inventions and test fly a plane of your design. Discover famous (and not-so-famous, but super neat) Canadian inventions. Learn about amazing, local innovations like robots that teach themselves, one-person electric cars and a computer that uses parallel universes.

Imagine what you can create here, eh!!

You can find more information here.

One quick question, why would Vancouver International Airport be presenting this show? I asked that question of Science World’s Communications Coordinator, Jason Bosher, and received this response,

 YVR is the presenting sponsor. They donated money to the exhibition and they also contributed an exhibit for the “We Move” themed zone in the Made in Canada exhibition. The YVR exhibit details the history of the YVR airport, it’s geographic advantage and some of the planes they have seen there.

I also asked if there was any connection between this show and the ‘Ingenious’ book launch,

Some folks here are aware of the book launch. It has to do with the Canada 150 initiative and nothing to do with the Made in Canada exhibition, which was developed here at Science World. It is our own original exhibition.

So there you have it.

(4) Robotics, AI, and the future of work (Ottawa)

I’m glad to finally stumble across a Canadian event focusing on the topic of artificial intelligence (AI), robotics and the future of work. Sadly (for me), this is taking place in Ottawa. Here are more details  from the May 25, 2017 notice (received via email) from the Canadian Science Policy Centre (CSPC),

CSPC is Partnering with CIFAR {Canadian Institute for Advanced Research]
The Second Annual David Dodge Lecture

Join CIFAR and Senior Fellow Daron Acemoglu for
the Second Annual David Dodge CIFAR Lecture in Ottawa on June 13.
June 13, 2017 | 12 – 2 PM [emphasis mine]
Fairmont Château Laurier, Drawing Room | 1 Rideau St, Ottawa, ON
Along with the backlash against globalization and the outsourcing of jobs, concern is also growing about the effect that robotics and artificial intelligence will have on the labour force in advanced industrial nations. World-renowned economist Acemoglu, author of the best-selling book Why Nations Fail, will discuss how technology is changing the face of work and the composition of labour markets. Drawing on decades of data, Acemoglu explores the effects of widespread automation on manufacturing jobs, the changes we can expect from artificial intelligence technologies, and what responses to these changes might look like. This timely discussion will provide valuable insights for current and future leaders across government, civil society, and the private sector.

Daron Acemoglu is a Senior Fellow in CIFAR’s Insitutions, Organizations & Growth program, and the Elizabeth and James Killian Professor of Economics at the Massachusetts Institute of Technology.

Tickets: $15 (A light lunch will be served.)

You can find a registration link here. Also, if you’re interested in the Canadian efforts in the field of artificial intelligence you can find more in my March 24, 2017 posting (scroll down about 25% of the way and then about 40% of the way) on the 2017 Canadian federal budget and science where I first noted the $93.7M allocated to CIFAR for launching a Pan-Canadian Artificial Intelligence Strategy.

(5) June 2017 edition of the Curiosity Collider Café (Vancouver)

This is an art/science (also known called art/sci and SciArt) that has taken place in Vancouver every few months since April 2015. Here’s more about the June 2017 edition (from the Curiosity Collider events page),

Collider Cafe

When
8:00pm on Wednesday, June 21st, 2017. Door opens at 7:30pm.

Where
Café Deux Soleils. 2096 Commercial Drive, Vancouver, BC (Google Map).

Cost
$5.00-10.00 cover at the door (sliding scale). Proceeds will be used to cover the cost of running this event, and to fund future Curiosity Collider events. Curiosity Collider is a registered BC non-profit organization.

***

#ColliderCafe is a space for artists, scientists, makers, and anyone interested in art+science. Meet, discover, connect, create. How do you explore curiosity in your life? Join us and discover how our speakers explore their own curiosity at the intersection of art & science.

The event will start promptly at 8pm (doors open at 7:30pm). $5.00-10.00 (sliding scale) cover at the door. Proceeds will be used to cover the cost of running this event, and to fund future Curiosity Collider events. Curiosity Collider is a registered BC non-profit organization.

Enjoy!

*I changed ‘three’ events to ‘five’ events and added a number to each event for greater reading ease on May 31, 2017.

An explanation of neural networks from the Massachusetts Institute of Technology (MIT)

I always enjoy the MIT ‘explainers’ and have been a little sad that I haven’t stumbled across one in a while. Until now, that is. Here’s an April 14, 201 neural network ‘explainer’ (in its entirety) by Larry Hardesty (?),

In the past 10 years, the best-performing artificial-intelligence systems — such as the speech recognizers on smartphones or Google’s latest automatic translator — have resulted from a technique called “deep learning.”

Deep learning is in fact a new name for an approach to artificial intelligence called neural networks, which have been going in and out of fashion for more than 70 years. Neural networks were first proposed in 1944 by Warren McCullough and Walter Pitts, two University of Chicago researchers who moved to MIT in 1952 as founding members of what’s sometimes called the first cognitive science department.

Neural nets were a major area of research in both neuroscience and computer science until 1969, when, according to computer science lore, they were killed off by the MIT mathematicians Marvin Minsky and Seymour Papert, who a year later would become co-directors of the new MIT Artificial Intelligence Laboratory.

The technique then enjoyed a resurgence in the 1980s, fell into eclipse again in the first decade of the new century, and has returned like gangbusters in the second, fueled largely by the increased processing power of graphics chips.

“There’s this idea that ideas in science are a bit like epidemics of viruses,” says Tomaso Poggio, the Eugene McDermott Professor of Brain and Cognitive Sciences at MIT, an investigator at MIT’s McGovern Institute for Brain Research, and director of MIT’s Center for Brains, Minds, and Machines. “There are apparently five or six basic strains of flu viruses, and apparently each one comes back with a period of around 25 years. People get infected, and they develop an immune response, and so they don’t get infected for the next 25 years. And then there is a new generation that is ready to be infected by the same strain of virus. In science, people fall in love with an idea, get excited about it, hammer it to death, and then get immunized — they get tired of it. So ideas should have the same kind of periodicity!”

Weighty matters

Neural nets are a means of doing machine learning, in which a computer learns to perform some task by analyzing training examples. Usually, the examples have been hand-labeled in advance. An object recognition system, for instance, might be fed thousands of labeled images of cars, houses, coffee cups, and so on, and it would find visual patterns in the images that consistently correlate with particular labels.

Modeled loosely on the human brain, a neural net consists of thousands or even millions of simple processing nodes that are densely interconnected. Most of today’s neural nets are organized into layers of nodes, and they’re “feed-forward,” meaning that data moves through them in only one direction. An individual node might be connected to several nodes in the layer beneath it, from which it receives data, and several nodes in the layer above it, to which it sends data.

To each of its incoming connections, a node will assign a number known as a “weight.” When the network is active, the node receives a different data item — a different number — over each of its connections and multiplies it by the associated weight. It then adds the resulting products together, yielding a single number. If that number is below a threshold value, the node passes no data to the next layer. If the number exceeds the threshold value, the node “fires,” which in today’s neural nets generally means sending the number — the sum of the weighted inputs — along all its outgoing connections.

When a neural net is being trained, all of its weights and thresholds are initially set to random values. Training data is fed to the bottom layer — the input layer — and it passes through the succeeding layers, getting multiplied and added together in complex ways, until it finally arrives, radically transformed, at the output layer. During training, the weights and thresholds are continually adjusted until training data with the same labels consistently yield similar outputs.

Minds and machines

The neural nets described by McCullough and Pitts in 1944 had thresholds and weights, but they weren’t arranged into layers, and the researchers didn’t specify any training mechanism. What McCullough and Pitts showed was that a neural net could, in principle, compute any function that a digital computer could. The result was more neuroscience than computer science: The point was to suggest that the human brain could be thought of as a computing device.

Neural nets continue to be a valuable tool for neuroscientific research. For instance, particular network layouts or rules for adjusting weights and thresholds have reproduced observed features of human neuroanatomy and cognition, an indication that they capture something about how the brain processes information.

The first trainable neural network, the Perceptron, was demonstrated by the Cornell University psychologist Frank Rosenblatt in 1957. The Perceptron’s design was much like that of the modern neural net, except that it had only one layer with adjustable weights and thresholds, sandwiched between input and output layers.

Perceptrons were an active area of research in both psychology and the fledgling discipline of computer science until 1959, when Minsky and Papert published a book titled “Perceptrons,” which demonstrated that executing certain fairly common computations on Perceptrons would be impractically time consuming.

“Of course, all of these limitations kind of disappear if you take machinery that is a little more complicated — like, two layers,” Poggio says. But at the time, the book had a chilling effect on neural-net research.

“You have to put these things in historical context,” Poggio says. “They were arguing for programming — for languages like Lisp. Not many years before, people were still using analog computers. It was not clear at all at the time that programming was the way to go. I think they went a little bit overboard, but as usual, it’s not black and white. If you think of this as this competition between analog computing and digital computing, they fought for what at the time was the right thing.”

Periodicity

By the 1980s, however, researchers had developed algorithms for modifying neural nets’ weights and thresholds that were efficient enough for networks with more than one layer, removing many of the limitations identified by Minsky and Papert. The field enjoyed a renaissance.

But intellectually, there’s something unsatisfying about neural nets. Enough training may revise a network’s settings to the point that it can usefully classify data, but what do those settings mean? What image features is an object recognizer looking at, and how does it piece them together into the distinctive visual signatures of cars, houses, and coffee cups? Looking at the weights of individual connections won’t answer that question.

In recent years, computer scientists have begun to come up with ingenious methods for deducing the analytic strategies adopted by neural nets. But in the 1980s, the networks’ strategies were indecipherable. So around the turn of the century, neural networks were supplanted by support vector machines, an alternative approach to machine learning that’s based on some very clean and elegant mathematics.

The recent resurgence in neural networks — the deep-learning revolution — comes courtesy of the computer-game industry. The complex imagery and rapid pace of today’s video games require hardware that can keep up, and the result has been the graphics processing unit (GPU), which packs thousands of relatively simple processing cores on a single chip. It didn’t take long for researchers to realize that the architecture of a GPU is remarkably like that of a neural net.

Modern GPUs enabled the one-layer networks of the 1960s and the two- to three-layer networks of the 1980s to blossom into the 10-, 15-, even 50-layer networks of today. That’s what the “deep” in “deep learning” refers to — the depth of the network’s layers. And currently, deep learning is responsible for the best-performing systems in almost every area of artificial-intelligence research.

Under the hood

The networks’ opacity is still unsettling to theorists, but there’s headway on that front, too. In addition to directing the Center for Brains, Minds, and Machines (CBMM), Poggio leads the center’s research program in Theoretical Frameworks for Intelligence. Recently, Poggio and his CBMM colleagues have released a three-part theoretical study of neural networks.

The first part, which was published last month in the International Journal of Automation and Computing, addresses the range of computations that deep-learning networks can execute and when deep networks offer advantages over shallower ones. Parts two and three, which have been released as CBMM technical reports, address the problems of global optimization, or guaranteeing that a network has found the settings that best accord with its training data, and overfitting, or cases in which the network becomes so attuned to the specifics of its training data that it fails to generalize to other instances of the same categories.

There are still plenty of theoretical questions to be answered, but CBMM researchers’ work could help ensure that neural networks finally break the generational cycle that has brought them in and out of favor for seven decades.

This image from MIT illustrates a ‘modern’ neural network,

Most applications of deep learning use “convolutional” neural networks, in which the nodes of each layer are clustered, the clusters overlap, and each cluster feeds data to multiple nodes (orange and green) of the next layer. Image: Jose-Luis Olivares/MIT

h/t phys.org April 17, 2017

One final note, I wish the folks at MIT had an ‘explainer’ archive. I’m not sure how to find any more ‘explainers on MIT’s website.

Machine learning programs learn bias

The notion of bias in artificial intelligence (AI)/algorithms/robots is gaining prominence (links to other posts featuring algorithms and bias are at the end of this post). The latest research concerns machine learning where an artificial intelligence system trains itself with ordinary human language from the internet. From an April 13, 2017 American Association for the Advancement of Science (AAAS) news release on EurekAlert,

As artificial intelligence systems “learn” language from existing texts, they exhibit the same biases that humans do, a new study reveals. The results not only provide a tool for studying prejudicial attitudes and behavior in humans, but also emphasize how language is intimately intertwined with historical biases and cultural stereotypes. A common way to measure biases in humans is the Implicit Association Test (IAT), where subjects are asked to pair two concepts they find similar, in contrast to two concepts they find different; their response times can vary greatly, indicating how well they associated one word with another (for example, people are more likely to associate “flowers” with “pleasant,” and “insects” with “unpleasant”). Here, Aylin Caliskan and colleagues developed a similar way to measure biases in AI systems that acquire language from human texts; rather than measuring lag time, however, they used the statistical number of associations between words, analyzing roughly 2.2 million words in total. Their results demonstrate that AI systems retain biases seen in humans. For example, studies of human behavior show that the exact same resume is 50% more likely to result in an opportunity for an interview if the candidate’s name is European American rather than African-American. Indeed, the AI system was more likely to associate European American names with “pleasant” stimuli (e.g. “gift,” or “happy”). In terms of gender, the AI system also reflected human biases, where female words (e.g., “woman” and “girl”) were more associated than male words with the arts, compared to mathematics. In a related Perspective, Anthony G. Greenwald discusses these findings and how they could be used to further analyze biases in the real world.

There are more details about the research in this April 13, 2017 Princeton University news release on EurekAlert (also on ScienceDaily),

In debates over the future of artificial intelligence, many experts think of the new systems as coldly logical and objectively rational. But in a new study, researchers have demonstrated how machines can be reflections of us, their creators, in potentially problematic ways. Common machine learning programs, when trained with ordinary human language available online, can acquire cultural biases embedded in the patterns of wording, the researchers found. These biases range from the morally neutral, like a preference for flowers over insects, to the objectionable views of race and gender.

Identifying and addressing possible bias in machine learning will be critically important as we increasingly turn to computers for processing the natural language humans use to communicate, for instance in doing online text searches, image categorization and automated translations.

“Questions about fairness and bias in machine learning are tremendously important for our society,” said researcher Arvind Narayanan, an assistant professor of computer science and an affiliated faculty member at the Center for Information Technology Policy (CITP) at Princeton University, as well as an affiliate scholar at Stanford Law School’s Center for Internet and Society. “We have a situation where these artificial intelligence systems may be perpetuating historical patterns of bias that we might find socially unacceptable and which we might be trying to move away from.”

The paper, “Semantics derived automatically from language corpora contain human-like biases,” published April 14  [2017] in Science. Its lead author is Aylin Caliskan, a postdoctoral research associate and a CITP fellow at Princeton; Joanna Bryson, a reader at University of Bath, and CITP affiliate, is a coauthor.

As a touchstone for documented human biases, the study turned to the Implicit Association Test, used in numerous social psychology studies since its development at the University of Washington in the late 1990s. The test measures response times (in milliseconds) by human subjects asked to pair word concepts displayed on a computer screen. Response times are far shorter, the Implicit Association Test has repeatedly shown, when subjects are asked to pair two concepts they find similar, versus two concepts they find dissimilar.

Take flower types, like “rose” and “daisy,” and insects like “ant” and “moth.” These words can be paired with pleasant concepts, like “caress” and “love,” or unpleasant notions, like “filth” and “ugly.” People more quickly associate the flower words with pleasant concepts, and the insect terms with unpleasant ideas.

The Princeton team devised an experiment with a program where it essentially functioned like a machine learning version of the Implicit Association Test. Called GloVe, and developed by Stanford University researchers, the popular, open-source program is of the sort that a startup machine learning company might use at the heart of its product. The GloVe algorithm can represent the co-occurrence statistics of words in, say, a 10-word window of text. Words that often appear near one another have a stronger association than those words that seldom do.

The Stanford researchers turned GloVe loose on a huge trawl of contents from the World Wide Web, containing 840 billion words. Within this large sample of written human culture, Narayanan and colleagues then examined sets of so-called target words, like “programmer, engineer, scientist” and “nurse, teacher, librarian” alongside two sets of attribute words, such as “man, male” and “woman, female,” looking for evidence of the kinds of biases humans can unwittingly possess.

In the results, innocent, inoffensive biases, like for flowers over bugs, showed up, but so did examples along lines of gender and race. As it turned out, the Princeton machine learning experiment managed to replicate the broad substantiations of bias found in select Implicit Association Test studies over the years that have relied on live, human subjects.

For instance, the machine learning program associated female names more with familial attribute words, like “parents” and “wedding,” than male names. In turn, male names had stronger associations with career attributes, like “professional” and “salary.” Of course, results such as these are often just objective reflections of the true, unequal distributions of occupation types with respect to gender–like how 77 percent of computer programmers are male, according to the U.S. Bureau of Labor Statistics.

Yet this correctly distinguished bias about occupations can end up having pernicious, sexist effects. An example: when foreign languages are naively processed by machine learning programs, leading to gender-stereotyped sentences. The Turkish language uses a gender-neutral, third person pronoun, “o.” Plugged into the well-known, online translation service Google Translate, however, the Turkish sentences “o bir doktor” and “o bir hem?ire” with this gender-neutral pronoun are translated into English as “he is a doctor” and “she is a nurse.”

“This paper reiterates the important point that machine learning methods are not ‘objective’ or ‘unbiased’ just because they rely on mathematics and algorithms,” said Hanna Wallach, a senior researcher at Microsoft Research New York City, who was not involved in the study. “Rather, as long as they are trained using data from society and as long as society exhibits biases, these methods will likely reproduce these biases.”

Another objectionable example harkens back to a well-known 2004 paper by Marianne Bertrand of the University of Chicago Booth School of Business and Sendhil Mullainathan of Harvard University. The economists sent out close to 5,000 identical resumes to 1,300 job advertisements, changing only the applicants’ names to be either traditionally European American or African American. The former group was 50 percent more likely to be offered an interview than the latter. In an apparent corroboration of this bias, the new Princeton study demonstrated that a set of African American names had more unpleasantness associations than a European American set.

Computer programmers might hope to prevent cultural stereotype perpetuation through the development of explicit, mathematics-based instructions for the machine learning programs underlying AI systems. Not unlike how parents and mentors try to instill concepts of fairness and equality in children and students, coders could endeavor to make machines reflect the better angels of human nature.

“The biases that we studied in the paper are easy to overlook when designers are creating systems,” said Narayanan. “The biases and stereotypes in our society reflected in our language are complex and longstanding. Rather than trying to sanitize or eliminate them, we should treat biases as part of the language and establish an explicit way in machine learning of determining what we consider acceptable and unacceptable.”

Here’s a link to and a citation for the Princeton paper,

Semantics derived automatically from language corpora contain human-like biases by Aylin Caliskan, Joanna J. Bryson, Arvind Narayanan. Science  14 Apr 2017: Vol. 356, Issue 6334, pp. 183-186 DOI: 10.1126/science.aal4230

This paper appears to be open access.

Links to more cautionary posts about AI,

Aug 5, 2009: Autonomous algorithms; intelligent windows; pretty nano pictures

June 14, 2016:  Accountability for artificial intelligence decision-making

Oct. 25, 2016 Removing gender-based stereotypes from algorithms

March 1, 2017: Algorithms in decision-making: a government inquiry in the UK

There’s also a book which makes some of the current use of AI programmes and big data quite accessible reading: Cathy O’Neal’s ‘Weapons of Math Destruction: How Big Data Increases Inequality and Threatens Democracy’.