Tag Archives: artificial intelligence (AI)

Neurotransistor for brainlike (neuromorphic) computing

According to researchers at Helmholtz-Zentrum Dresden-Rossendorf and the rest of the international team collaborating on the work, it’s time to look more closely at plasticity in the neuronal membrane,.

From the abstract for their paper, Intrinsic plasticity of silicon nanowire neurotransistors for dynamic memory and learning functions by Eunhye Baek, Nikhil Ranjan Das, Carlo Vittorio Cannistraci, Taiuk Rim, Gilbert Santiago Cañón Bermúdez, Khrystyna Nych, Hyeonsu Cho, Kihyun Kim, Chang-Ki Baek, Denys Makarov, Ronald Tetzlaff, Leon Chua, Larysa Baraban & Gianaurelio Cuniberti. Nature Electronics volume 3, pages 398–408 (2020) DOI: https://doi.org/10.1038/s41928-020-0412-1 Published online: 25 May 2020 Issue Date: July 2020

Neuromorphic architectures merge learning and memory functions within a single unit cell and in a neuron-like fashion. Research in the field has been mainly focused on the plasticity of artificial synapses. However, the intrinsic plasticity of the neuronal membrane is also important in the implementation of neuromorphic information processing. Here we report a neurotransistor made from a silicon nanowire transistor coated by an ion-doped sol–gel silicate film that can emulate the intrinsic plasticity of the neuronal membrane.

Caption: Neurotransistors: from silicon chips to neuromorphic architecture. Credit: TU Dresden / E. Baek Courtesy: Helmholtz-Zentrum Dresden-Rossendorf

A July 14, 2020 news item on Nanowerk announced the research (Note: A link has been removed),

Especially activities in the field of artificial intelligence, like teaching robots to walk or precise automatic image recognition, demand ever more powerful, yet at the same time more economical computer chips. While the optimization of conventional microelectronics is slowly reaching its physical limits, nature offers us a blueprint how information can be processed and stored quickly and efficiently: our own brain.

For the very first time, scientists at TU Dresden and the Helmholtz-Zentrum Dresden-Rossendorf (HZDR) have now successfully imitated the functioning of brain neurons using semiconductor materials. They have published their research results in the journal Nature Electronics (“Intrinsic plasticity of silicon nanowire neurotransistors for dynamic memory and learning functions”).

A July 14, 2020 Helmholtz-Zentrum Dresden-Rossendorf press release (also on EurekAlert), which originated the news items delves further into the research,

Today, enhancing the performance of microelectronics is usually achieved by reducing component size, especially of the individual transistors on the silicon computer chips. “But that can’t go on indefinitely – we need new approaches”, Larysa Baraban asserts. The physicist, who has been working at HZDR since the beginning of the year, is one of the three primary authors of the international study, which involved a total of six institutes. One approach is based on the brain, combining data processing with data storage in an artificial neuron.

“Our group has extensive experience with biological and chemical electronic sensors,” Baraban continues. “So, we simulated the properties of neurons using the principles of biosensors and modified a classical field-effect transistor to create an artificial neurotransistor.” The advantage of such an architecture lies in the simultaneous storage and processing of information in a single component. In conventional transistor technology, they are separated, which slows processing time and hence ultimately also limits performance.

Silicon wafer + polymer = chip capable of learning

Modeling computers on the human brain is no new idea. Scientists made attempts to hook up nerve cells to electronics in Petri dishes decades ago. “But a wet computer chip that has to be fed all the time is of no use to anybody,” says Gianaurelio Cuniberti from TU Dresden. The Professor for Materials Science and Nanotechnology is one of the three brains behind the neurotransistor alongside Ronald Tetzlaff, Professor of Fundamentals of Electrical Engineering in Dresden, and Leon Chua [emphasis mine] from the University of California at Berkeley, who had already postulated similar components in the early 1970s.

Now, Cuniberti, Baraban and their team have been able to implement it: “We apply a viscous substance – called solgel – to a conventional silicon wafer with circuits. This polymer hardens and becomes a porous ceramic,” the materials science professor explains. “Ions move between the holes. They are heavier than electrons and slower to return to their position after excitation. This delay, called hysteresis, is what causes the storage effect.” As Cuniberti explains, this is a decisive factor in the functioning of the transistor. “The more an individual transistor is excited, the sooner it will open and let the current flow. This strengthens the connection. The system is learning.”

Cuniberti and his team are not focused on conventional issues, though. “Computers based on our chip would be less precise and tend to estimate mathematical computations rather than calculating them down to the last decimal,” the scientist explains. “But they would be more intelligent. For example, a robot with such processors would learn to walk or grasp; it would possess an optical system and learn to recognize connections. And all this without having to develop any software.” But these are not the only advantages of neuromorphic computers. Thanks to their plasticity, which is similar to that of the human brain, they can adapt to changing tasks during operation and, thus, solve problems for which they were not originally programmed.

I highlighted Dr. Leon Chua’s name as he was one of the first to conceptualize the notion of a memristor (memory resistor), which is what the press release seems to be referencing with the mention of artificial synapses. Dr. Chua very kindly answered a few questions for me about his work which I published in an April 13, 2010 posting (scroll down about 40% of the way).

Brain-inspired computer with optimized neural networks

Caption: Left to right: The experiment was performed on a prototype of the BrainScales-2 chip; Schematic representation of a neural network; Results for simple and complex tasks. Credit: Heidelberg University

I don’t often stumble across research from the European Union’s flagship Human Brain Project. So, this is a delightful occurrence especially with my interest in neuromorphic computing. From a July 22, 2020 Human Brain Project press release (also on EurekAlert),

Many computational properties are maximized when the dynamics of a network are at a “critical point”, a state where systems can quickly change their overall characteristics in fundamental ways, transitioning e.g. between order and chaos or stability and instability. Therefore, the critical state is widely assumed to be optimal for any computation in recurrent neural networks, which are used in many AI [artificial intelligence] applications.

Researchers from the HBP [Human Brain Project] partner Heidelberg University and the Max-Planck-Institute for Dynamics and Self-Organization challenged this assumption by testing the performance of a spiking recurrent neural network on a set of tasks with varying complexity at – and away from critical dynamics. They instantiated the network on a prototype of the analog neuromorphic BrainScaleS-2 system. BrainScaleS is a state-of-the-art brain-inspired computing system with synaptic plasticity implemented directly on the chip. It is one of two neuromorphic systems currently under development within the European Human Brain Project.

First, the researchers showed that the distance to criticality can be easily adjusted in the chip by changing the input strength, and then demonstrated a clear relation between criticality and task-performance. The assumption that criticality is beneficial for every task was not confirmed: whereas the information-theoretic measures all showed that network capacity was maximal at criticality, only the complex, memory intensive tasks profited from it, while simple tasks actually suffered. The study thus provides a more precise understanding of how the collective network state should be tuned to different task requirements for optimal performance.

Mechanistically, the optimal working point for each task can be set very easily under homeostatic plasticity by adapting the mean input strength. The theory behind this mechanism was developed very recently at the Max Planck Institute. “Putting it to work on neuromorphic hardware shows that these plasticity rules are very capable in tuning network dynamics to varying distances from criticality”, says senior author Viola Priesemann, group leader at MPIDS. Thereby tasks of varying complexity can be solved optimally within that space.

The finding may also explain why biological neural networks operate not necessarily at criticality, but in the dynamically rich vicinity of a critical point, where they can tune their computation properties to task requirements. Furthermore, it establishes neuromorphic hardware as a fast and scalable avenue to explore the impact of biological plasticity rules on neural computation and network dynamics.

“As a next step, we now study and characterize the impact of the spiking network’s working point on classifying artificial and real-world spoken words”, says first author Benjamin Cramer of Heidelberg University.

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

Control of criticality and computation in spiking neuromorphic networks with plasticity by Benjamin Cramer, David Stöckel, Markus Kreft, Michael Wibral, Johannes Schemmel, Karlheinz Meier & Viola Priesemann. Nature Communications volume 11, Article number: 2853 (2020) DOI: https://doi.org/10.1038/s41467-020-16548-3 Published: 05 June 2020

This paper is open access.

News from the Canadian Light Source (CLS), Canadian Science Policy Conference (CSPC) 2020, the International Symposium on Electronic Arts (ISEA) 2020, and HotPopRobot

I have some news about conserving art; early bird registration deadlines for two events, and, finally, an announcement about contest winners.

Canadian Light Source (CLS) and modern art

Rita Letendre. Victoire [Victory], 1961. Oil on canvas, Overall: 202.6 × 268 cm. Art Gallery of Ontario. Gift of Jessie and Percy Waxer, 1974, donated by the Ontario Heritage Foundation, 1988. © Rita Letendre L74.8. Photography by Ian Lefebvre

This is one of three pieces by Rita Letendre that underwent chemical mapping according to an August 5, 2020 CLS news release by Victoria Martinez (also received via email),

Research undertaken at the Canadian Light Source (CLS) at the University of Saskatchewan was key to understanding how to conserve experimental oil paintings by Rita Letendre, one of Canada’s most respected living abstract artists.

The work done at the CLS was part of a collaborative research project between the Art Gallery of Ontario (AGO) and the Canadian Conservation Institute (CCI) that came out of a recent retrospective Rita Letendre: Fire & Light at the AGO. During close examination, Meaghan Monaghan, paintings conservator from the Michael and Sonja Koerner Centre for Conservation, observed that several of Letendre’s oil paintings from the fifties and sixties had suffered significant degradation, most prominently, uneven gloss and patchiness, snowy crystalline structures coating the surface known as efflorescence, and cracking and lifting of the paint in several areas.

Kate Helwig, Senior Conservation Scientist at the Canadian Conservation Institute, says these problems are typical of mid-20th century oil paintings. “We focused on three of Rita Letendre’s paintings in the AGO collection, which made for a really nice case study of her work and also fits into the larger question of why oil paintings from that period tend to have degradation issues.”

Growing evidence indicates that paintings from this period have experienced these problems due to the combination of the experimental techniques many artists employed and the additives paint manufacturers had begun to use.

In order to determine more precisely how these factors affected Letendre’s paintings, the research team members applied a variety of analytical techniques, using microscopic samples taken from key points in the works.

“The work done at the CLS was particularly important because it allowed us to map the distribution of materials throughout a paint layer such as an impasto stroke,” Helwig said. The team used Mid-IR chemical mapping at the facility, which provides a map of different molecules in a small sample.

For example, chemical mapping at the CLS allowed the team to understand the distribution of the paint additive aluminum stearate throughout the paint layers of the painting Méduse. This painting showed areas of soft, incompletely dried paint, likely due to the high concentration and incomplete mixing of this additive. 

The painting Victoire had a crumbling base paint layer in some areas and cracking and efflorescence at the surface in others.  Infrared mapping at the CLS allowed the team to determine that excess free fatty acids in the paint were linked to both problems; where the fatty acids were found at the base they formed zing “soaps” which led to crumbling and cracking, and where they had moved to the surface they had crystallized, causing the snowflake-like efflorescence.

AGO curators and conservators interviewed Letendre to determine what was important to her in preserving and conserving her works, and she highlighted how important an even gloss across the surface was to her artworks, and the philosophical importance of the colour black in her paintings. These priorities guided conservation efforts, while the insights gained through scientific research will help maintain the works in the long term.

In order to restore the black paint to its intended even finish for display, conservator Meaghan Monaghan removed the white crystallization from the surface of Victoire, but it is possible that it could begin to recur. Understanding the processes that lead to this degradation will be an important tool to keep Letendre’s works in good condition.

“The world of modern paint research is complicated; each painting is unique, which is why it’s important to combine theoretical work on model paint systems with this kind of case study on actual works of art” said Helwig. The team hopes to collaborate on studying a larger cross section of Letendre’s paintings in oil and acrylic in the future to add to the body of knowledge.

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

Rita Letendre’s Oil Paintings from the 1960s: The Effect of Artist’s Materials on Degradation Phenomena by Kate Helwig, Meaghan Monaghan, Jennifer Poulin, Eric J. Henderson & Maeve Moriarty. Studies in Conservation (2020): 1-15 DOI: https://doi.org/10.1080/00393630.2020.1773055 Published online: 06 Jun 2020

This paper is behind a paywall.

Canadian Science Policy Conference (CSPC) 2020

The latest news from the CSPC 2020 (November 16 – 20 with preconference events from Nov. 1 -14) organizers is that registration is open and early birds have a deadline of September 27, 2020 (from an August 6, 2020 CSPC 2020 announcement received via email),

It’s time! Registration for the 12th Canadian Science Policy Conference (CSPC 2020) is open now. Early Bird registration is valid until Sept. 27th [2020].

CSPC 2020 is coming to your offices and homes:

Register for full access to 3 weeks of programming of the biggest science and innovation policy forum of 2020 under the overarching theme: New Decade, New Realities: Hindsight, Insight, Foresight.

2500+ Participants

300+ Speakers from five continents

65+ Panel sessions, 15 pre conference sessions and symposiums

50+ On demand videos and interviews with the most prominent figures of science and innovation policy 

20+ Partner-hosted functions

15+ Networking sessions

15 Open mic sessions to discuss specific topics

The virtual conference features an exclusive array of offerings:

3D Lounge and Exhibit area

Advance access to the Science Policy Magazine, featuring insightful reflections from the frontier of science and policy innovation

Many more

Don’t miss this unique opportunity to engage in the most important discussions of science and innovation policy with insights from around the globe, just from your office, home desk, or your mobile phone.

Benefit from significantly reduced registration fees for an online conference with an option for discount for multiple ticket purchases

Register now to benefit from the Early Bird rate!

The preliminary programme can be found here. This year there will be some discussion of a Canadian synthetic biology roadmap, presentations on various Indigenous concerns (mostly health), a climate challenge presentation focusing on Mexico and social vulnerability and another on parallels between climate challenges and COVID-19. There are many presentations focused on COVID-19 and.or health.

There doesn’t seem to be much focus on cyber security and, given that we just lost two ice caps (see Brandon Spektor’s August 1, 2020 article [Two Canadian ice caps have completely vanished from the Arctic, NASA imagery shows] on the Live Science website), it’s surprising that there are no presentations concerning the Arctic.

International Symposium on Electronic Arts (ISEA) 2020

According to my latest information, the early bird rate for ISEA 2020 (Oct. 13 -18) ends on August 13, 2020. (My June 22, 2020 posting describes their plans for the online event.)

You can find registration information here.

Margaux Davoine has written up a teaser for the 2020 edition of ISEA in the form of an August 6, 2020 interview with Yan Breuleux. I’ve excerpted one bit,

Finally, thinking about this year’s theme [Why Sentience?], there might be something a bit ironic about exploring the notion of sentience (historically reserved for biological life, and quite a small subsection of it) through digital media and electronic arts. There’s been much work done in the past 25 years to loosen the boundaries between such distinctions: how do you imagine ISEA2020 helping in that?

The similarities shared between humans, animals, and machines are fundamental in cybernetic sciences. According to the founder of cybernetics Norbert Wiener, the main tenets of the information paradigm – the notion of feedback – can be applied to humans, animals as well as the material world. Famously, the AA predictor (as analysed by Peter Galison in 1994) can be read as a first attempt at human-machine fusion (otherwise known as a cyborg).

The infamous Turing test also tends to blur the lines between humans and machines, between language and informational systems. Second-order cybernetics are often associated with biologists Francisco Varela and Humberto Maturana. The very notion of autopoiesis (a system capable of maintaining a certain level of stability in an unstable environment) relates back to the concept of homeostasis formulated by Willam Ross [William Ross Ashby] in 1952. Moreover, the concept of “ecosystems” emanates directly from the field of second-order cybernetics, providing researchers with a clearer picture of the interdependencies between living and non-living organisms. In light of these theories, the absence of boundaries between animals, humans, and machines constitutes the foundation of the technosciences paradigm. New media, technological arts, virtual arts, etc., partake in the dialogue between humans and machines, and thus contribute to the prolongation of this paradigm. Frank Popper nearly called his book “Techno Art” instead of “Virtual Art”, in reference to technosciences (his editor suggested the name change). For artists in the technological arts community, Jakob von Uexkull’s notion of “human-animal milieu” is an essential reference. Also present in Simondon’s reflections on human environments (both natural and artificial), the notion of “milieu” is quite important in the discourses about art and the environment. Concordia University’s artistic community chose the concept of “milieu” as the rallying point of its research laboratories.

ISEA2020’s theme resonates particularly well with the recent eruption of processing and artificial intelligence technologies. For me, Sentience is a purely human and animal idea: machines can only simulate our ways of thinking and feeling. Partly in an effort to explore the illusion of sentience in computers, Louis-Philippe Rondeau, Benoît Melançon and I have established the Mimesis laboratory at NAD University. Processing and AI technologies are especially useful in the creation of “digital doubles”, “Vactors”, real-time avatar generation, Deep Fakes and new forms of personalised interactions.

I adhere to the epistemological position that the living world is immeasurable. Through their ability to simulate, machines can merely reduce complex logics to a point of understandability. The utopian notion of empathetic computers is an idea mostly explored by popular science-fiction movies. Nonetheless, research into computer sentience allows us to devise possible applications, explore notions of embodiment and agency, and thereby develop new forms of interaction. Beyond my own point of view, the idea that machines can somehow feel emotions gives artists and researchers the opportunity to experiment with certain findings from the fields of the cognitive sciences, computer sciences and interactive design. For example, in 2002 I was particularly marked by an immersive installation at Universal Exhibition in Neuchatel, Switzerland titled Ada: Intelligence Space. The installation comprised an artificial environment controlled by a computer, which interacted with the audience on the basis of artificial emotion. The system encouraged visitors to participate by intelligently analysing their movements and sounds. Another example, Louis-Philippe Demers’ Blind Robot (2012),  demonstrates how artists can be both critical of, and amazed by, these new forms of knowledge. Additionally, the 2016 BIAN (Biennale internationale d’art numérique), organized by ELEKTRA (Alain Thibault) explored the various ways these concepts were appropriated in installation and interactive art. The way I see it, current works of digital art operate as boundary objects. The varied usages and interpretations of a particular work of art allow it to be analyzed from nearly every angle or field of study. Thus, philosophers can ask themselves: how does a computer come to understand what being human really is?

I have yet to attend conferences or exchange with researchers on that subject. Although the sheer number of presentation propositions sent to ISEA2020, I have no doubt that the symposium will be the ideal context to reflect on the concept of Sentience and many issues raised therein.

For the last bit of news.

HotPopRobot, one of six global winners of 2020 NASA SpaceApps COVID-19 challenge

I last wrote about HotPopRobot’s (Artash and Arushi with a little support from their parents) response to the 2020 NASA (US National Aeronautics and Space Administration) SpaceApps challenge in my July 1, 2020 post, Toronto COVID-19 Lockdown Musical: a data sonification project from HotPopRobot. (You’ll find a video of the project embedded in the post.)

Here’s more news from HotPopRobot’s August 4, 2020 posting (Note: Links have been removed),

Artash (14 years) and Arushi (10 years). Toronto.

We are excited to become the global winners of the 2020 NASA SpaceApps COVID-19 Challenge from among 2,000 teams from 150 countries. The six Global Winners will be invited to visit a NASA Rocket Launch site to view a spacecraft launch along with the SpaceApps Organizing team once travel is deemed safe. They will also receive an invitation to present their projects to NASA, ESA [European Space Agency], JAXA [Japan Aerospace Exploration Agency], CNES [Centre National D’Etudes Spatiales; France], and CSA [Canadian Space Agency] personnel. https://covid19.spaceappschallenge.org/awards

15,000 participants joined together to submit over 1400 projects for the COVID-19 Global Challenge that was held on 30-31 May 2020. 40 teams made to the Global Finalists. Amongst them, 6 teams became the global winners!

The 2020 SpaceApps was an international collaboration between NASA, Canadian Space Agency, ESA, JAXA, CSA,[sic] and CNES focused on solving global challenges. During a period of 48 hours, participants from around the world were required to create virtual teams and solve any of the 12 challenges related to the COVID-19 pandemic posted on the SpaceApps website. More details about the 2020 SpaceApps COVID-19 Challenge:  https://sa-2019.s3.amazonaws.com/media/documents/Space_Apps_FAQ_COVID_.pdf

We have been participating in NASA Space Challenge for the last seven years since 2014. We were only 8 years and 5 years respectively when we participated in our very first SpaceApps 2014.

We have grown up learning more about space, tacking global challenges, making hardware and software projects, participating in meetings, networking with mentors and teams across the globe, and giving presentations through the annual NASA Space Apps Challenges. This is one challenge we look forward to every year.

It has been a fun and exciting journey meeting so many people and astronauts and visiting several fascinating places on the way! We hope more kids, youths, and families are inspired by our space journey. Space is for all and is yours to discover!

If you have the time, I recommend reading HotPopRobot’s August 4, 2020 posting in its entirety.

China’s neuromorphic chips: Darwin and Tianjic

I believe that China has more than two neuromorphic chips. The two being featured here are the ones for which I was easily able to find information.

The Darwin chip

The first information (that I stumbled across) about China and a neuromorphic chip (Darwin) was in a December 22, 2015 Science China Press news release on EurekAlert,

Artificial Neural Network (ANN) is a type of information processing system based on mimicking the principles of biological brains, and has been broadly applied in application domains such as pattern recognition, automatic control, signal processing, decision support system and artificial intelligence. Spiking Neural Network (SNN) is a type of biologically-inspired ANN that perform information processing based on discrete-time spikes. It is more biologically realistic than classic ANNs, and can potentially achieve much better performance-power ratio. Recently, researchers from Zhejiang University and Hangzhou Dianzi University in Hangzhou, China successfully developed the Darwin Neural Processing Unit (NPU), a neuromorphic hardware co-processor based on Spiking Neural Networks, fabricated by standard CMOS technology.

With the rapid development of the Internet-of-Things and intelligent hardware systems, a variety of intelligent devices are pervasive in today’s society, providing many services and convenience to people’s lives, but they also raise challenges of running complex intelligent algorithms on small devices. Sponsored by the college of Computer science of Zhejiang University, the research group led by Dr. De Ma from Hangzhou Dianzi university and Dr. Xiaolei Zhu from Zhejiang university has developed a co-processor named as Darwin.The Darwin NPU aims to provide hardware acceleration of intelligent algorithms, with target application domain of resource-constrained, low-power small embeddeddevices. It has been fabricated by 180nm standard CMOS process, supporting a maximum of 2048 neurons, more than 4 million synapses and 15 different possible synaptic delays. It is highly configurable, supporting reconfiguration of SNN topology and many parameters of neurons and synapses.Figure 1 shows photos of the die and the prototype development board, which supports input/output in the form of neural spike trains via USB port.

The successful development ofDarwin demonstrates the feasibility of real-time execution of Spiking Neural Networks in resource-constrained embedded systems. It supports flexible configuration of a multitude of parameters of the neural network, hence it can be used to implement different functionalities as configured by the user. Its potential applications include intelligent hardware systems, robotics, brain-computer interfaces, and others.Since it uses spikes for information processing and transmission,similar to biological neural networks, it may be suitable for analysis and processing of biological spiking neural signals, and building brain-computer interface systems by interfacing with animal or human brains. As a prototype application in Brain-Computer Interfaces, Figure 2 [not included here] describes an application example ofrecognizingthe user’s motor imagery intention via real-time decoding of EEG signals, i.e., whether he is thinking of left or right, and using it to control the movement direction of a basketball in the virtual environment. Different from conventional EEG signal analysis algorithms, the input and output to Darwin are both neural spikes: the input is spike trains that encode EEG signals; after processing by the neural network, the output neuron with the highest firing rate is chosen as the classification result.

The most recent development for this chip was announced in a September 2, 2019 Zhejiang University press release (Note: Links have been removed),

The second generation of the Darwin Neural Processing Unit (Darwin NPU 2) as well as its corresponding toolchain and micro-operating system was released in Hangzhou recently. This research was led by Zhejiang University, with Hangzhou Dianzi University and Huawei Central Research Institute participating in the development and algorisms of the chip. The Darwin NPU 2 can be primarily applied to smart Internet of Things (IoT). It can support up to 150,000 neurons and has achieved the largest-scale neurons on a nationwide basis.

The Darwin NPU 2 is fabricated by standard 55nm CMOS technology. Every “neuromorphic” chip is made up of 576 kernels, each of which can support 256 neurons. It contains over 10 million synapses which can construct a powerful brain-inspired computing system.

“A brain-inspired chip can work like the neurons inside a human brain and it is remarkably unique in image recognition, visual and audio comprehension and naturalistic language processing,” said MA De, an associate professor at the College of Computer Science and Technology on the research team.

“In comparison with traditional chips, brain-inspired chips are more adept at processing ambiguous data, say, perception tasks. Another prominent advantage is their low energy consumption. In the process of information transmission, only those neurons that receive and process spikes will be activated while other neurons will stay dormant. In this case, energy consumption can be extremely low,” said Dr. ZHU Xiaolei at the School of Microelectronics.

To cater to the demands for voice business, Huawei Central Research Institute designed an efficient spiking neural network algorithm in accordance with the defining feature of the Darwin NPU 2 architecture, thereby increasing computing speeds and improving recognition accuracy tremendously.

Scientists have developed a host of applications, including gesture recognition, image recognition, voice recognition and decoding of electroencephalogram (EEG) signals, on the Darwin NPU 2 and reduced energy consumption by at least two orders of magnitude.

In comparison with the first generation of the Darwin NPU which was developed in 2015, the Darwin NPU 2 has escalated the number of neurons by two orders of magnitude from 2048 neurons and augmented the flexibility and plasticity of the chip configuration, thus expanding the potential for applications appreciably. The improvement in the brain-inspired chip will bring in its wake the revolution of computer technology and artificial intelligence. At present, the brain-inspired chip adopts a relatively simplified neuron model, but neurons in a real brain are far more sophisticated and many biological mechanisms have yet to be explored by neuroscientists and biologists. It is expected that in the not-too-distant future, a fascinating improvement on the Darwin NPU 2 will come over the horizon.

I haven’t been able to find a recent (i.e., post 2017) research paper featuring Darwin but there is another chip and research on that one was published in July 2019. First, the news.

The Tianjic chip

A July 31, 2019 article in the New York Times by Cade Metz describes the research and offers what seems to be a jaundiced perspective about the field of neuromorphic computing (Note: A link has been removed),

As corporate giants like Ford, G.M. and Waymo struggle to get their self-driving cars on the road, a team of researchers in China is rethinking autonomous transportation using a souped-up bicycle.

This bike can roll over a bump on its own, staying perfectly upright. When the man walking just behind it says “left,” it turns left, angling back in the direction it came.

It also has eyes: It can follow someone jogging several yards ahead, turning each time the person turns. And if it encounters an obstacle, it can swerve to the side, keeping its balance and continuing its pursuit.

… Chinese researchers who built the bike believe it demonstrates the future of computer hardware. It navigates the world with help from what is called a neuromorphic chip, modeled after the human brain.

Here’s a video, released by the researchers, demonstrating the chip’s abilities,

Now back to back to Metz’s July 31, 2019 article (Note: A link has been removed),

The short video did not show the limitations of the bicycle (which presumably tips over occasionally), and even the researchers who built the bike admitted in an email to The Times that the skills on display could be duplicated with existing computer hardware. But in handling all these skills with a neuromorphic processor, the project highlighted the wider effort to achieve new levels of artificial intelligence with novel kinds of chips.

This effort spans myriad start-up companies and academic labs, as well as big-name tech companies like Google, Intel and IBM. And as the Nature paper demonstrates, the movement is gaining significant momentum in China, a country with little experience designing its own computer processors, but which has invested heavily in the idea of an “A.I. chip.”

If you can get past what seems to be a patronizing attitude, there are some good explanations and cogent criticisms in the piece (Metz’s July 31, 2019 article, Note: Links have been removed),

… it faces significant limitations.

A neural network doesn’t really learn on the fly. Engineers train a neural network for a particular task before sending it out into the real world, and it can’t learn without enormous numbers of examples. OpenAI, a San Francisco artificial intelligence lab, recently built a system that could beat the world’s best players at a complex video game called Dota 2. But the system first spent months playing the game against itself, burning through millions of dollars in computing power.

Researchers aim to build systems that can learn skills in a manner similar to the way people do. And that could require new kinds of computer hardware. Dozens of companies and academic labs are now developing chips specifically for training and operating A.I. systems. The most ambitious projects are the neuromorphic processors, including the Tianjic chip under development at Tsinghua University in China.

Such chips are designed to imitate the network of neurons in the brain, not unlike a neural network but with even greater fidelity, at least in theory.

Neuromorphic chips typically include hundreds of thousands of faux neurons, and rather than just processing 1s and 0s, these neurons operate by trading tiny bursts of electrical signals, “firing” or “spiking” only when input signals reach critical thresholds, as biological neurons do.

Tiernan Ray’s August 3, 2019 article about the chip for ZDNet.com offers some thoughtful criticism with a side dish of snark (Note: Links have been removed),

Nature magazine’s cover story [July 31, 2019] is about a Chinese chip [Tianjic chip]that can run traditional deep learning code and also perform “neuromorophic” operations in the same circuitry. The work’s value seems obscured by a lot of hype about “artificial general intelligence” that has no real justification.

The term “artificial general intelligence,” or AGI, doesn’t actually refer to anything, at this point, it is merely a placeholder, a kind of Rorschach Test for people to fill the void with whatever notions they have of what it would mean for a machine to “think” like a person.

Despite that fact, or perhaps because of it, AGI is an ideal marketing term to attach to a lot of efforts in machine learning. Case in point, a research paper featured on the cover of this week’s Nature magazine about a new kind of computer chip developed by researchers at China’s Tsinghua University that could “accelerate the development of AGI,” they claim.

The chip is a strange hybrid of approaches, and is intriguing, but the work leaves unanswered many questions about how it’s made, and how it achieves what researchers claim of it. And some longtime chip observers doubt the impact will be as great as suggested.

“This paper is an example of the good work that China is doing in AI,” says Linley Gwennap, longtime chip-industry observer and principal analyst with chip analysis firm The Linley Group. “But this particular idea isn’t going to take over the world.”

The premise of the paper, “Towards artificial general intelligence with hybrid Tianjic chip architecture,” is that to achieve AGI, computer chips need to change. That’s an idea supported by fervent activity these days in the land of computer chips, with lots of new chip designs being proposed specifically for machine learning.

The Tsinghua authors specifically propose that the mainstream machine learning of today needs to be merged in the same chip with what’s called “neuromorphic computing.” Neuromorphic computing, first conceived by Caltech professor Carver Mead in the early ’80s, has been an obsession for firms including IBM for years, with little practical result.

[Missing details about the chip] … For example, the part is said to have “reconfigurable” circuits, but how the circuits are to be reconfigured is never specified. It could be so-called “field programmable gate array,” or FPGA, technology or something else. Code for the project is not provided by the authors as it often is for such research; the authors offer to provide the code “on reasonable request.”

More important is the fact the chip may have a hard time stacking up to a lot of competing chips out there, says analyst Gwennap. …

What the paper calls ANN and SNN are two very different means of solving similar problems, kind of like rotating (helicopter) and fixed wing (airplane) are for aviation,” says Gwennap. “Ultimately, I expect ANN [?] and SNN [spiking neural network] to serve different end applications, but I don’t see a need to combine them in a single chip; you just end up with a chip that is OK for two things but not great for anything.”

But you also end up generating a lot of buzz, and given the tension between the U.S. and China over all things tech, and especially A.I., the notion China is stealing a march on the U.S. in artificial general intelligence — whatever that may be — is a summer sizzler of a headline.

ANN could be either artificial neural network or something mentioned earlier in Ray’s article, a shortened version of CANN [continuous attractor neural network].

Shelly Fan’s August 7, 2019 article for the SingularityHub is almost as enthusiastic about the work as the podcasters for Nature magazine  were (a little more about that later),

The study shows that China is readily nipping at the heels of Google, Facebook, NVIDIA, and other tech behemoths investing in developing new AI chip designs—hell, with billions in government investment it may have already had a head start. A sweeping AI plan from 2017 looks to catch up with the US on AI technology and application by 2020. By 2030, China’s aiming to be the global leader—and a champion for building general AI that matches humans in intellectual competence.

The country’s ambition is reflected in the team’s parting words.

“Our study is expected to stimulate AGI [artificial general intelligence] development by paving the way to more generalized hardware platforms,” said the authors, led by Dr. Luping Shi at Tsinghua University.

Using nanoscale fabrication, the team arranged 156 FCores, containing roughly 40,000 neurons and 10 million synapses, onto a chip less than a fifth of an inch in length and width. Initial tests showcased the chip’s versatility, in that it can run both SNNs and deep learning algorithms such as the popular convolutional neural network (CNNs) often used in machine vision.

Compared to IBM TrueNorth, the density of Tianjic’s cores increased by 20 percent, speeding up performance ten times and increasing bandwidth at least 100-fold, the team said. When pitted against GPUs, the current hardware darling of machine learning, the chip increased processing throughput up to 100 times, while using just a sliver (1/10,000) of energy.

BTW, Fan is a neuroscientist (from her SingularityHub profile page),

Shelly Xuelai Fan is a neuroscientist-turned-science writer. She completed her PhD in neuroscience at the University of British Columbia, where she developed novel treatments for neurodegeneration. While studying biological brains, she became fascinated with AI and all things biotech. Following graduation, she moved to UCSF [University of California at San Francisco] to study blood-based factors that rejuvenate aged brains. She is the co-founder of Vantastic Media, a media venture that explores science stories through text and video, and runs the award-winning blog NeuroFantastic.com. Her first book, “Will AI Replace Us?” (Thames & Hudson) will be out April 2019.

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

Towards artificial general intelligence with hybrid Tianjic chip architecture by Jing Pei, Lei Deng, Sen Song, Mingguo Zhao, Youhui Zhang, Shuang Wu, Guanrui Wang, Zhe Zou, Zhenzhi Wu, Wei He, Feng Chen, Ning Deng, Si Wu, Yu Wang, Yujie Wu, Zheyu Yang, Cheng Ma, Guoqi Li, Wentao Han, Huanglong Li, Huaqiang Wu, Rong Zhao, Yuan Xie & Luping Shi. Nature volume 572, pages106–111(2019) DOI: https//doi.org/10.1038/s41586-019-1424-8 Published: 31 July 2019 Issue Date: 01 August 2019

This paper is behind a paywall.

The July 31, 2019 Nature podcast, which includes a segment about the Tianjic chip research from China, which is at the 9 mins. 13 secs. mark (AI hardware) or you can scroll down about 55% of the way to the transcript of the interview with Luke Fleet, the Nature editor who dealt with the paper.

Some thoughts

The pundits put me in mind of my own reaction when I heard about phones that could take pictures. I didn’t see the point but, as it turned out, there was a perfectly good reason for combining what had been two separate activities into one device. It was no longer just a telephone and I had completely missed the point.

This too may be the case with the Tianjic chip. I think it’s too early to say whether or not it represents a new type of chip or if it’s a dead end.

A tangle of silver nanowires for brain-like action

I’ve been meaning to get to this news item from late 2019 as it features work from a team that I’ve been following for a number of years now. First mentioned here in an October 17, 2011 posting, James Gimzewski has been working with researchers at the University of California at Los Angeles (UCLA) and researchers at Japan’s National Institute for Materials Science (NIMS) on neuromorphic computing.

This particular research had a protracted rollout with the paper being published in October 2019 and the last news item about it being published in mid-December 2019.

A December 17, 2029 news item on Nanowerk was the first to alert me to this new work (Note: A link has been removed),

UCLA scientists James Gimzewski and Adam Stieg are part of an international research team that has taken a significant stride toward the goal of creating thinking machines.

Led by researchers at Japan’s National Institute for Materials Science, the team created an experimental device that exhibited characteristics analogous to certain behaviors of the brain — learning, memorization, forgetting, wakefulness and sleep. The paper, published in Scientific Reports (“Emergent dynamics of neuromorphic nanowire networks”), describes a network in a state of continuous flux.

A December 16, 2019 UCLA news release, which originated the news item, offers more detail (Note: A link has been removed),

“This is a system between order and chaos, on the edge of chaos,” said Gimzewski, a UCLA distinguished professor of chemistry and biochemistry, a member of the California NanoSystems Institute at UCLA and a co-author of the study. “The way that the device constantly evolves and shifts mimics the human brain. It can come up with different types of behavior patterns that don’t repeat themselves.”

The research is one early step along a path that could eventually lead to computers that physically and functionally resemble the brain — machines that may be capable of solving problems that contemporary computers struggle with, and that may require much less power than today’s computers do.

The device the researchers studied is made of a tangle of silver nanowires — with an average diameter of just 360 nanometers. (A nanometer is one-billionth of a meter.) The nanowires were coated in an insulating polymer about 1 nanometer thick. Overall, the device itself measured about 10 square millimeters — so small that it would take 25 of them to cover a dime.

Allowed to randomly self-assemble on a silicon wafer, the nanowires formed highly interconnected structures that are remarkably similar to those that form the neocortex, the part of the brain involved with higher functions such as language, perception and cognition.

One trait that differentiates the nanowire network from conventional electronic circuits is that electrons flowing through them cause the physical configuration of the network to change. In the study, electrical current caused silver atoms to migrate from within the polymer coating and form connections where two nanowires overlap. The system had about 10 million of these junctions, which are analogous to the synapses where brain cells connect and communicate.

The researchers attached two electrodes to the brain-like mesh to profile how the network performed. They observed “emergent behavior,” meaning that the network displayed characteristics as a whole that could not be attributed to the individual parts that make it up. This is another trait that makes the network resemble the brain and sets it apart from conventional computers.

After current flowed through the network, the connections between nanowires persisted for as much as one minute in some cases, which resembled the process of learning and memorization in the brain. Other times, the connections shut down abruptly after the charge ended, mimicking the brain’s process of forgetting.

In other experiments, the research team found that with less power flowing in, the device exhibited behavior that corresponds to what neuroscientists see when they use functional MRI scanning to take images of the brain of a sleeping person. With more power, the nanowire network’s behavior corresponded to that of the wakeful brain.

The paper is the latest in a series of publications examining nanowire networks as a brain-inspired system, an area of research that Gimzewski helped pioneer along with Stieg, a UCLA research scientist and an associate director of CNSI.

“Our approach may be useful for generating new types of hardware that are both energy-efficient and capable of processing complex datasets that challenge the limits of modern computers,” said Stieg, a co-author of the study.

The borderline-chaotic activity of the nanowire network resembles not only signaling within the brain but also other natural systems such as weather patterns. That could mean that, with further development, future versions of the device could help model such complex systems.

In other experiments, Gimzewski and Stieg already have coaxed a silver nanowire device to successfully predict statistical trends in Los Angeles traffic patterns based on previous years’ traffic data.

Because of their similarities to the inner workings of the brain, future devices based on nanowire technology could also demonstrate energy efficiency like the brain’s own processing. The human brain operates on power roughly equivalent to what’s used by a 20-watt incandescent bulb. By contrast, computer servers where work-intensive tasks take place — from training for machine learning to executing internet searches — can use the equivalent of many households’ worth of energy, with the attendant carbon footprint.

“In our studies, we have a broader mission than just reprogramming existing computers,” Gimzewski said. “Our vision is a system that will eventually be able to handle tasks that are closer to the way the human being operates.”

The study’s first author, Adrian Diaz-Alvarez, is from the International Center for Material Nanoarchitectonics at Japan’s National Institute for Materials Science. Co-authors include Tomonobu Nakayama and Rintaro Higuchi, also of NIMS; and Zdenka Kuncic at the University of Sydney in Australia.

Caption: (a) Micrograph of the neuromorphic network fabricated by this research team. The network contains of numerous junctions between nanowires, which operate as synaptic elements. When voltage is applied to the network (between the green probes), current pathways (orange) are formed in the network. (b) A Human brain and one of its neuronal networks. The brain is known to have a complex network structure and to operate by means of electrical signal propagation across the network. Credit: NIMS

A November 11, 2019 National Institute for Materials Science (Japan) press release (also on EurekAlert but dated December 25, 2019) first announced the news,

An international joint research team led by NIMS succeeded in fabricating a neuromorphic network composed of numerous metallic nanowires. Using this network, the team was able to generate electrical characteristics similar to those associated with higher order brain functions unique to humans, such as memorization, learning, forgetting, becoming alert and returning to calm. The team then clarified the mechanisms that induced these electrical characteristics.

The development of artificial intelligence (AI) techniques has been rapidly advancing in recent years and has begun impacting our lives in various ways. Although AI processes information in a manner similar to the human brain, the mechanisms by which human brains operate are still largely unknown. Fundamental brain components, such as neurons and the junctions between them (synapses), have been studied in detail. However, many questions concerning the brain as a collective whole need to be answered. For example, we still do not fully understand how the brain performs such functions as memorization, learning and forgetting, and how the brain becomes alert and returns to calm. In addition, live brains are difficult to manipulate in experimental research. For these reasons, the brain remains a “mysterious organ.” A different approach to brain research?in which materials and systems capable of performing brain-like functions are created and their mechanisms are investigated?may be effective in identifying new applications of brain-like information processing and advancing brain science.

The joint research team recently built a complex brain-like network by integrating numerous silver (Ag) nanowires coated with a polymer (PVP) insulating layer approximately 1 nanometer in thickness. A junction between two nanowires forms a variable resistive element (i.e., a synaptic element) that behaves like a neuronal synapse. This nanowire network, which contains a large number of intricately interacting synaptic elements, forms a “neuromorphic network”. When a voltage was applied to the neuromorphic network, it appeared to “struggle” to find optimal current pathways (i.e., the most electrically efficient pathways). The research team measured the processes of current pathway formation, retention and deactivation while electric current was flowing through the network and found that these processes always fluctuate as they progress, similar to the human brain’s memorization, learning, and forgetting processes. The observed temporal fluctuations also resemble the processes by which the brain becomes alert or returns to calm. Brain-like functions simulated by the neuromorphic network were found to occur as the huge number of synaptic elements in the network collectively work to optimize current transport, in the other words, as a result of self-organized and emerging dynamic processes..

The research team is currently developing a brain-like memory device using the neuromorphic network material. The team intends to design the memory device to operate using fundamentally different principles than those used in current computers. For example, while computers are currently designed to spend as much time and electricity as necessary in pursuit of absolutely optimum solutions, the new memory device is intended to make a quick decision within particular limits even though the solution generated may not be absolutely optimum. The team also hopes that this research will facilitate understanding of the brain’s information processing mechanisms.

This project was carried out by an international joint research team led by Tomonobu Nakayama (Deputy Director, International Center for Materials Nanoarchitectonics (WPI-MANA), NIMS), Adrian Diaz Alvarez (Postdoctoral Researcher, WPI-MANA, NIMS), Zdenka Kuncic (Professor, School of Physics, University of Sydney, Australia) and James K. Gimzewski (Professor, California NanoSystems Institute, University of California Los Angeles, USA).

Here at last is a link to and a citation for the paper,

Emergent dynamics of neuromorphic nanowire networks by Adrian Diaz-Alvarez, Rintaro Higuchi, Paula Sanz-Leon, Ido Marcus, Yoshitaka Shingaya, Adam Z. Stieg, James K. Gimzewski, Zdenka Kuncic & Tomonobu Nakayama. Scientific Reports volume 9, Article number: 14920 (2019) DOI: https://doi.org/10.1038/s41598-019-51330-6 Published: 17 October 2019

This paper is open access.

Comedy club performances show how robots and humans connect via humor

Caption: Naomi Fitter and Jon the Robot. Credit: Johanna Carson, OSU College of Engineering

Robot comedian is not my first thought on seeing that image; ventriloquist’s dummy is what came to mind. However, it’s not the first time I’ve been wrong about something. A May 19, 2020 news item on ScienceDaily reveals the truth about Jon, a comedian in robot form,

Standup comedian Jon the Robot likes to tell his audiences that he does lots of auditions but has a hard time getting bookings.

“They always think I’m too robotic,” he deadpans.

If raucous laughter follows, he comes back with, “Please tell the booking agents how funny that joke was.”

If it doesn’t, he follows up with, “Sorry about that. I think I got caught in a loop. Please tell the booking agents that you like me … that you like me … that you like me … that you like me.”

Jon the Robot, with assistance from Oregon State University researcher Naomi Fitter, recently wrapped up a 32-show tour of comedy clubs in greater Los Angeles and in Oregon, generating guffaws and more importantly data that scientists and engineers can use to help robots and people relate more effectively with one another via humor.

A May 18, 2020 Oregon State University (OSU) news release (also on EurekAlert), which originated the news item, delves furthers into this intriguing research,

“Social robots and autonomous social agents are becoming more and more ingrained in our everyday lives,” said Fitter, assistant professor of robotics in the OSU College of Engineering. “Lots of them tell jokes to engage users – most people understand that humor, especially nuanced humor, is essential to relationship building. But it’s challenging to develop entertaining jokes for robots that are funny beyond the novelty level.”

Live comedy performances are a way for robots to learn “in the wild” which jokes and which deliveries work and which ones don’t, Fitter said, just like human comedians do.

Two studies comprised the comedy tour, which included assistance from a team of Southern California comedians in coming up with material true to, and appropriate for, a robot comedian.

The first study, consisting of 22 performances in the Los Angeles area, demonstrated that audiences found a robot comic with good timing – giving the audience the right amounts of time to react, etc. – to be significantly more funny than one without good timing.

The second study, based on 10 routines in Oregon, determined that an “adaptive performance” – delivering post-joke “tags” that acknowledge an audience’s reaction to the joke – wasn’t necessarily funnier overall, but the adaptations almost always improved the audience’s perception of individual jokes. In the second study, all performances featured appropriate timing.

“In bad-timing mode, the robot always waited a full five seconds after each joke, regardless of audience response,” Fitter said. “In appropriate-timing mode, the robot used timing strategies to pause for laughter and continue when it subsided, just like an effective human comedian would. Overall, joke response ratings were higher when the jokes were delivered with appropriate timing.”

The number of performances, given to audiences of 10 to 20, provide enough data to identify significant differences between distinct modes of robot comedy performance, and the research helped to answer key questions about comedic social interaction, Fitter said.

“Audience size, social context, cultural context, the microphone-holding human presence and the novelty of a robot comedian may have influenced crowd responses,” Fitter said. “The current software does not account for differences in laughter profiles, but future work can account for these differences using a baseline response measurement. The only sensing we used to evaluate joke success was audio readings. Future work might benefit from incorporating additional types of sensing.”

Still, the studies have key implications for artificial intelligence efforts to understand group responses to dynamic, entertaining social robots in real-world environments, she said.

“Also, possible advances in comedy from this work could include improved techniques for isolating and studying the effects of comedic techniques and better strategies to help comedians assess the success of a joke or routine,” she said. “The findings will guide our next steps toward giving autonomous social agents improved humor capabilities.”

The studies were published by the Association for Computing Machinery [ACM]/Institute of Electrical and Electronics Engineering’s [IEEE] International Conference on Human-Robot Interaction [HRI].

Here’s another link to the two studies published in a single paper, which were first presented at the 2020 International Conference on Human-Robot Interaction [HRI]. along with a citation for the title of the published presentation,

Comedians in Cafes Getting Data: Evaluating Timing and Adaptivity in Real-World Robot Comedy Performance by John Vilk and Naomi T Fitter. HRI ’20: Proceedings of the 2020 ACM/IEEE International Conference on Human-Robot InteractionMarch 2020 Pages 223–231 DOI: https://doi.org/10.1145/3319502.3374780

The paper is open access and the researchers have embedded an mp4 file which includes parts of the performances. Enjoy!

Artificial intelligence (AI) consumes a lot of energy but tree-like memory may help conserve it

A simulation of a quantum material’s properties reveals its ability to learn numbers, a test of artificial intelligence. (Purdue University image/Shakti Wadekar)

A May 7, 2020 Purdue University news release (also on EurekAlert) describes a new approach for energy-efficient hardware in support of artificial intelligence (AI) systems,

To just solve a puzzle or play a game, artificial intelligence can require software running on thousands of computers. That could be the energy that three nuclear plants produce in one hour.

A team of engineers has created hardware that can learn skills using a type of AI that currently runs on software platforms. Sharing intelligence features between hardware and software would offset the energy needed for using AI in more advanced applications such as self-driving cars or discovering drugs.

“Software is taking on most of the challenges in AI. If you could incorporate intelligence into the circuit components in addition to what is happening in software, you could do things that simply cannot be done today,” said Shriram Ramanathan, a professor of materials engineering at Purdue University.

AI hardware development is still in early research stages. Researchers have demonstrated AI in pieces of potential hardware, but haven’t yet addressed AI’s large energy demand.

As AI penetrates more of daily life, a heavy reliance on software with massive energy needs is not sustainable, Ramanathan said. If hardware and software could share intelligence features, an area of silicon might be able to achieve more with a given input of energy.

Ramanathan’s team is the first to demonstrate artificial “tree-like” memory in a piece of potential hardware at room temperature. Researchers in the past have only been able to observe this kind of memory in hardware at temperatures that are too low for electronic devices.

The results of this study are published in the journal Nature Communications.

The hardware that Ramanathan’s team developed is made of a so-called quantum material. These materials are known for having properties that cannot be explained by classical physics. Ramanathan’s lab has been working to better understand these materials and how they might be used to solve problems in electronics.

Software uses tree-like memory to organize information into various “branches,” making that information easier to retrieve when learning new skills or tasks.

The strategy is inspired by how the human brain categorizes information and makes decisions.

“Humans memorize things in a tree structure of categories. We memorize ‘apple’ under the category of ‘fruit’ and ‘elephant’ under the category of ‘animal,’ for example,” said Hai-Tian Zhang, a Lillian Gilbreth postdoctoral fellow in Purdue’s College of Engineering. “Mimicking these features in hardware is potentially interesting for brain-inspired computing.”

The team introduced a proton to a quantum material called neodymium nickel oxide. They discovered that applying an electric pulse to the material moves around the proton. Each new position of the proton creates a different resistance state, which creates an information storage site called a memory state. Multiple electric pulses create a branch made up of memory states.

“We can build up many thousands of memory states in the material by taking advantage of quantum mechanical effects. The material stays the same. We are simply shuffling around protons,” Ramanathan said.

Through simulations of the properties discovered in this material, the team showed that the material is capable of learning the numbers 0 through 9. The ability to learn numbers is a baseline test of artificial intelligence.

The demonstration of these trees at room temperature in a material is a step toward showing that hardware could offload tasks from software.

“This discovery opens up new frontiers for AI that have been largely ignored because implementing this kind of intelligence into electronic hardware didn’t exist,” Ramanathan said.

The material might also help create a way for humans to more naturally communicate with AI.

“Protons also are natural information transporters in human beings. A device enabled by proton transport may be a key component for eventually achieving direct communication with organisms, such as through a brain implant,” Zhang said.

Here’s a link to and a citation for the published study,

Perovskite neural trees by Hai-Tian Zhang, Tae Joon Park, Shriram Ramanathan. Nature Communications volume 11, Article number: 2245 (2020) DOI: https://doi.org/10.1038/s41467-020-16105-y Published: 07 May 2020

This paper is open access.

Brain-inspired electronics with organic memristors for wearable computing

I went down a rabbit hole while trying to figure out the difference between ‘organic’ memristors and standard memristors. I have put the results of my investigation at the end of this post. First, there’s the news.

An April 21, 2020 news item on ScienceDaily explains why researchers are so focused on memristors and brainlike computing,

The advent of artificial intelligence, machine learning and the internet of things is expected to change modern electronics and bring forth the fourth Industrial Revolution. The pressing question for many researchers is how to handle this technological revolution.

“It is important for us to understand that the computing platforms of today will not be able to sustain at-scale implementations of AI algorithms on massive datasets,” said Thirumalai Venkatesan, one of the authors of a paper published in Applied Physics Reviews, from AIP Publishing.

“Today’s computing is way too energy-intensive to handle big data. We need to rethink our approaches to computation on all levels: materials, devices and architecture that can enable ultralow energy computing.”

An April 21, 2020 American Institute of Physics (AIP) news release (also on EurekAlert), which originated the news item, describes the authors’ approach to the problems with organic memristors,

Brain-inspired electronics with organic memristors could offer a functionally promising and cost- effective platform, according to Venkatesan. Memristive devices are electronic devices with an inherent memory that are capable of both storing data and performing computation. Since memristors are functionally analogous to the operation of neurons, the computing units in the brain, they are optimal candidates for brain-inspired computing platforms.

Until now, oxides have been the leading candidate as the optimum material for memristors. Different material systems have been proposed but none have been successful so far.

“Over the last 20 years, there have been several attempts to come up with organic memristors, but none of those have shown any promise,” said Sreetosh Goswami, lead author on the paper. “The primary reason behind this failure is their lack of stability, reproducibility and ambiguity in mechanistic understanding. At a device level, we are now able to solve most of these problems,”

This new generation of organic memristors is developed based on metal azo complex devices, which are the brainchild of Sreebata Goswami, a professor at the Indian Association for the Cultivation of Science in Kolkata and another author on the paper.

“In thin films, the molecules are so robust and stable that these devices can eventually be the right choice for many wearable and implantable technologies or a body net, because these could be bendable and stretchable,” said Sreebata Goswami. A body net is a series of wireless sensors that stick to the skin and track health.

The next challenge will be to produce these organic memristors at scale, said Venkatesan.

“Now we are making individual devices in the laboratory. We need to make circuits for large-scale functional implementation of these devices.”

Caption: The device structure at a molecular level. The gold nanoparticles on the bottom electrode enhance the field enabling an ultra-low energy operation of the molecular device. Credit Sreetosh Goswami, Sreebrata Goswami and Thirumalai Venky Venkatesan

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

An organic approach to low energy memory and brain inspired electronics by Sreetosh Goswami, Sreebrata Goswami, and T. Venkatesan. Applied Physics Reviews 7, 021303 (2020) DOI: https://doi.org/10.1063/1.5124155

This paper is open access.

Basics about memristors and organic memristors

This undated article on Nanowerk provides a relatively complete and technical description of memristors in general (Note: A link has been removed),

A memristor (named as a portmanteau of memory and resistor) is a non-volatile electronic memory device that was first theorized by Leon Ong Chua in 1971 as the fourth fundamental two-terminal circuit element following the resistor, the capacitor, and the inductor (IEEE Transactions on Circuit Theory, “Memristor-The missing circuit element”).

Its special property is that its resistance can be programmed (resistor function) and subsequently remains stored (memory function). Unlike other memories that exist today in modern electronics, memristors are stable and remember their state even if the device loses power.

However, it was only almost 40 years later that the first practical device was fabricated. This was in 2008, when a group led by Stanley Williams at HP Research Labs realized that switching of the resistance between a conducting and less conducting state in metal-oxide thin-film devices was showing Leon Chua’s memristor behavior. …

The article on Nanowerk includes an embedded video presentation on memristors given by Stanley Williams (also known as R. Stanley Williams).

Mention of an ‘organic’memristor can be found in an October 31, 2017 article by Ryan Whitwam,

The memristor is composed of the transition metal ruthenium complexed with “azo-aromatic ligands.” [emphasis mine] The theoretical work enabling this material was performed at Yale, and the organic molecules were synthesized at the Indian Association for the Cultivation of Sciences. …

I highlighted ‘ligands’ because that appears to be the difference. However, there is more than one type of ligand on Wikipedia.

First, there’s the Ligand (biochemistry) entry (Note: Links have been removed),

In biochemistry and pharmacology, a ligand is a substance that forms a complex with a biomolecule to serve a biological purpose. …

Then, there’s the Ligand entry,

In coordination chemistry, a ligand[help 1] is an ion or molecule (functional group) that binds to a central metal atom to form a coordination complex …

Finally, there’s the Ligand (disambiguation) entry (Note: Links have been removed),

  • Ligand, an atom, ion, or functional group that donates one or more of its electrons through a coordinate covalent bond to one or more central atoms or ions
  • Ligand (biochemistry), a substance that binds to a protein
  • a ‘guest’ in host–guest chemistry

I did take a look at the paper and did not see any references to proteins or other biomolecules that I could recognize as such. I’m not sure why the researchers are describing their device as an ‘organic’ memristor but this may reflect a shortcoming in the definitions I have found or shortcomings in my reading of the paper rather than an error on their parts.

Hopefully, more research will be forthcoming and it will be possible to better understand the terminology.

Fourth Industrial Revolution and its impact on charity organizations

Andy Levy-Ajzenkopf’s February 21, 2020 article (Technology and innovation: How the Fourth Industrial Revolution is impacting the charitable sector) for Charity Village has an ebullient approach to adoption of new and emerging technologies in the charitable sector (Note: A link has been removed),

Almost daily, new technologies are being developed to help innovate the way people give or the way organizations offer opportunities to advance their causes. There is no going back.

The charitable sector – along with society at large – is now fully in the midst of what is being called the Fourth Industrial Revolution, a term first brought to prominence among CEOs, thought leaders and policy makers at the 2016 World Economic Forum. And if you haven’t heard the phrase yet, get ready to hear it tons more as economies around the world embrace it.

To be clear, the Fourth Industrial Revolution is the newest disruption in the way our world works. When you hear someone talk about it, what they’re describing is the massive technological shift in our business and personal ecosystems that now rely heavily on things like artificial intelligence, quantum computing, 3D printing and the general “Internet of things.”

Still, now more than ever, charitable business is getting done and being advanced by sector pioneers who aren’t afraid to make use of new technologies on offer to help civil society.

It seems like everywhere one turns, the topic of artificial intelligence (A.I.) is increasingly becoming subject of choice.

This is no different in the charitable sector, and particularly so for a new company called Fundraise Wisely (aka Wisely). Its co-founder and CEO, Artiom Komarov, explains a bit about what exactly his tech is doing for the sector.

“We help accelerate fundraising, with A.I. At a product level, we connect to your CRM (content relationship management system) and predict the next gift and next gift date for every donor. We then use that information to help you populate and prioritize donor portfolios,” Komarov states.

He notes that his company is seeing increased demand for innovative technologies from charities over the last while.

“What we’re hearing is that… A.I. tech is compelling because at the end of the day it’s meant to move the bottom line, helping nonprofits grow their revenue. We’ve also found that internally [at a charitable organization] there’s always a champion that sees the potential impact of technology; and that’s a great place to start with change,” Komarov says. “If it’s done right, tech can be an enabler of better work for organizations. From both research and experience, we know that tech adoption usually fails because of culture rather than the underlying technology. We’re here to work with the client closely to help that transition.”

I would like to have seen some numbers. For example, Komarov says that AI is having a positive impact on a charity’s bottom line. So, how much money did one of these charities raise? Was it more money than they would have made without AI? Assuming they did manage to raise greater funds, could another technology been more cost effective?

For another perspective (equally positive) on technology and charity, there’s a November 29, 2012 posting (Why technology and innovation are key to increasing charity donations) on the Guardian blogs by Henna Butt and Renita Shah (Note: Links have been removed),

At the beginning of this year the [UK] Cabinet Office and Nesta [formerly National Endowment for Science, Technology and the Arts {NESTA}] announced a £10m fund to invest in innovation in giving. The first tranche of this money has already been invested in promising initiatives such as Timto which allows you to create a gift list that includes a charity donation and Pennies, whose electronic money box allows customers to donate when paying for something in a shop using a credit card. Small and sizeable organisations alike are now using web and mobile technologies to make giving more convenient, more social and more compelling.

Butt’s and Shah’s focus was on mobile technologies and social networks. Like Levy-Ajzenkopf’s article, there’s no discussion of any possible downside to these technologies, e.g., privacy issues. As well, the inevitability of this move toward more technology for charity is explicitly stated by Levy-Ajzenkopf “There is no going back” and noted less starkly by Butt and Shah “… innovation is becoming increasingly important for the success of charities.” To rephrase my concern, are we utilizing technology in our work or are we serving the needs of our technology?

Finally, for anyone who’s curious about the Fourth Industrial Revolution, I have a December 3, 2015 posting about it.

MIT Media Lab releases new educational site for kids K-12: it’s all about artificial intelligence (AI)

Mark Wilson announces a timely new online programme from the Massachusetts Institute of Technology (MIT) in his April 9, 2020 article for Fast Company (Note: Links have been removed).

Not every child will grow up to attend MIT, but that doesn’t mean they can’t get a jump start on its curriculum. In response to the COVID-19 pandemic, which has forced millions of students to learn from home, MIT Media Lab associate professor Cynthia Breazeal has released [April 7, 2020] a website for K-12 students to learn about one of the most important topics in STEM [science, technology, engineering, and mathematics]: artificial intelligence.

The site provides 60 activities, lesson plans, and links to interactive AI experiments that MIT and companies like Google have developed in the past. Projects include coding robots to doodle, developing an image classifier (a tool that can identify images), writing speculative fiction to tackle the murky ethics of AI, and developing a chatbot (your grade schooler cannot possibly be worse at that task than I was). Everything is free, but schools are supposed to license lesson plans from MIT before adopting them.

Various associated MIT groups are covering a wide range of topics including the already mentioned AI ethics, as well as, cyber security and privacy issues, creativity, and more. Here’s a little something from a programme for the Girl Scouts of America, which focused on data privacy and tech policy,

The Girl Scouts awarded the Brownie (7-9) and Junior (9-11) troops with Cybersecurity badges at the end of the full event. 
Credit: Daniella DiPaola [downloaded from https://www.media.mit.edu/posts/data-privacy-policy-to-practice-with-the-girl-scouts/]

You can find MIT’s AI education website here. While the focus is largely on children, it seems they are inviting adults to participate as well. At least that’s what I infer from what one of the groups associated with this AI education website, the LifeLong Kindergarten group states on their webpage,

The Lifelong Kindergarten group develops new technologies and activities that, in the spirit of the blocks and finger paint of kindergarten, engage people in creative learning experiences. Our ultimate goal is to foster a world full of playfully creative people, who are constantly inventing new possibilities for themselves and their communities.

The website is a little challenging with regard to navigation but perhaps these links to the Research Projects page will help you get started quickly or, for those who like to investigate a little further before jumping in, this News page (which is a blog) might prove helpful.

That’s it for today. I wish everyone a peaceful long weekend while we all observe as joyfully and carefully as possible our various religious and seasonal traditions. From my tradition to yours, Joyeuses Pâques!