If artificial intelligence is going to spread to trillions of devices, those devices will have to operate in a way that doesn't need a human to run them, a Google executive who leads a key part of the search giant's machine learning software told a conference of chip designers this week.

"The only way to scale up to the kinds of hundreds of billions or trillions of devices we are expecting to emerge into the world in the next few years is if we take people out of the care and maintenance loop," said Pete Warden, who runs Google's effort to bring deep learning to even the simplest embedded devices.

"You need to have peel-and-stick sensors," said Warden, ultra-simple, dirt-cheap devices that require only tiny amounts of power and cost pennies.

"And the only way to do that is to make sure that you don't need to have people going around and doing maintenance."

Warden was the keynote speaker Tuesday at a microprocessor conference held virtually, The Linley Fall Processor Conference, hosted by chip analysts The Linley Group.

Warden offered the assembled, mostly chip industry executives, a wish list, as he put it, for hardware for devices.

That wish list includes ultra-low-power chips that do away with complex memory access and file access mechanisms, and instead focus on the repetitive arithmetic operations required in machine learning. Machine learning makes heavy use of linear algebra, consisting of vector-matrix and matrix-matrix multiplication operations.

Embedded deep learning needs chips that have "more arithmetic," said Warden.

"ML workloads are usually compute-bound," he told the audience. "We load a few activation and weight values, and do a lot of arithmetic on them in registers."

Warden's vision is that of self-sufficient devices that would run on battery power, perhaps for years, without needing to connect to a wall socket very often, perhaps not ever.

That would exclude the Raspberry Pi, said Warden, and anything else that requires "mains power," being plugged into a wall, and things that draw watts of power from a battery, such as a smartphone.

Instead, "We are aiming at the edge of the edge," said Warden, devices that are even more resource-constrained than cell phones, things such as peel-and-stick sensors that can be used in industrial applications.

"We are really looking at running on devices that are less than a dollar, maybe even 50 cents in price, that have a very small form factor."

Also: What is edge computing? Here's why the edge matters and where it's headed

Such devices might draw a single milliwatt to operate, he said, which "is really important, because that means you have a device that can run on double-A batteries for a year or two years, or even via energy harvesting from solar or vibration."

The challenge at present for deep learning forms of machine learning, Warden told the audience, is that many deep learning neural networks can't run at all on embedded devices because of the diffuse requirements of all the many micro-controller platforms that exist.

"We interact with a lot of product teams inside Google trying to build very interesting new products, and product teams at companies all over the world, and we often have to say, No, that's not quite possible yet," Warden told the audience.

"Because what's happening is the technology around deep learning, and the kinds of models that you can actually build on the training side that would be useful for product features, they often can't actually be deployed on the kinds of devices that people have in their actual hardware platforms."

If such models could be made to run on those billions of devices, "they would enable a whole bunch of new experiences for users," he said.

Embedded machine learning of the kind Warden discussed is part of a broader movement called TinyML. Today, examples of TinyML are fairly limited, things such as the wake word that activates a phone, such as "Hey, Google," or "Hey, Siri." (Warden confided to the audience, with a chuckle, that he and colleagues have to refer to "Hey, Google" around the office as "Hey, G," in order not to have one another's phones going off constantly.)

Warden has been leading the software effort to make possible the kinds of ultra-light-weight devices he was talking about. That effort is called TensorFlow Lite Micro, or TF Micro.

Warden and colleagues built on the existing TensorFlow Lite framework that exports trained machine learning models to run on embedded devices. While TF Lite removes some of the complexity of TensorFlow to make it feasible in a smaller-footprint device, TF Micro goes even further, to make machine learning able to run in devices with as little as 20 kilobytes of RAM.

TF Micro was introduced this month in a formal research paper by Warden and colleagues. The researchers had to build a framework that would work across numerous chip instruction sets, work with low-power microcontrollers, and they had to design it to support a greatly-reduced number of operations, excluding functions such as loading files from external locations.

The team also had to handle refinement of machine learning models for low-resource devices, which meant optimizing the quantization of models, representing operands in 8-bit integer form, say, rather than 32-bit floating point.

What Warden and team settled on is an interpreter that runs multiple models simultaneously. Using an interpreter not only makes it possible to run across the plethora of embedded platforms, it also makes it possible to update machine learning models as they improve without having to recompile models for a given device.

Chips to run TF Micro will have to do things that get around the limited nature of the embedded framework, Warden said. While full-blown TensorFlow supports 1,200 operations, TF Micro only supports a small fraction of those.

As a result, chips for running inference have to be able to "fall back to general-purpose code" rather than supporting every single last instruction.

"One of the real drawbacks of a lot of hardware accelerators is that they fail to run a lot of the models that people want to run on them," said Warden. "We want custom accelerators to fall back to run general-propose code without a massive performance penalty."

Summing up his wish list, Warden told the audience, "Really, what I'm looking for is tens or hundreds of billions of operations per second per milliwatt."

Some of the demands may be beyond what's feasible at present, he acknowledged. "I would love to have megabytes of model storage space instead of kilobytes," although, "I understand that's challenging."

"And, of course, I want it cheaper," he said.

The Linley conference, now in its fifteenth year, has over 1,000 attendees this year, conference organizer Linley Gwennap told ZDNet, which is more than three times as many attendees as in prior years, when the event was held at hotel ballrooms in the Silicon Valley area.

The conference continues through today.

The rest is here:

Google AI executive sees a world of trillions of devices untethered from human care - ZDNet

The year 2020 continues to be difficult here on Earth, where the pandemic is exploding again in regions of the world that were once successful in containing it. Germany reported a record number of cases this week alongside Poland and the Czech Republic, as the U.S. counted 500,000 new cases. Its the backdrop to a tumultuous U.S. election, which experts fear will turn violent on election day. Meanwhile, Western and Southern states like Oregon, Washington, California, and Louisiana are reeling from historically destructive wildfires, severe droughts, and hurricanes.

Things are calmer in outer space, where scientists are applying AI to make exciting new finds. Processes that would have taken hours each day if performed by humans have been reduced to minutes, a testament to the good AI can achieve when used in a thoughtful way. While not necessarily groundbreaking, unprecedented, or state-of-the-art with regard to technique, the innovations are inspiring stories of discovery at a time when there isnt a surfeit of hope.

Earlier this month, researchers at NASAs Jet Propulsion Laboratory in California announced they had fed an algorithm 6,830 images taken by the Context Camera on NASAs Mars Reconnaissance Orbiter (MRO) to identify changes to the Martian surface. Given 112,000 images taken by the Context Camera, the AI tool spotted a cluster of craters in the Noctis Fossae region of Mars, including 20 new areas of interest that might have formed from a meteor impact between March 2010 and May 2012. NASA hopes to use similar classification technology on future Mars orbiters, which might provide a more complete picture of how often meteors strike Mars.

In August, researchers at the University of Warwick built a separate AI algorithm to dig through NASA data containing thousands of potential planet candidates. The team trained the system on data collected by NASAs now-retired Kepler Space Telescope, which spent nine years in deep space searching for new worlds. Once it learned to separate planets from false positives, it was used to analyze datasets that hadnt yet been validated, which is when it found 50 exoplanets.

And last week, Intel, the European Space Agency (ESA), and startupUbotica detailed what they claim is the first AI-powered satellite to orbit Earth: the desktop-sized PhiSat-1. It aims to solve the problem of clouds obscuring satellite photos by collecting a large number of images from space in the visible, near-infrared, and thermal-infrared parts of the electromagnetic spectrum and then filtering out cloud-covered images using AI algorithms. Future versions of the PhiSat-1 could look for fires when flying over areas prone to wildfire and notify responders in minutes rather than hours. Over oceans, which are typically ignored, they might spot rogue ships or environmental accidents, and over ice, they could track thickness and melting ponds to help monitor climate change.

AI is problematic in many respects; its biased, discriminatory, and harmful at its worst. We have written about how facial recognition algorithms tend to be less accurate when applied to certain racial and ethnic groups. Natural language processing models embed implicit and explicit gender biases, as well as toxic theories and conspiracies. And governments are investigating the use of AI and machine learning to wage deadly warfare.

This being the case, some AI like that applied to Martian landscapes, telescope snapshots, and cloudy satellite images can be a force for good. And in a year marked by tragedy and general skepticism about technology (and the tech industry), this positivity isnt just encouraging, but sorely needed.

For AI coverage, send news tips to Khari Johnson and Kyle Wiggers and be sure to subscribe to the AI Weekly newsletter and bookmark our AI channel, The Machine.

Thanks for reading,

Kyle Wiggers

AI Staff Writer

Here is the original post:

AI Weekly: In a chaotic year, AI is quietly accelerating the pace of space exploration - VentureBeat

Executive Summary

It is very, very easy for a well-intentioned AI practitioner to inadvertently do harm when they set out to do good AI has the power to amplify unfair biases, making innate biases exponentially more harmful. Because AI often interacts with complex social systems, where correlation and causation might not be immediately clear or even easily discernible AI practitioners need to build partnerships with community members, stakeholders, and experts to help them better understand the world theyre interacting with and the implications of making mistakes. Community-based system dynamics (CBSD) is a promising participatory approach to understanding complex social systems that does just that.

Artificial Intelligence (AI) has become one of the biggest drivers of technological change, impacting industries and creating entirely new opportunities. From an engineering standpoint, AI is just a more advanced form of data engineering. Most good AI projects function more like muddy pickup trucks than spotless race cars they are a workhorse technology that humbly makes a production line 5% safer or movie recommendations a little more on point. However, more so than many other technologies, it is very, very easy for a well-intentioned AI practitioner to inadvertently do harm when they set out to do good. AI has the power to amplify unfair biases, making innate biases exponentially more harmful.

As Google AI practitioners, we understand that how AI technology is developed and used will have a significant impact on society for many years to come. As such, its crucial to formulate best practices. This starts with the responsible development of the technology and mitigating any potential unfair bias which may exist, both of which require technologists to look more than one step ahead: not Will this delivery automation save 15% on the delivery cost? but How will this change affect the cities where we operate and the people at-risk populations in particular who live there?

This has to be done the old-fashioned way: by human data scientists understanding the process that generates the variables that end up in datasets and models. Whats more, that understanding can only be achieved in partnership with the people represented by and impacted by these variables community members and stakeholders, such as experts who understand the complex systems that AI will ultimately interact with.

How do we actually implement this goal of building fairness into these new technologies especially when they often work in ways we might not expect? As a first step, computer scientists need to do more to understand the contexts in which their technologies are being developed and deployed.

Despite our advances in measuring and detecting unfair bias, causation mistakes can still lead to harmful outcomes for marginalized communities. Whats a causation mistake? Take, for example, the observation during the Middle Ages that sick people attracted fewer lice, which led to an assumption that lice were good for you. In actual fact, lice dont like living on people with fevers. Causation mistakes like this, where a correlation is wrongly thought to signal a cause and effect, can be extremely harmful in high-stakes domains such as health care and criminal justice. AI system developers who usually do not have social science backgrounds typically do not understand the underlying societal systems and structures that generate the problems their systems are intended to solve. This lack of understanding can lead to designs based on oversimplified, incorrect causal assumptions that exclude critical societal factors and can lead to unintended and harmful outcomes.

For instance, the researchers who discovered that a medical algorithm widely used in the U.S. health care was racially biased against Black patients identified that the root cause was the mistaken causal assumption, made by the algorithm designers, that people with more complex health needs will have spent more money on health care. This assumption ignores critical factors such as lack of trust in the health care system and lack of access to affordable health care that tend to decrease spending on health care by Black patients regardless of the complexity of their health care needs.

Researchers make this kind of causation/correlation mistake all the time.But things are worse for a deep learning computer, which searches billions of possible correlations in order to find the most accurate way to predict data, and thus has billions of opportunities to make causal mistakes. Complicating the issue further, it is very hard, even with modern tools, such as Shapely analysis, to understand why such a mistake was made a human data scientist sitting in a lab with their supercomputer can never deduce from the data itself what the causation mistakes may be. This is why, among scientists, it is never acceptable to claim to have found a causal relationship in nature just by passively looking at data. You must formulate the hypothesis and then conduct an experiment in order to tease out the causation.

Addressing these causal mistakes requires taking a step back. Computer scientists need to do more to understand and account for the underlying societal contexts in which these technologies are developed and deployed.

Here at Google, we started to lay the foundations for what this approach might look like. In a recent paper co-written by DeepMind, Google AI, and our Trust & Safety team, we argue that considering these societal contexts requires embracing the fact that they are dynamic, complex, non-linear, adaptive systems governed by hard-to-see feedback mechanisms. We all participate in these systems, but no individual person or algorithm can see them in their entirety or fully understand them. So, to account for these inevitable blindspots and innovate responsibly, technologists must collaborate with stakeholders representatives from sociology, behavioral science, and the humanities, as well as from vulnerable communities to form a shared hypothesis of how they work. This process should happen at the earliest stages of product development even before product design starts and be done in full partnership with communities most vulnerable to algorithmic bias.

This participatory approach to understanding complex social systems called community-based system dynamics (CBSD) requires building new networks to bring these stakeholders into the process. CBSD isgrounded in systems thinking and incorporates rigorous qualitative and quantitative methods for collaboratively describing and understanding complex problem domains, and weve identified it as a promising practice in our research. Building the capacity topartner with communities in fair and ethical ways that provide benefits to all participants needs to be a top priority. It wont be easy. But the societal insights gained from a deep understanding of the problems that matter most to the most vulnerable in society can lead to technological innovations that are safer and more beneficial for everyone.

When communities are underrepresented in the product development design process, they are underserved by the products that result. Right now, were designing what the future of AI will look like. Will it be inclusive and equitable? Or will it reflect the most unfair and unjust elements of our society? The more just option isnt a foregone conclusion we have to work towards it. Our vision for the technology is one where a full range of perspectives, experiences and structural inequities are accounted for. We work to seek out and include these perspectives in a range of ways, including human rights diligence processes, research sprints, direct input from vulnerable communities and organizations focused on inclusion, diversity, and equity such as WiML (Women in ML) and Latinx in AI; many of these organizations are also co-founded and co-led by Googler researchers, such as Black in AI and Queer in AI.

If we, as a field, want this technology to live up to our ideals, then we need to change how we think about what were building to shift to our mindset from building because we can to building what we should. This means fundamentally shifting our focus to understanding deep problems and working to ethically partner and collaborate with marginalized communities. This will give us a more reliable view of both the data that fuels our algorithms and the problems we seek to solve. This deeper understanding could allow organizations in every sector to unlock new possibilities of what they have to offer while being inclusive, equitable and socially beneficial.

Go here to see the original:

AI Engineers Need to Think Beyond Engineering - Harvard Business Review

93% of security operations center employing AI and machine learning tools to detect advanced threats | 2020-10-30 | Security Magazine This website requires certain cookies to work and uses other cookies to help you have the best experience. By visiting this website, certain cookies have already been set, which you may delete and block. By closing this message or continuing to use our site, you agree to the use of cookies. Visit our updated privacy and cookie policy to learn more. This Website Uses CookiesBy closing this message or continuing to use our site, you agree to our cookie policy. Learn MoreThis website requires certain cookies to work and uses other cookies to help you have the best experience. By visiting this website, certain cookies have already been set, which you may delete and block. By closing this message or continuing to use our site, you agree to the use of cookies. Visit our updated privacy and cookie policy to learn more.

Read this article:

93% of security operations centers employing AI and machine learning tools to detect advanced threats - Security Magazine

Photo: Baona | Getty Images

I Asked AI to Write This Post for Me. Here Are the Results.

In June 2020 an Artificial Intelligence system called GPT-3 went live. This AI model is focused on Natural Language Programming and was trained by reading trillions of words and sentences online. The net result is that it can generate impressive text that humans can barely tell was created by a computer.

A growing number of developers are being given access to GPT-3 to create real-world applications. In the coming months, you are going to start to see a plethora of AI applications that create content such as blogs, articles, reports, emails, advertising copy, and sales scripts.Its likely that huge amounts of text that will be written in 2021and beyond will be written by computers and edited by humans.

Related:3 Ways Small Businesses Can Quickly EmbraceAIWithout Big Data or Programmers

I was given access to a beta version of an application called Copy.ai which is specifically designed to create copy that could be suitable for marketing purposes. I asked the system to write me a short blog post titled Can an AI write business blogs. In under 3 seconds, it produced the following blog post:

Here are 10 things you need to know about using AI to write your business blogs:

1. AI is fast Very fast.

2. AI produces unique content Every time.

3. AI allows you to scale your content production Using AI to produce your content means you dont have to spend your time writing.

4. AI is customized for your industry AI can track the key topics in your industry and make sure that the content it produces is relevant and on-trend.

5. AI is a creative writing tool.

6. AI is not a replacement for copywriters.

7. AI will give you consistent quality content.

8. AI will allow you to produce better quality content than before.

9. AI can help you to produce content for your website, blog, social media and emails.

10. AI is not a one-size-fits-all solution.

Our AI software is completely customizable to your needs and is designed to produce content that is relevant to your business. If youre looking for a business blogging tool that can help you to produce better content at a faster rate then please contact us today.

As you can see, this isnt a perfect or complete blog post. I'm not even sure if all of the 10 points are true.On the whole, though, its not a bad starting point for an interesting article and it would be a lot easier for a busy entrepreneur to edit or expand this blog post rather than starting from scratch. In practical terms, you could run the AI system a dozen times in under a minute and pick a reasonable starting point.

For now, I believe AI applications have the power to generate directionally correct content that an experienced professional could edit and improve. Within afew years, the next version of this system will be ready (GPT-4) and it will likely be ten times more powerful. Its possible that this more advanced system will write most things better than most humans by 2025.

Related:HowAISolutions Are Solving 5 Long-Standing Business Challenges

I can imagine a group of farmers in 1890 looking at a tractor for the first time and wondering what will happen to the millions of farmhands being paid to plow fields. They couldnt have imagined the types of jobs or lives we live today that are made possible because a machine does a job humans had to do for thousands of years. As we move into the 2020s, we are crossing a similar frontier for humanity which may displace millions of peoples jobsbut may also give rise to completely new ways of living and working.

Related:I Asked AI to Write This Post for Me. Here Are the Results.Become an Expert on the Future of FinTech and Blockchain with This $40 BundleHow Technology is Revolutionizing Beauty Ecommerce

Here is the original post:

I Asked AI to Write This Post for Me. Here Are the Results. - The Advocate

For countless Americans, the United States military epitomizes nonpareil technological advantage. Thankfully, in many cases, we live up to it.

But our present digital reality is quite different, even sobering. Fighting terrorists for nearly 20 years after 9/11, we remained a flip-phone military in what is now a smartphone world. Infrastructure to support a robust digital force remains painfully absent. Consequently, service members lead personal lives digitally connected to almost everything and military lives connected to almost nothing. Imagine having some of the worlds best hardwarestealth fighters or space planessupported by the worlds worst data plan.

Meanwhile, the accelerating global information age remains dizzying. The year 2020 is on track to produce 59 zetabytes of data. Thats a one with 21 zeroes after itover 50 times the number of stars in the observable universe. On average, every person online contributes 1.7 megabytes of content per second, and counting. Taglines like Data is the new oil emphasize the economic import, but not its full potential. Data is more reverently captures its ever evolving, artificially intelligent future.

WIRED OPINION

ABOUT

Will Roper is the Air Force and Space Force acquisition executive.

The rise of artificial intelligence has come a long way since 1945, when visionary mathematician Alan Turing hypothesized that machines would one day perform intelligent functions, like playing chess. Aided by meteoric advances in data processinga million-billion-fold over the past 70 yearsTurings vision was achieved only 52 years later, when IBMs Deep Blue defeated the reigning world chess champion, Garry Kasparov, with select moves described as almost human. But this impressive feat would be dwarfed in 2016 when Googles AlphaGo shocked the world with a beyond-human, even beautiful move on its way to defeating 18-time world Go champion Lee Sedol. That now famous move 37 of game two was the death knell of human preeminence in strategy games. Machines now teach the worlds elite how to play.

China took more notice of this than usual. Weve become frustratingly accustomed to them copying or stealing US military secretstwo decades of post-9/11 operations provides a lot of time to watch and learn. But Chinas ambitions far outstrip merely copying or surpassing our military. AlphaGos victory was a Sputnik moment for the Chinese Communist Party, triggering its own NASA-like response: a national Mega-Project in AI. Though there is no moon in this digital space race, its giant leap may be the next industrial revolution. The synergy of 5G and cloud-to-edge AI could radically evolve the internet of things, enabling ubiquitous AI and all the economic and military advantages it could bestow. It's not just our military that needs digital urgency: Our nation must wake up fast. The only thing worse than fearing AI itself is fearing not having it.

There is a gleam of hope. The Air Force and Space Force had their own move 37 moment last month during the first AI-enabled shoot-down of a cruise missile at blistering machine speeds. Though happening in a literal flash, this watershed event was seven years in the making, integrating technologies as diverse as hypervelocity guns, fighters, computing clouds, virtual reality, 4G LTE and 5G, and even Project Maventhe Pentagons first AI initiative. In the blink of a digital eye, we birthed an internet of military things.

Working at unprecedented speeds (at least for the Pentagon), the Air Force and Space Force are expanding this IoT.mil across the militaryand not a moment too soon. With AI surpassing human performance in more than just chess and Go, traditional roles in warfare are not far behind. Whose AI will overtake them? is an operative question in the digital space race. Another is how our military finally got off the launch pad.

More than seven years ago, I spearheaded the development of hypervelocity guns to defeat missile attacks with low-cost, rapid-fire projectiles. I also launched Project Maven to pursue machine-speed targeting of potential threats. But with no defense plug-n-play infrastructure, these systems remained stuck in airplane mode. The Air Force and Space Force later offered me the much-needed chance to create that digital infrastructurecloud, software platforms, enterprise data, even coding skillsfrom the ground up. We had to become a good software company to become a software-enabled force.

Read more here:

There's No Turning Back on AI in the Military - WIRED

In the end most teams used smaller models that produced specific parts of a song, like the chords or melodies, and then stitched these together by hand. Uncanny Valley used an algorithm to match up lyrics and melodies that had been produced by different AIs, for example.

Another team, Dadabots x Portrait XO, did not want to repeat their chorus twice but couldnt find a way to direct the AI to change the second version. In the end the team used seven models and cobbled together different results to get the variation they wanted.

It was like assembling a jigsaw puzzle, says Huang: Some teams felt like the puzzle was unreasonably hard, but some found it exhilarating, because they had so many raw materials and colorful puzzle pieces to put together.

Uncanny Valley used the AIs to provide the ingredients, including melodies produced by a model trained on koala, kookaburra, and Tasmanian devil noises. The people on the team then put these together.

Its like having a quirky human collaborator that isn't that great at songwriting but very prolific, says Sandra Uitdenbogerd, a computer scientist at RMIT University in Melbourne and a member of Uncanny Valley. We choose the bits that we can work with.

But this was more compromise than collaboration. Honestly, I think humans could have done it equally well, she says.

Generative AI models produce output at the level of single notesor pixels, in the case of image generation. They dont perceive the bigger picture. Humans, on the other hand, typically compose in terms of verse and chorus and how a song builds. There's a mismatch between what AI produces and how we think, says Cai.

Cai wants to change how AI models are designed to make them easier to work with. I think that could really increase the sense of control for users, she says.

Its not just musicians and artists who will benefit. Making AIs easier to use, by giving people more ways to interact with their output, will make them more trustworthy wherever theyre used, from policing to health care.

We've seen that giving doctors the tools to steer AI can really make a difference in their willingness to use AI at all, says Cai.

Read the original here:

To see what makes AI hard to use, ask it to write a pop song - MIT Technology Review

"There are considerable benefits of setting up a transatlantic digital community cooperating on Artificial Intelligence (AI) and emerging and disruptive technologies, where NATO can play a key role as a facilitator for innovation and exchange", said NATO Deputy Secretary General Mircea Geoan. On Wednesday (28 October 2020) he took part in a high-level virtual discussion on transatlantic cooperation in the era of AI, organised by the Atlantic Council's Future Europe Initiative and GeoTech Center.

Mr. Geoan engaged in this conversation alongside the Chair and Vice Chair of the National Security Commission on Artificial Intelligence (NSCAI), Dr. Eric Schmidt and Secretary Robert O. Work, and the Head of Cabinet of European Commission Executive Vice-President Margrethe Vestager, Ambassador Kim Jrgensen. They discussed what modern technologies mean for European and American defence and security stakeholders, why the United States and the European Union should cooperate on AI, and how best to promote shared values in the field.

"NATO is a natural platform for transatlantic cooperation of AI," the Deputy Secretary General underlined. NATO offers its consultative mechanisms and unique networks for collaboration on defence and security questions. Bringing together Allies and partners, public and private sector, innovators and industry. We have great communities in areas like military capability development, science and technology, standardisation - and of course our Command Structure and military exercises. We also have new cross-cutting policy teams on Innovation Policy, who cover AI, and on Data Policy, he pointed out.

See more here:

Cooperation on Artificial Intelligence will boost security and prosperity on both sides of the Atlantic - NATO HQ

Every day we are likely to interact with some form of artificial intelligence (AI). It works behind the scenes in everything from social media and traffic navigation apps to product recommendations and virtual assistants.

AI systems can perform tasks or make predictions, recommendations or decisions that would usually require human intelligence. Their objectives are set by humans but the systems act without explicit human instructions.

As AI plays a greater role in our lives both at work and at home, questions arise. How willing are we to trust AI systems? And what are our expectations for how AI should be deployed and managed?

To find out, we surveyed a nationally representative sample of more than 2,500 Australians in June and July 2020. Our report, produced with KPMG and led by Nicole Gillespie, shows Australians on the whole dont know a lot about how AI is used, have little trust in AI systems, and believe it should be carefully regulated.

Trust is central to the widespread acceptance and adoption of AI. However, our research suggests the Australian public is ambivalent about trusting AI systems.

Nearly half of our respondents (45%) are unwilling to share their information or data with an AI system. Two in five (40%) are unwilling to rely on recommendations or other output of an AI system.

Further, many Australians are not convinced about the trustworthiness of AI systems, but more are likely to perceive AI as competent than to be designed with integrity and humanity.

Despite this, Australians generally accept (42%) or tolerate AI (28%), but few approve (16%) or embrace (7%) it.

When it comes to developing and using AI systems, our respondents had the most confidence in Australian universities, research institutions and defence organisations to do so in the public interest. (More than 81% were at least moderately confident.)

Australians have least confidence in commercial organisations to develop and use AI (37% no or low confidence). This may be due to the fact that most (76%) believe commercial organisations use AI for financial gain rather than societal benefit.

These findings suggest an opportunity for businesses to partner with more trusted entities, such as universities and research institutions, to ensure that AI is developed and deployed in an ethical and trustworthy way that protects human rights. They also suggest businesses need to think further about how they can use AI in ways that create positive outcomes for stakeholders and society more broadly.

Read more: Your questions answered on artificial intelligence

Overwhelmingly (96%), Australians expect AI to be regulated and most expect external, independent oversight. Most Australians (over 68%) have moderate to high confidence in the federal government and regulatory agencies to regulate and govern AI in the best interests of the public.

However, the current regulation and laws fall short of community expectations.

Our findings show the strongest driver of trust in AI is the belief that the current regulations and laws are sufficient to make the use of AI safe. However, most Australians either disagree (45%) or are ambivalent (20%) that this is the case.

These findings highlight the need to strengthen the regulatory and legal framework governing AI in Australia, and to communicate this to the public, to help them feel comfortable with the use of AI.

What do Australians expect when AI systems are deployed? Most of our respondents (more than 83%) have clear expectations of the principles and practices they expect organisations to uphold in the design, development and use of AI systems in order to be trusted.

These include:

high standards of robust performance and accuracy

data privacy, security and governance

human agency and oversight

transparency and explainability

fairness, inclusion and non-discrimination

accountability and contestability

risk and impact mitigation.

Read more: Will we ever agree to just one set of rules on the ethical development of artificial intelligence?

Most Australians (more than 70%) would also be more willing to use AI systems if there were assurance mechanisms in place to bolster standards and oversight. These include independent AI ethics reviews, AI ethics certifications, national standards for AI explainability and transparency, and AI codes of conduct.

Organisations can build trust and make consumers more willing to use AI systems, when they are appropriate, by clearly supporting and implementing ethical practices, oversight and accountability.

Most Australians (61%) report having a low understanding of AI, including low awareness of how and when it is used. For example, even though 78% of Australians report using social media, almost two in three (59%) were unaware that social media apps use AI. Only 51% report even hearing or reading about AI in the past year. This low awareness and understanding is a problem given how much AI is being used in our daily lives.

The good news is most Australians (86%) want to know more about AI. When we consider these factors together, there is a need and an appetite for a public literacy program in AI.

One model for this comes from Finland, where a government-backed course in AI literacy aims to teach more than 5 million EU citizens. More than 530,000 students have enrolled in the course so far.

Overall, our findings suggest public trust in AI systems can be improved by strengthening the regulatory framework for governing AI, living up to Australians expectations of trustworthy AI, and strengthening Australias AI literacy.

Read more: Your questions answered on artificial intelligence

Here is the original post:

Australians have low trust in artificial intelligence and want it to be better regulated - The Conversation Australia

Humans have been teaming up with machines throughout history to achieve goals, be it by using simple machines to move materials or complex machines to travel in space. But advances in artificial intelligence today bring possibilities for even more sophisticated teamwork true human-machine teams that cooperate to solve complex problems.

Much of the development of these human-machine teams focuses on the machine, tackling the technology challenges of training AI algorithms to perform their role in a mission effectively. But less focus, MIT Lincoln Laboratory researchers say, has been given to the human side of the team. What if the machine works perfectly, but the human is struggling?

"In the area of human-machine teaming, we often think about the technology for example, how do we monitor it, understand it, make sure it's working right. But teamwork is a two-way street, and these considerations aren't happening both ways. What we're doing is looking at the flip side, where the machine is monitoring and enhancing the other side the human," says Michael Pietrucha, a tactical systems specialist at the laboratory.

Pietrucha is among a team of laboratory researchers that aims to develop AI systems that can sense when a person's cognitive fatigue is interfering with their performance. The system would then suggest interventions, or even take action in dire scenarios, to help the individual recover or to prevent harm.

"Throughout history, we see human error leading to mishaps, missed opportunities, and sometimes disastrous consequences," says Megan Blackwell, former deputy lead of internally funded biological science and technology research at the laboratory. "Today, neuromonitoring is becoming more specific and portable. We envision using technology to monitor for fatigue or cognitive overload. Is this person attending to too much? Will they run out of gas, so to speak? If you can monitor the human, you could intervene before something bad happens."

This vision has its roots in decades-long research at the laboratory in using technology to "read" a person's cognitive or emotional state. By collecting biometric data such as video and audio recordings of a person speaking and processing these data with advanced AI algorithms, researchers have uncovered biomarkers of various psychological and neurobehavioral conditions. These biomarkers have been used to train models that can accurately estimate the level of a person's depression, for example.

In this work, the team will apply their biomarker research to AI that can analyze an individual's cognitive state, encapsulating how fatigued, stressed, or overloaded a person is feeling. The system will use biomarkers derived from physiological data such as vocal and facial recordings, heart rate, EEG and optical indications of brain activity, and eye movement to gain these insights.

The first step will be to build a cognitive model of an individual. "The cognitive model will integrate the physiological inputs and monitor the inputs to see how they change as a person performs particular fatiguing tasks," says Thomas Quatieri, who leads several neurobehavioral biomarker research efforts at the laboratory. "Through this process, the system can establish patterns of activity and learn a person's baseline cognitive state involving basic task-related functions needed to avoid injury or undesirable outcomes, such as auditory and visual attention and response time."

Once this individualized baseline is established, the system can start to recognize deviations from normal and predict if those deviations will lead to mistakes or poor performance.

"Building a model is hard. You know you got it right when it predicts performance," says William Streilein, principal staff in the Lincoln Lab's Homeland Protection and Air Traffic Control Division. "We've done well if the system can identify a deviation, and then actually predict that the deviation is going to interfere with the person's performance on a task. Humans are complex; we compensate naturally to stress or fatigue. What's important is building a system that can predict when that deviation won't be compensated for, and to only intervene then."

The possibilities for interventions are wide-ranging. On one end of the spectrum are minor adjustments a human can make to restore performance: drink coffee, change the lighting, get fresh air. Other interventions could suggest a shift change or transfer of a task to a machine or other teammate. Another possibility is using transcranial direct current stimulation, aperformance-restoringtechniquethat uses electrodes to stimulate parts of the brain and has been show to bemore effective than caffeinein countering fatigue, with fewer side effects.

On the other end of the spectrum, the machine might take actions necessary to ensure the survival of the human team member when the human is incapable of doing so. For example, an AI teammate could make the "ejection decision" for a fighter pilot who has lost consciousness or the physical ability to eject themselves. Pietrucha, a retired colonel in the U.S. Air Force who has had many flight hours as a fighter/attack aviator, sees the promise of such a system that "goes beyond the mere analysis of flight parameters and includes analysis of the cognitive state of the aircrew, intervening only when the aircrew can't or wont," he says.

Determining the most helpful intervention, and its effectiveness, depends on a number of factors related to the task at hand, dosage of the intervention, and even a user's demographic background. "There's a lot of work to be done still in understanding the effects of different interventions and validating their safety," Streilein says. "Eventually, we want to introduce personalized cognitive interventions and assess their effectiveness on mission performance."

Beyond its use in combat aviation, the technology could benefit other demanding or dangerous jobs, such as those related to air traffic control, combat operations, disaster response, or emergency medicine. "There are scenarios where combat medics are vastly outnumbered, are in taxing situations, and are as every bit as tired as everyone else. Having this kind of over-the-shoulder help, something to help monitor their mental status and fatigue, could help prevent medical errors or even alert others to their level of fatigue," Blackwell says.

Today, the team is pursuing sponsorship to help develop the technology further. The coming year will be focused on collecting data to train their algorithms. The first subjects will be intelligence analysts, outfitted with sensors as they play a serious game that simulates the demands of their job. "Intelligence analysts are often overwhelmed by data and could benefit from this type of system," Streilein says. "The fact that they usually do their job in a 'normal' room environment, on a computer, allows us to easily instrument them to collect physiological data and start training."

"We'll be working on a basis set of capabilities in the near term," Quatieri says, "but an ultimate goal would be to leverage those capabilities so that, while the system is still individualized, it could be a more turnkey capability that could be deployed widely, similar to how Siri, for example, is universal but adapts quickly to an individual." In the long view, the team sees the promise of a universal background model that could represent anyone and be adapted for a specific use.

Such a capability may be key to advancing human-machine teams of the future. As AI progresses to achieve more human-like capabilities, while being immune from the human condition of mental stress, it's possible that humans may present the greatest risk to mission success. An AI teammate may know just how to lift their partner up.

View post:

Stressed on the job? An AI teammate may know how to help - MIT News

The diffusion trend for artificial intelligence inventor-patentees started at 1% in 1976 and increased to 25% in 2018, which means that 25% of all unique inventor-patentees in 2018 used AI technologies in their granted patents.

On October 27, the United States Patent and Trademark Office (USPTO) released a report titled Inventing AI: Tracing the diffusion of artificial intelligence with U.S. patents. The study showed that artificial intelligence (AI) patent applications increased by more than 100% between 2002 and 2018, from 30,000 to over 60,000, and the overall share of patent applications containing AI subject matter rose from 9% to nearly 16%.

According to the U.S. National Institute of Standards and Technology (NIST), AI technologies and systems comprise software and/or hardware that can learn to solve complex problems, make predictions or undertake tasks that require human-like sensing (such as vision, speech, and touch), perception, cognition, planning, learning, communication, or physical action. However, for purposes of patent applications and grants, the USPTO defines AI as including one or more of eight component technologies: vision, planning/control, knowledge processing, speech, AI hardware, evolutionary computation, natural language processing, and machine learning. Between the years of 1990 and 2018, the largest AI technological areas were planning/control and knowledge processing, which include inventions directed to controlling systems, developing plans, and processing information. In addition, the study showed that patent applications in the areas of machine learning and computer visions have shown a pronounced increase since 2012.

The study explained that, since 1976, AI technologies have been diffusing across a large percentage of technology subclasses, spreading from 10% in 1976 to more than 42% of all patent technology subclasses in 2018. The study identified three distinct clusters with different diffusion rates in order from the fastest to the slowest growing: 1.) knowledge processing and planning/control, 2.) vision, machine learning, and AI hardware, 3.) revolutionary computing, speech, and natural language processing. The study noted that the clusters suggest a form of technological interdependence among the AI component technologies, but also noted that additional research is required to understand the factors behind the patterns.

The study also identified the growth in the number of AI inventors as an indicator of diffusion. In particular, the diffusion trend for inventor-patentees started at 1% in 1976 and increased to 25% in 2018, which means that 25% of all unique inventor-patentees in 2018 used AI technologies in their granted patents.

Noting that AI requires specialized knowledge, the study pointed out that diffusion is generally slower and can be restricted to a narrow set of organizations in areas where skilled labor and technical information are harder to obtain, such as in AI. The study identified the top 30 U.S. companies that held 29% of all AI patents granted from 1976 to 2018. The leading company was IBM Corp. with 46,752 patents, followed by Microsoft Corp. with 22,067 patents and Google Inc. with 10,928 patents.

With respect to geographic diffusion of AI, the study indicated that, between 1976 and 2000, AI inventor-patentees tended to be concentrated in larger cities or established technology hubs, such as Silicon Valley, California, because those regions were home to companies with employees having the specialized knowledge required to understand AI technologies. Since 2001, AI inventor-patentees have diffused widely across the U.S. For example, Maine and South Carolina are active in digital data processing and data processing adapted for businesses, Oregon is active in fitness training and equipment, and Montana is active in inventions analyzing the chemical and physical properties of materials. The study also showed that the American Midwest is adopting AI technology, but at a slower rate. For example, Wisconsin leads in medical instruments and processes for diagnosis, surgery, and identification and Iowa, Kansas, Missouri, Nebraska, and Ohio are contributing to AI technologies relating to telephonic communications. Further, inventor-patentees in North Dakota are actively contributing to AI technologies as applied to agriculture.

The USPTO noted that the study suggests that AI has the potential to be as revolutionary as electricity or the semiconductor and depends, at least in part, on the ability of innovators and firms to successfully incorporate AI inventions into existing and new products, processes, and services.

The report results were obtained from a machine learning AI algorithm that determined the volume, nature, and evolution of AI and its component technologies as contained in U.S. patents from 1976 through 2018. This methodology improved the accuracy of identifying AI patents by better capturing the diffusion of AI across technology, companies, inventor-patentees, and geography.

Rebecca Tapscott is an intellectual property attorney who has joined IPWatchdog as our Staff Writer. She received her Bachelor of Science degree in chemistry from the University of Central Florida and received her Juris Doctorate in 2002 from the George Mason School of Law in Arlington, VA.

Prior to joining IPWatchdog, Rebecca has worked as a senior associate attorney for the Bilicki Law Firm and Diederiks & Whitelaw, PLC. Her practice has involved intellectual property litigation, the preparation and prosecution of patent applications in the chemical, mechanical arts, and electrical arts, strategic alliance and development agreements, and trademark prosecution and opposition matters. In addition, she is admitted to the Virginia State Bar and is a registered patent attorney with the United States Patent and Trademark Office. She is also a member of the American Bar Association and the American Intellectual Property Law Association.

Continue reading here:

USPTO Releases Benchmark Study on the Artificial Intelligence Patent Landscape - IPWatchdog.com

"COPAN's PhenoMATRIX not only was able to detect more true positive cultures than manual review of digital culture images, but it shows that chromogenic cultures together with artificial intelligence algorithms can detect GBS colonization with the same high sensitivity as molecular detection systems," said COPAN Diagnostics' Scientific Director Dr. Susan Sharp.

The study, which was published on October 21, 2020, evaluated the performance of the PhenoMatrix Chromogenic Detection Module digital imaging software's ability to detect GBS from LIM broth plated on bioMrieux's CHROMID Strepto Battwo clinical laboratories.

After 48 hours of incubation, the sensitivity of COPAN's PhenoMATRIX was similar to the BD MAX GBS molecular test 95.5% to 96.8% respectively and significantly higher than manual at 90.3%.2

Another noteworthy discovery was that COPAN's software never inaccurately called a culture that was determined to be a positive a negative, and it identified an additional eight true positive specimens that were missed by manual reading. This finding establishes that the innovative PhenoMATRIX AI, plus classic culture, is quite the powerful combination at a fraction of the cost of molecular testing.

"COPAN's AI software, along with the use of chromogenic agars has made our decades-old agar culture for the detection of pathogens 'new' again," Sharp added.

PhenoMATRIX is an advanced set of highly sophisticated AI that gives WASPLab users the power to automatically pre-assess and pre-sort culture plates, read, interpret and segregate bacterial cultures. By grouping negative cultures, which are the majority of the cultures screened, staff can quickly review up to 40 plates per computer screen and batch release negative results eliminating the need to review each plate manually saving time and freeing up technicians to focus on more complex tasks.

Contact us for more information about COPAN's state-of-the-art PhenoMATRIX software and how you can add these intelligent algorithms to your WASPLab system.

References: 1. CDC. Group B Strep (GBS) Fast Facts. https://www.cdc.gov/groupbstrep/about/fast-facts.html. Last reviewed June 11, 2020. Accessed October 27, 2020. 2. Baker J, et al.Digital image analysis for the detection of Group BStreptococcusfrom ChromID StreptoB Media using a PhenoMatrix Artificial Intelligence Software Algorithm. J Clin Microbiol. 2020;doi:10.1128/JCM.01902-19

About COPANWith a reputation for innovation, COPAN is the leading manufacturer of collection and transport systems in the world. COPAN's collaborative approach to pre-analytics has resulted in Flocked Swabs, ESwab, UTM Universal Transport Medium, and laboratory automation, WASP and WASPLab. COPAN carries a range of microbial sampling products, inoculation loops, and pipettes. For more information, visitwww.copanusa.com.

SOURCE COPAN Diagnostics, Inc.

Home page

Visit link:

COPAN's PhenoMATRIX Fuses the Power of Artificial Intelligence and Culture for Highly Sensitive GBS Detection Using Breakthrough Reading Algorithm -...