Monthly Archives: May 2021

A new education focus around artificial intelligence for healthcare professionals and patients could be the way forward when it comes to the future of health. Jane Rendall from Sectra and Rachel Dunscombe, CEO of the NHS Digital Academy, explore.

A crisis point could be on the horizon for NHS imaging disciplines. Rising demand and pervasive recruitment challenges mean there will be too few experts to go around based on current ways of working.

We certainly dont want to reach that point, and to achieve that the health service will need to adopt artificial intelligence in new ways as an important mechanism in redesigning services.

For this to happen radiologists, pathologists and other ologists must master how AI works and how it could be used to achieve maximum impact.

These professionals, together with organisational and process experts, need to be given the headspace to work out how their profession will evolve in coming years, having taken the potential of this technology into account. They need to understand what part of their profession requires or can be strengthened by human judgement and engagement. And they need to be able to establish when decisions could be made quickly and automatically by AI.

What can be safely automated, should be automated, or have the option of being automated. More than an efficiency drive, this is a necessity to be able to deliver healthcare expected by citizens, and to facilitate early engagement and prevention.

The iceberg

There is a big education piece that needs to be undertaken in order for this complex redesign to happen effectively, and for AI to be used in more sophisticated ways than narrower diagnostic support uses often seen today.

Clinical professions are changing and will become more data driven. This will require a new skillset currently absent from learning, like understanding the technology and mathematical concepts behind algorithms.

There are four key areas where people need education and orientation, and the technology is just the tip of the iceberg.

Unless we tackle this iceberg whole, we wont achieve impact at scale and pace instead we risk creating orphaned silos of technology that dont fit into the healthcare system.

Thats why this needs to be part of continuous professional development and education for anyone in healthcare using AI. People need to understand what problems they are trying to solve, and ways in which that can be done safely.

Educating patients

When talking pathway redesign our radiologists, pathologists and others will need to understand how this AI is communicated to citizens. That includes the explanations that patients see, the outcomes and measures patients see, and informing choices presented to the patient, potentially via their patient portal. Many patients already get choices around how they receive information; this could extend to their diagnostic choices.

A potential future option to have a preliminary diagnosis in 30 seconds by choosing to use an algorithm to look at your image, rather than 15 days for a human counterpart to examine it, could be a valid option in many cases.

And if we can gather evidence over time of the efficacy of those choices, we can show that to patients.

We can move from prescribing a set of pathways to citizens to giving them more choice, to informing how they interact with an algorithm.

Conversely some patients might have a complex history and prefer an analogue approach. Patients might be advised to rely on a radiologist for complex cases. But for a relatively simple bone break, you might choose an algorithm. Humans add most value where there is complexity. Some of this is about choice, some will be about advice. And part of this equation is about determining where choice is appropriate.

Digitally ready workforce

This is transformation it is about how we are going to practice medicine or radiology in the future not orphaning tech along the way.

It is about empowering a digital and AI ready workforce to reimagine their own careers, their workplace and workflow.

The potential crisis point creates a sense of urgency, but this is also an opportunity to make service redesign everyones job so they are not just part of the service, they are part of the future.

Read the original post:

The NHS AI iceberg: below the surface - Digital Health

On the video chat, a woman waves and reaches out to a baby, bouncing on its mother's lap. The picture is so lifelike -- the baby has its fingers in its mouth -- it's as if the woman thinks she can touch the people on the other side of the screen.

It isn't a hologram, but it seems close.

In the throes of the coronavirus pandemic, Zoom became a lifeline. Now Google is trying to reinvent video chat with computer vision and machine learning to create realistic 3D images you can talk to without donning a special headset or glasses. The result is what Google calls a "magic window" that connects people across cities and countries. Right now it's being tested at Google offices in the San Francisco Bay Area, New York and Seattle, but the world's largest search provider will run trials with a few business partners later this year.

Called Project Starline, the initiative is one of the marquee announcements at the Google I/O developer conference, which made its return on Tuesday after skipping a year because of COVID-19. In 2020, people cocooned themselves at home to fend off the fast-spreading virus. Hospitals overflowed, and loved ones were lost. Tech executives announcing new products weren't on people's minds.

Starline creates 3D images of people for video chats.

Now much of the world is in a different place. Vaccination distribution is continuing in many countries, and more of the world is opening up. The search giant, too, hopes to turn the page. It kicked off the three-day I/O developer fest in a fashion that seems almost throwback: It was live.

True to the times, the new normal was a little different. The keynote was broadcast from a courtyard at Google's headquarters in Mountain View, California, rather than at the Shoreline Amphitheatre, the nearby outdoor venue that's hosted the confab since 2016. Instead of a large crowd of developers, the audience was a small group of Google employees. It was a step away from the glossy pretaped productions that Apple and Samsung have delivered in the pandemic era, even if it was still streamed.

"Over the past year, we've seen how technology can be used to help billions of people through the most difficult of times," CEO Sundar Pichai said in a statement ahead of his keynote speech. "It's made us more committed than ever to our goal of building a more helpful Google for everyone."

For Google, the chaos extended well beyond the pandemic. The company has weathered a host of controversies over the past 12 months. It's the target of three major antitrust lawsuits, including a landmark case by the US Department of Justice and another by a bipartisan coalition of nearly 40 states. Google's ouster of prominent artificial intelligence researchers sparked outrage across the industry. Pichai, who's been CEO since 2015, was hauled before Congress three times in the last year to defend the company in front of lawmakers concerned about everything from disinformation to alleged anti-conservative bias.

Over the past year, we've seen how technology can be used to help billions of people through the most difficult of times.

Sundar Pichai

Days before the virtual conference, Google offered a comprehensive look at its lineup, including new efforts in quantum computing, updates to familiar services such as Search and Maps, and a tool that uses your phone's camera to help people identify skin abnormalities. The announcements underscore how keen Pichai and his team are to put the spotlight on Google's product and engineering efforts and move past the unwanted attention on its business practices and corporate culture.

"They're focused on the product story, and that's the story they want to tell," said Bob O'Donnell, founder of Technalysis Research. "They're eager to show they're pushing forward."

One area of Google's business -- artificial intelligence -- neatly captures the company's complicated year. Google is a powerhouse in the field, and its I/O announcements put that prowess on display. But in the background, thorny issues underscore struggles that Google would like to move past.

One of the biggest unveilings at I/O is the announcement of a new campus in Santa Barbara, California, focused on quantum computing. It will be host to hundreds of Google employees and include a quantum data center, research labs and a chip manufacturing facility.

"This is a longer-term bet on an interesting technology that might be able to solve important problems that no classical computer can solve," says Jeff Dean, Google's AI chief.

Another highlight is a technology for Google Search called the "multitask unified model," or MUM, which Google calls a "milestone" in understanding information. The goal of MUM is to decipher complex questions that involve several steps and layers. For example, if someone asks Google how to prepare for climbing two different mountains, the search engine could in the future give that person information on the elevation of each mountain, as well as details on training and what gear to buy.

Google's quantum computing campus in Santa Barbara, California.

There's also a new service called Derm Assist, which uses AI to help people learn more about possible skin, hair and nail conditions. Using phone cameras, people upload pictures of an ailment and answer a series of questions. The software then matches the pictures to a database of 288 known conditions and gives them information on what the condition might be.

As helpful as that may sound, the service raises potential privacy concerns. Sending pictures of a worrisome mole to Google may be a little too creepy for some people. But Karen DeSalvo, Google's chief health officer, says people are already coming to Google with their health concerns, with almost 10 billion Google searches a year related to hair, skin and nail issues. Still, she acknowledges the trust barriers, especially as people become increasingly wary of big tech.

"It's something we really think about," DeSalvo says. "What we hope over time is that, as people see the information getting better-quality, as they're seeing that they're getting navigated to authoritative sources, they're going to increasingly trust us."

Though Google has made strides in AI, the division has been roiled for months by high-profile terminations and resignations over ethics and diversity. In December, Timnit Gebru, one of the few prominent Black women in the field, announced on Twitter that she had been fired over a research paper that called out risks for bias in AI, including in systems used by Google's search engine.

The fallout led to Google's firing of Margaret Mitchell, who founded the company's ethical AI unit and co-led it with Gebru, after an investigation over data security. Samy Bengio, who managed Gebru and Mitchell and voiced support for them, resigned last month.

"The reputational hit is a real thing," Google's Dean said in his first interview since the controversy became public in December. "But we have to move past this, and we are deeply committed to doing work in the space and feel it's a really important area."

Putting the spotlight back on products also means highlighting changes to Google products that billions of people use every day, including the Android mobile operating system, Maps and Search.

At I/O, the company trumpeted a new milestone for Android: It's now running on 3 billion devices globally, another sign of dominance for the most popular mobile software in the world. Android powers almost nine out of every 10 smartphones on the planet.

Other changes to Android include a major design makeover, the biggest aesthetic change to the platform since 2014. The new look includes a color extraction feature, which generates a color palette for phones by pulling similar hues from their wallpaper. Another feature now lets people unlock their cars using their phone as a digital key. It will also let people send their keys to other phones when they lend out their car. Google is partnering with BMW to debut the tool, and it will be available first on Google's Pixel phones and Samsung's Galaxy phones.

Google also updated its Maps app. One change is a direct result of the pandemic: a feature that tracks how crowded neighborhoods and other large areas are. Google reckons you'll want to know if a farmer's market, for example, has attracted a rush of people so you can avoid it if you want. Another feature adds more context to a map based on the time of day. So it won't show you, say, a closed breakfast place if you look at the map at 9 p.m.

The tech giant also introduced a product update to its most iconic service -- its search engine. Now results will include a label called "About this result" that provides users in the US more context about that source, in an attempt to combat misinformation. Google says it's working with Wikipedia to provide background on websites, including short descriptions. People can also see when the site was indexed and whether or not your connection to the site is secure.

"This allows you to see, from a given piece of information, more about the source," says Liz Reid, a vice president of engineering for Google Search. "So you can really understand: Why is this source the one that's telling you this information? And how much do you trust the source?"

A new Google Maps feature will tell you how busy an area or neighborhood is.

The change comes as Google and other tech giants face intense scrutiny over false information being spread on their platforms. When it comes to Google's services, the video site YouTube often gets the brunt of criticism for spreading misinformation and conspiracies, but shady search results can also be the culprit. In March, Pichai appeared virtually before Congress alongside Facebook CEO Mark Zuckerberg and Twitter chief Jack Dorsey to testify about the danger of misinformation on tech platforms.

"Staying ahead of these challenges and keeping users safe and secure on our platforms is a top priority," Pichai said at the time.

The search giant, which has long been criticized for its data practices, is also touting its efforts in privacy and security at I/O.

Many of Google's privacy changes are coming to Android 12. The most intriguing one is called Private Compute Core, which cordons off some data processing like speech recognition and machine vision from the rest of the operating system. The processing is done locally on the device and unconnected from the network, keeping the data more private. It powers Google features like Live Caption, which generates captions in real time, and Now Playing, which recognizes music playing in the area, similar to Shazam.

Another new feature allows people to limit what they share with apps, like only disclosing their approximate location instead of their exact whereabouts. The distinction could be useful, for example, if someone doesn't want to share their precise location with a weather app.

The pandemic has accelerated a lot of trends in terms of the amount of time people are spending online, on their various devices.

Jennifer Fitzpatrick

Privacy across apps has become a hot topic in the last few months, with Apple cracking down on data tracking on iPhones. One change by Apple requires developers to ask people for permission to gather data and track them across apps and websites. The update has riled some players in the broader tech industry. Facebook has been particularly vocal, and the policy has prompted a war of words between Zuckerberg and Apple CEO Tim Cook.

Asked why Google hasn't introduced a similar feature, Sameer Samat, vice president of product management for Android, suggested the company is working on it. "Who says we're not?" he replied. When pressed, he said the company has "nothing to announce" right now.

Jen Fitzpatrick, a senior vice president in charge of core experiences and infrastructure across Google, says the last year, in which people across the globe have come to rely on technology to stay connected while physically apart, has made privacy more urgent.

"The pandemic has accelerated a lot of trends in terms of the amount of time people are spending online, on their various devices," Fitzpatrick says. "It just reinforces how important it is to give users experiences that are safe."

Now Google will have to prove that privacy is a part of its normal.

More:

Google unveils Android, AI and chat features as it tries to move past chaotic year - CNET

TEL AVIV, Israel--(BUSINESS WIRE)--Visionary.ai, an AI image processing start-up, announced that it has closed a $4.5 million Seed Round led by Ibex Investors with the participation of Spring Ventures of Aviv Refuah, Capital Point of Yossi Tamar, and additional investors. Their mission is to enable high-quality digital imaging in all circumstances and conditions, and especially in situations in which visibility is currently limited (low light, fog, WDR, etc.).

Co-founders Oren Debbi (CEO) and Yoav Taieb (CTO) come with deep domain expertise in computer vision and AI. Oren has years of experience in business development and sales in the Visual AI space, and Yoav spent almost two decades pioneering algorithm development at Mobileye (Intel).

The company has developed AI based image signal processing algorithms that significantly improve on current technologies. This allows Visionarys software to drastically improve the quality of images and videos taken from all cameras in ways that were not possible before.

The company has entered into major contracts and has a robust pipeline with major players from various sectors in both Israel and the US. According to Oren Debbi, Visionary.ai offers camera manufacturers a new capability needed by many companies through the utilization of artificial intelligence. We are surprised and excited by the quick uptake of our technology by leading companies.

The edge computing market CAGR is expected to reach 37% between now and 2027 which will rapidly accelerate the adoption of this new technology. Nicole Priel from Ibex Investors says The companys software will have wide implications for AI on the edge and I truly believe that Visionary.ais ISP will become the new standard in every camera.

Over the next year, Visonariy.ai intends to invest heavily in its product and algorithms and expand its R&D team to change how and what we see today.

Read the original:

Visionary.ai Raises $4.5 Million to Help Revolutionize the Quality of Digital Imaging in Poor Light, Wide Dynamic Range, and Low Visibility - Business...

Unbounce, a Vancouver startup best known for helping marketers create automated landing pages, added a new wrinkle this morning when it announced it has acquired Snazzy.ai, an early-stage automated copywriting startup. The two companies did not share the terms.

Unbounce Chief Strategy Officer Tamara Grominsky says that her company focuses on helping customers convert their customers into sales, and with Snazzy, it gets some pretty nifty technology based on GPT-3 artificial intelligence technology.

Were focused right now on building conversion intelligence software that will allow marketers to work with machines to really unlock their true conversion potential [] and we saw a huge opportunity with Snazzy to focus particularly on the content creation and copy creation space to help us accelerate that strategy, Grominsky explained.

She points out that the product is really aimed at the marketing generalist charged with overseeing landing pages, and who is responsible for a range of tasks including writing copy. The average Unbounce customer isnt a specialized copywriter, so they dont spend [their work] day writing copy. Theyre what we would consider a marketing generalist or really someone whos responsible for a wide range of marketing responsibilities, she said.

Snazzy co-founder Chris Frantz says the tech is really about getting people started, and then they can tweak the results as needed. The hardest part has always been to get that first line, that first page, the first couple of words in and we eliminate that entirely. That might not always result in amazing copy, but on the plus side you can always click the button again and give it another try, he said.

Frantz says that with so much competition in the space, he and his co-founder felt they could build a market much faster as part of a larger and broader marketing platform solution like Unbounce.

I love Tamaras vision for the future of Unbounce. I think she has a very ambitious vision. She sold me on that very early on in the process. At the same time, there was a lot of competition in the space, and to have a key differentiator with a company like Unbounce, which has a decade of marketing experience and a lot of trust within this community, I think its a very powerful wedge that we can use to further grow our audience, Frantz said.

The tool lets you write a range of copy, from landing pages to Google ad copy. The company launched in alpha last October and already had 30,000 customers, which Grominsky says Unbounce hopes to convert into customers. The good news for those customers is that the company plans to leave Snazzy as a standalone product, while incorporating the tech into the platform in ways that make sense in the coming year.

See the rest here:

Unbounce snags Snazzy.ai to add automated copywriting to platform - TechCrunch

At its I/O developer conference, Google today announced the next generation of its custom Tensor Processing Units (TPU) AI chips. This is the fourth generation of these chips, which Google says are twice as fast as the last version. As Google CEO Sundar Pichai noted, these chips are then combined in pods with 4,096 v4 TPUs. A single pod delivers over one exaflop of computing power.

Google, of course, uses the custom chips to power many of its own machine learning services, but it will also make this latest generation available to developers as part of its Google Cloud platform.

This is the fastest system weve ever deployed at Google and a historic milestone for us, Pichai said. Previously to get an exaflop you needed to build a custom supercomputer, but we already have many of these deployed today and will soon have dozens of TPUv4 pods in our data centers, many of which will be operating at or near 90% carbon-free energy. And our TPUv4 pods will be available to our cloud customers later this year.

The TPUs were among Googles first custom chips. While others, including Microsoft, decided to go with more flexible FPGAs for its machine learning services, Google made an early bet on these custom chips. They take a bit longer to develop and quickly become outdated as technologies change but can deliver significantly better performance.

Read the rest here:

Google launches the next generation of its custom AI chips - TechCrunch

Building on a deal first struck in 2019, artificial intelligence (AI) specialist Exscientia Ltd. has agreed to take responsibility for a multitarget drug discovery collaboration with Bristol Myers Squibb Co. that could be worth more than $1.2 billion in all. The expanded collaboration, first established with BMS-acquired Celgene Corp., includes $50 million in up-front funding, up to $125 million in near to mid-term potential milestones, plus additional clinical, regulatory and commercial payments. It remains focused on small-molecule drug candidates in areas including oncology and immunology.

Under terms of the agreement, Exscientia will take responsibility for AI-design and experimental work necessary to discover drug candidates for BMS. Neither party has disclosed specific products of their work yet, though Exscientia CEO Andrew Hopkins told BioWorld that his company plans to announce milestones as the projects develop. Should they succeed, Exscientia would receive tiered royalties on net sales of any marketed drug products resulting from the collaboration.

The deal provided yet further recognition of the tremendous interest and belief in AI-driven discovery approaches that have bloomed within big pharma. Momentum behind the trend has driven deals for Exscientia with Roche Holding AG, Glaxosmithkline plc, Sanofi SA and Evotec AG, as well as substantial investments in the company.

In April, Exscientia closed a $225 million series D financing led by Softbank Vision Fund 2, replete with an additional $300 million equity commitment. Prior to that in March, the Oxford, U.K.-based company extended its series C financing, adding a further $40 million to the $60 million raised in May 2020. Earlier financings included a $26 million series B round and $17.6 million in series A support.

Underpinning the BMS deal, as well as the others, is Exscientia's Centaur AI platform, an approach that the company said allows it to "identify emerging hotspots of opportunity" from genetic data and global biological literature before applying those insights to "learn which areas of chemistry are most likely to balance complex requirements" for each discovery project.

In a statement about the deal, BMS president of research and early development Rupert Vessey said Exscientia would be applying AI technologies proving "capable of generating best-in-class molecules while also reducing discovery times." BMS declined to make Vessey available for an interview to discuss the deal.

While its not yet clear how AI-centric approaches to discovery will ultimately stack up to more traditional approaches in terms of efficiency or clinical success, progress is clearly underway.

Since its founding in 2012, Exscientia has advanced three AI-designed molecules to the clinic: EXS-21546, an in-house adenosine 2A receptor antagonist in testing for the potential oral treatment of cancer, and two programs partnered with Sumitomo Dainippon Pharma Co. Ltd. (DSP), of Osaka, Japan.

The first of the DSP-partnered programs is DSP-1181, an oral long-acting 5-HT1A receptor agonist under development as a potential treatment for obsessive-compulsive disorder. Clinical development of the candidate began in January 2020, less than 12 months after the project began, contrasting with a typical average of 4.5 years from discovery to the clinic using conventional techniques, Exscientia said last year.

Adding to the progress, on May 13 Exscientia announced a second DSP-partnered candidate is entering a phase I study in the U.S. for the treatment of Alzheimer's disease psychosis. The molecule, DSP-0038, is both an antagonist for the 5-HT2A receptor and agonist for the 5-HT1A receptor, while selectively avoiding similar receptors and unwanted targets, such as the dopamine D2 receptor, the partners said.

As made clear in an announcement of its most recent financing, Exscientia has no small ambitions around the potential for its AI-driven platforms, noting it will "continue expanding the technology platform toward autonomous drug design." The objective, Hopkins said, is "enabling automated systems to also make key decisions by themselves, minimizing the requirements for human intervention. For autonomous driving, for example, the car will sense when it needs to brake and will then perform that response."

"It is the same principle for drug design and discovery where an autonomous system would determine what type of experiment would be required," he said.

See the article here:

Exscientia and BMS expand AI drug discovery deal, with potential $1.2B+ value - BioWorld Online

Skilling up.

The purpose of automation is to reduce the amount of human effort needed to do a task. But we need more human effort than ever to build and maintain those automated systems.

Thats the paradox of technology, and now that parts of the world are emerging on the other side of the Covid crisis (hopefully), the job market is heating up again, and companies are hungry for human talent who can alleviate their need to hire human talent. Amazon which turned many labor markets on their heads by accelerating e-commerce and on-site automation recognizes this shift, providing free cloud computing training to 29 million people around the world by 2025, in addition to committing more than $700 million to upskill 100,000 of its own employees in that time.

Cognizant released its quarterly Jobs of the Future Index, predicting a strong recovery for the US jobs market this coming year, especially those involving technology. The technology market, which saw lower rates of hiring six to eight months ago, but is primed for strong recovery as organizations accelerate their adoption of cloud strategies and AI solutions, the survey shows. Of all the jobs tracked in the Cognizant report, AI, algorithm and automation jobs saw a 28% gain over the last quarter, with robotics engineer and video game designer being two of the fastest-growing jobs - with 73% and 54% growth in jobs postings for these positions in the last three months. Algorithms, automation, and AI, the largest family in the index, realized a 28% gain over the quarter.

Ardine Williams, VP of workforce development at Amazon, believes the past year was one of revitalization and reinvention for many people as they reconsider their career options. The scale of the disruption was staggering and so was the pace of adaptation, she says. Many used this disruption as an opportunity to revitalize or to reinvent their skills. Last year we saw unprecedented participation in cloud skills training during the lockdown. This interest came from learners on both ends of the spectrum from those who were completely new to the cloud and looking to understand the fundamentals, to seasoned professionals who were looking to take advanced courses in areas like machine learning.

This interest was reflected in AWSs array of cloud courses. The number of learners taking the vendors free, fundamental cloud courses AWS Training and Certification in May 2020 versus May 2019 jumped 152%, Williams shares. Were seeing the cloud support accelerated growth across industries, and this has been especially true during the pandemic, Williams says. When we look back, well see that the pandemic accelerated cloud adoption.

The impending shortage of STEM skills poses the greatest risk to post-Covid growth, Williams says. There is a shortage of talent with the necessary cloud skills that is leaving many technical roles unfilled and hampering businesses digital transformation objectives. There is no way around this its incumbent for enterprises to invest in training their own talent to tackle this skills shortage, she adds. Amazons own investment in close to $1 billion illustrates its efforts to address potential skill shortages.

Along with sharpening technology skills, the drive to re-invent extends to business innovation. What weve seen over the past year is a wellspring of creativity and reinvention, driven by necessity, says Williams. The greatest barrier to reinvention is often how difficult it can be to take apart something youve worked hard to build. The pandemic forced the issue for many. A key feature of businesses that survived and flourished was innovation.

Williams sees tremendous opportunity to integrate that innovation into the culture of organizations. Looking forward, leaders need to be intentional about creating and nurturing an environment that places employee skills, learning, and experimentation at the forefront.

She adds that its hard to think of an industry that is not evolving, and the pace of innovation seems to be accelerating. The area that has been very interesting to watch and experience is the restaurant industry. Integrated technology solutions helped businesses get their restaurants online, connect with delivery services, and run outdoor operations with minimal staff via QR code menus, contactless payments, and the scale of the cloud.

As the world re-opens, these changes are helping restaurants cope with limited labor availability and with their bottom line, says Williams. I suspect that like the profound changes we saw in oil-field staffing models after the Great Recession, many of these shifts in restaurant business models are here to stay. That means that a lot of workers from the food service industry will need to upskill.

Originally posted here:

Artificial Intelligence And Automations Paradox: More Human Talent Needed To Reduce Need For Human Talent - Forbes

Any smartphone owner or Google user is already intimately connected with artificial intelligence, but knowing what that means is a different matter. AIs ubiquity has not yet translated to a corresponding understanding of what and how this revolutionary technology system works, according to one of the pioneers in the industry.

I think the challenge for us is it's both everywhere and it's kind of receding into the background and people are not necessarily aware, Sir Nigel Shadbolt, one of the UKs pre-eminent computer scientists tells The National from his home in Oxford.

AI is a totally pervasive technology. It literally has become a new utility. We don't recognise it that way but the supercomputers we carry around in our pockets - our mobile phones - are running all sorts of AI-inspired and directly AI implemented algorithms to recognise your voice or recognise a photo of a face in a photo you've just taken and label it, or when it's reaching back into the cloud services to decide what to recommend to you, or how to route you efficiently to your next meeting. These things are all running."

The professor in computer science at Oxford University likens our relationship with AI to that with electricity: were highly dependent on it without a full understanding of the complex engineering feats behind a power grid.

Mainstream AI is a process of combining datasets and algorithms, or rules, to develop predictive patterns based on the data provided. To the purist, AI is a machine or algorithm which can perform tasks that would ordinarily require human intelligence.

AI is used for geographical navigation, Google searches, video-gaming and inventory management. Perhaps most universally, AI is used as recommender systems in social media platforms, on-demand video streaming services and online shopping platforms to tailor content and suggestions for users according to historical preferences.

The more information is gathered the more machine learning accelerates.

There is a duty for us to explain fundamentally what the basic principles are and what the issues are from the point of view, of safety, of fairness, of equity, availability of access, these have a moral dimension to them, says Sir Shadbolt.

For many people, artificial intelligence conjures up images of robotic humanoids or complex technology used by big tech giants to influence us. While this may be accurate in part, the fundamental misperceptions are widespread.

I sometimes reflect on the fact we might be moving back to almost an animistic culture where we imagine there's kind of a magic in our devices we don't need to worry about, Mr Shadbolt tells The National.

He has worked alongside Sir Tim Berners-Lee, inventor of the worldwide web, since 2009 and in 2012 the duo both went on to set up the Open Data Institute, which works with companies and governments to build an open, trustworthy data ecosystem.

Data is kind of an infrastructure just like your roads and your power grid but you can't see it, it's invisible in a certain sense. But you know it's important and building that kind of infrastructure is hugely important, says Mr Shadbolt, who was knighted in 2013 for his services to science and engineering.

Since the ODI was established, many national governments, regional authorities and public and private companies have gone on to publish their data online. In some countries, like France, the commitment to open public data is now enshrined in law.

The pandemic naturally pushed to the fore the importance of data, from the UK government's dashboard on hospital admissions rates to its track and trace system, information gathering and sharing was paramount in overcoming the virus.

With such pervasive influence on our lives, Mr Shadbolt says there is a growing renaissance of interest in the field of ethics and AI.

Civil rights groups have called for the banning of facial recognition software for fears that the system encroaches on privacy through mass surveillance as well as reinforcing racial discrimination. There are also other concerns that these complex learning models can be fooled.

Earlier this year a new Institute for Ethics in AI was created in Oxford University with Mr Shadbolt as its chair. He says the institutes aim is to examine the fairness and transparency of AIs many uses so that its power is used to empower and not oppress us.

The algorithms and the data of scale can be really transformational. But on the other hand, we need to reflect on the fact that there'll be two questions we've been talking about, about just how is that data used? And is it fair representation and have has the population consented?

Co-author of The Digital Ape: How to live (in peace) with smart machines, Mr Shadbolt says it is an ongoing conversation with science technologist and engineers on the one hand and legislators and ethicists on the other. Because these things at the end of the day, express our values, what we think are important to seek to preserve in the societies we build, he points out.

The Facebook-Cambridge Analytica scandal and the numerous online data breaches of other companies, has undoubtedly contributed to increasing public awareness over the perils of handing over personal information. A recent study by Penn State University researchers in the US suggests that users can become more willing to give over information when AIs offer or ask for help from users.

Nevertheless, fears around the uses of AI extend beyond its access to personal data to forecasting what a truly intelligent machine might be capable of. Scientists at the Center for Humans and Machines at the Max Planck Institute for Human Development in Berlin recently said that human control of any super-intelligent AI would be impossible.

AI has been steadily developing since the days of World War II and the code-breaking Turing Machine. It took a major leap forward in 1996 when the world chess champion, Garry Kasparov, said he could smell a new kind of intelligence across the table from the IBM supercomputer, Deep Blue.

Companies that are more open to adopting AI are likely to do better

David Egan

Mr Kasparov's defeat is often held up as a symbolic turning point in AI catching up with human intelligence. Nineteen years later, the power of AI made an exponential leap forward when AlphaGo became the first computer program to defeat a professional human player at Go, the incredibly complex and challenging 3000-year-old Chinese game.

The pandemic has accelerated the adoption of AI across sectors, particularly in healthcare and pushed it more towards becoming a necessity. In England AI systems were used to screen patients lung scans for Covid-19 and to sift through hundreds of research papers being published on the new virus.

AI received a battlefield promotion as the crisis forced the pace of innovation and adoption, said David Egan, a senior analyst at Columbia Threadneedle Investment, at a recent forum to discuss investor opportunities in the field.

Companies that are more open to adopting AI are likely to do better and the benefit to those companies will compound at an exponential rate each year.

Having surveyed the field for decades, Mr Shadbolt thinks we are now at the time to take hold of this "great opportunity" while also taking stock of the "bigger questions".

Technical development has to go hand in hand with an appreciation of our values, why we're doing this, what kind of society we want to build, where we want decision making to reside, where the value of all this insight actually ends up landing.

Read the rest here:

Artificial intelligence is already everywhere, we need to adapt - The National

Its interesting that youre talking about how, in these kinds of scenarios, you have to actively design distrust into the system to make it more safe.

Yes, thats what you have to do. Were actually trying an experiment right now around the idea of denial of service. We dont have results yet, and were wrestling with some ethical concerns. Because once we talk about it and publish the results, well have to explain why sometimes you may not want to give AI the ability to deny a service either. How do you remove service if someone really needs it?

But heres an example with the Tesla distrust thing. Denial of service would be: I create a profile of your trust, which I can do based on how many times you deactivated or disengaged from holding the wheel. Given those profiles of disengagement, I can then model at what point you are fully in this trust state. We have done this, not with Tesla data, but our own data. And at a certain point, the next time you come into the car, youd get a denial of service. You do not have access to the system for X time period.

Its almost like when you punish a teenager by taking away their phone. You know that teenagers will not do whatever it is that you didnt want them to do if you link it to their communication modality.

The other methodology weve explored is roughly called explainable AI, where the system provides an explanation with respect to some of its risks or uncertainties. Because all of these systems have uncertaintynone of them are 100%. And a system knows when its uncertain. So it could provide that as information in a way a human can understand, so people will change their behavior.

As an example, say Im a self-driving car, and I have all my map information, and I know certain intersections are more accident prone than others. As we get close to one of them, I would say, Were approaching an intersection where 10 people died last year. You explain it in a way where it makes someone go, Oh, wait, maybe I should be more aware.

The negatives are really linked to bias. Thats why I always talk about bias and trust interchangeably. Because if Im overtrusting these systems and these systems are making decisions that have different outcomes for different groups of individualssay, a medical diagnosis system has differences between women versus menwere now creating systems that augment the inequities we currently have. Thats a problem. And when you link it to things that are tied to health or transportation, both of which can lead to life-or-death situations, a bad decision can actually lead to something you cant recover from. So we really have to fix it.

The positives are that automated systems are better than people in general. I think they can be even better, but I personally would rather interact with an AI system in some situations than certain humans in other situations. Like, I know it has some issues, but give me the AI. Give me the robot. They have more data; they are more accurate. Especially if you have a novice person. Its a better outcome. It just might be that the outcome isnt equal.

Its important to me because I can identify times in my life where someone basically provided me access to engineering and computer science. I didnt even know it was a thing. And thats really why later on, I never had a problem with knowing that I could do it. And so I always felt that it was just my responsibility to do the same thing for those who have done it for me. As I got older as well, I noticed that there were a lot of people that didnt look like me in the room. So I realized: Wait, theres definitely a problem here, because people just dont have the role models, they dont have access, they dont even know this is a thing.

And why its important to the field is because everyone has a difference of experience. Just like Id been thinking about human-robot interaction before it was even a thing. It wasnt because I was brilliant. It was because I looked at the problem in a different way. And when Im talking to someone who has a different viewpoint, its like, Oh, lets try to combine and figure out the best of both worlds.

Airbags kill more women and kids. Why is that? Well, Im going to say that its because someone wasnt in the room to say, Hey, why dont we test this on women in the front seat? Theres a bunch of problems that have killed or been hazardous to certain groups of people. And I would claim that if you go back, its because you didnt have enough people who could say Hey, have you thought about this? because theyre talking from their own experience and from their environment and their community.

If you think about coding and programming, pretty much everyone can do it. There are so many organizations now like Code.org. The resources and tools are there. I would love to have a conversation with a student one day where I ask, Do you know about AI and machine learning? and they say, Dr. H, Ive been doing that since the third grade! I want to be shocked like that, because that would be wonderful. Of course, then Id have to think about what is my next job, but thats a whole other story.

But I think when you have the tools with coding and AI and machine learning, you can create your own jobs, you can create your own future, you can create your own solution. That would be my dream.

See original here:

We need to design distrust into AI systems to make them safer - MIT Technology Review

Simulations of a region of space 100 million light-years square. The leftmost simulation ran at low resolution. Using machine learning, researchers upscaled the low-res model to create a high-resolution simulation (right). That simulation captures the same details as a conventional high-res model (middle) while requiring significantly fewer computational resources. Credit: Y. Li et al./Proceedings of the National Academy of Sciences 2021

Using neural networks, researchers can now simulate universes in a fraction of the time, advancing the future of physics research.

A universe evolves over billions upon billions of years, but researchers have developed a way to create a complex simulated universe in less than a day. The technique, recently published in the journal Proceedings of the National Academy of Sciences, brings together machine learning, high-performance computing, and astrophysics and will help to usher in a new era of high-resolution cosmology simulations.

Cosmological simulations are an essential part of teasing out the many mysteries of the universe, including those of dark matter and dark energy. But until now, researchers faced the common conundrum of not being able to have it all simulations could focus on a small area at high resolution, or they could encompass a large volume of the universe at low resolution.

Carnegie Mellon University Physics Professors Tiziana Di Matteo and Rupert Croft, Flatiron Institute Research Fellow Yin Li, Carnegie Mellon Ph.D. candidate Yueying Ni, University of California Riverside Professor of Physics and Astronomy Simeon Bird and University of California Berkeleys Yu Feng surmounted this problem by teaching a machine learning algorithm based on neural networks to upgrade a simulation from low resolution to super resolution.

Cosmological simulations need to cover a large volume for cosmological studies, while also requiring high resolution to resolve the small-scale galaxy formation physics, which would incur daunting computational challenges. Our technique can be used as a powerful and promising tool to match those two requirements simultaneously by modeling the small-scale galaxy formation physics in large cosmological volumes, said Ni, who performed the training of the model, built the pipeline for testing and validation, analyzed the data and made the visualization from the data.

The trained code can take full-scale, low-resolution models and generate super-resolution simulations that contain up to 512 times as many particles. For a region in the universe roughly 500 million light-years across containing 134 million particles, existing methods would require 560 hours to churn out a high-resolution simulation using a single processing core. With the new approach, the researchers need only 36 minutes.

The results were even more dramatic when more particles were added to the simulation. For a universe 1,000 times as large with 134 billion particles, the researchers new method took 16 hours on a single graphics processing unit. Using current methods, a simulation of this size and resolution would take a dedicated supercomputer months to complete.

Reducing the time it takes to run cosmological simulations holds the potential of providing major advances in numerical cosmology and astrophysics, said Di Matteo. Cosmological simulations follow the history and fate of the universe, all the way to the formation of all galaxies and their black holes.

Scientists use cosmological simulations to predict how the universe would look in various scenarios, such as if the dark energy pulling the universe apart varied over time. Telescope observations then confirm whether the simulations predictions match reality.

With our previous simulations, we showed that we could simulate the universe to discover new and interesting physics, but only at small or low-res scales, said Croft. By incorporating machine learning, the technology is able to catch up with our ideas.

Di Matteo, Croft and Ni are part of Carnegie Mellons National Science Foundation (NSF) Planning Institute for Artificial Intelligence in Physics, which supported this work, and members of Carnegie Mellons McWilliams Center for Cosmology.

The universe is the biggest data sets there is artificial intelligence is the key to understanding the universe and revealing new physics, said Scott Dodelson, professor and head of the department of physics at Carnegie Mellon University and director of the NSF Planning Institute. This research illustrates how the NSF Planning Institute for Artificial Intelligence will advance physics through artificial intelligence, machine learning, statistics, and data science.

Its clear that AI is having a big effect on many areas of science, including physics and astronomy,said James Shank, a program director in NSFs Division of Physics. Our AI planning Institute program is working to push AI to accelerate discovery. This new result is a good example of how AI is transforming cosmology.

To create their new method, Ni and Li harnessed these fields to create a code that uses neural networks to predict how gravity moves dark matter around over time. The networks take training data, run calculations and compare the results to the expected outcome. With further training, the networks adapt and become more accurate.

The specific approach used by the researchers, called a generative adversarial network, pits two neural networks against each other. One network takes low-resolution simulations of the universe and uses them to generate high-resolution models. The other network tries to tell those simulations apart from ones made by conventional methods. Over time, both neural networks get better and better until, ultimately, the simulation generator wins out and creates fast simulations that look just like the slow conventional ones.

We couldnt get it to work for two years, Li said, and suddenly it started working. We got beautiful results that matched what we expected. We even did some blind tests ourselves, and most of us couldnt tell which one was real and which one was fake.

Despite only being trained using small areas of space, the neural networks accurately replicated the large-scale structures that only appear in enormous simulations.

The simulations didnt capture everything, though. Because they focused on dark matter and gravity, smaller-scale phenomena such as star formation, supernovae and the effects of black holes were left out. The researchers plan to extend their methods to include the forces responsible for such phenomena, and to run their neural networks on the fly alongside conventional simulations to improve accuracy.

Read AI Magic Just Removed One of the Biggest Roadblocks in Astrophysics for more on this research.

Reference: AI-assisted superresolution cosmological simulations by Yin Li, Yueying Ni, Rupert A. C. Croft, Tiziana Di Matteo, Simeon Bird and Yu Feng, 4 May 2021, Proceedings of the National Academy of Sciences.DOI: 10.1073/pnas.2022038118

The research was powered by the Frontera supercomputer at the Texas Advanced Computing Center (TACC), the fastest academic supercomputer in the world. The team is one of the largest users of this massive computing resource, which is funded by the NSF Office of Advanced Cyberinfrastructure.

This research was funded by the NSF, the NSF AI Institute: Physics of the Future and NASA.

Go here to see the original:

AI Technique Ushers In New Era of High-Resolution Simulations of the Universe - SciTechDaily