New Google-developed AI could help in diagnosing breast cancer

(January 2) Google is developing artificial Intelligence (AI) that will help doctors identify breast cancer according to a research paper published in Nature Today. The AI model which scans X-ray images known as mammograms reduces false negatives by 9.4 percent.

The current tests miss 20 percent of breast cancers. While as reported by the Wall Street Journal the AI does better than doctors, it nevertheless misses some cancers that radiologists find: "Google’s health research unit said it has developed an artificial-intelligence system that can match or outperform radiologists at detecting breast cancer, according to new research. But doctors still beat the machines in some cases."

Mammograms effective but have problems

Breast cancer is the second leading cause of cancer deaths in women, only lung cancer is more deadly and prevalent. Even though mammagrams are an effective common detection tool they nevertheless miss a large number of positive cases. Shravya Shetty a Google researcher and co-author of the paper said: “Mammograms are very effective but there’s still a significant problem with false negatives and false positives."

The study

The Google-financed study used anonymized mammograms from more than 25,000 women from both the UK and another 3,000 from the US. Shetty said that researchers attempted to follow the same principles as those of radiologists. The research team first trained AI to scan X-ray imagines to look for signs of breast cancer by identifying changes to the breasts of the 28,000 women and they then checked the AI program's analyses against that of the actual medical outcomes of the women.

The results

The researchers were ultimately able to reduce false negatives by 9.4 percent and false positives by 5.7 percent in the US. In the UK where two radiologists typically double-check the results the results were still positive but not by nearly the same degree. False negatives were cut by 2.7 percent but false positives by only 1.2 percent.

A recent Wired article notes: "The study claims the DeepMind algorithm performs better than a single radiologist, and is "non inferior" versus two."The model performs better than an individual radiologist in both the UK and the US," Kelly says. "In the UK we have this double reading system, where two radiologists or maybe three or four look at each scan… we're statistically the same as that, but not better than that." " However, the article notes that the UK is facing a shortage of radiologists.

The AI program called Deep Mind was not perfect. In some cases radiologists flagged cases which the AI system missed even though Deep Mind outperformed examinations by single cardiologists,

Google hopes AI system can be used in clinical settings

Daniel Tse a Google product manager and co-author of the paper said the research team was working to make sure the findings could be generalized across populations: “We’re very excited and encouraged by these results. There’s obviously quite a bit of nuance when you put this into clinical practice."

Shetty said that the AI system would help radiologists rather than replace them: “They each bring their own strength, it’s complementary. There are a number of cases where the radiologists catch something that the model misses, and vice versa. Bringing the two together could strengthen the overall results.”

Previously published in the Digital Journal

Amazon using Artificial Intelligence to make decisions on worker firings

Amazon's huge warehouses, called fulfillment centers are the engine of the company's retail business as it is where workers track, pack. sort, and shuffle each order before it is sent to the buyer.

Workers pressured to meet targets or be fired

An article in The Verge in April last year noted: "Amazon warehouse workers are forced to pee in bottles or forego their bathroom breaks entirely because fulfillment demands are too high, according to journalist James Bloodworth, who went undercover as an Amazon worker for his book, Hired: Six Months Undercover in Low-Wage Britain. Targets have reportedly increased exponentially, workers say in a new survey revealed over the weekend, and as result, they feel pressured and stressed to meet the new goals."

Workers are pressured to "make rate" with some packing hundreds of boxes per hour. If they do not work fast enough they lose their jobs. Stacy Mitchell, a staunch Amazon critic, said: "You've always got somebody right behind you who's ready to take your job." Mitchell also said: "One of the things that we hear consistently from workers is that they are treated like robots in effect because they're monitored and supervised by these automated systems."

Amazon's automated system

Amazon uses an automated system that tracks "time off task"(TOT) that tracks the productivity of each worker. The AI system automatically gathers information about the individual worker's productivity and can generate warnings or even firings without input from supervisors according to a letter obtained by Verge. The automatic firing can happen if the worker is pausing or taking breaks too often. However, a supervisor can override the system and there is also an appeal process by which a worker can attempt to get his or her job back.

Union criticizes system

The president of the United Food and Commercial Workers International Union (UFCW) said:"It's one thing for Jeff Bezos and Amazon to use a ruthless business model to destroy jobs for profit, but it is surreal to think that any company could fire their own workers without any human involvement. You’ve always got somebody right behind you who’s ready to take your job.”

Productivity firings at Amazon are common

In a signed letter last year, a lawyer representing Amazon reported that the company fired hundreds of employees at a single warehouse in Baltimore between August of 2017 and September 2018 for failure to meet production quotas. A company spokesperson confirmed that over 300 full-time workers , called associates by the company, were fired for inefficiency during that period. This is a substantial percentage of the workers in the facility who number about 2,500 full-time employees today.

An Amazon spokesperson said: ”Approximately 300 employees turned over in Baltimore related to productivity in this timeframe. In general, the number of employee terminations have decreased over the last two years at this facility as well as across North America.” Amazon gave no details of the present firing rate.

While Amazon touts the benefits of working for the company such as parental leave it seems that the company makes strenuous demands to reach productivity targets or workers will find themselves fired. The appended video discusses the Verge article upon which much of this article is based.

Previously published in Digital Journal

US legislators introduce bill to audit algorithms for bias

US legislators introduced a bill that requires large companies to audit machine learning powered systems such as facial recognition or ad targeting algorithms to check for bias.

	1 of 2

The Algorithmic Accountability Act

The act is sponsored by Senators Cory Booker a Democrat and Ron Wyden a Republican. In the House an equivalent bill is sponsored by Representative Yvette Clarke a Democrat. If the bill passes it would ask that the Federal Trade Commission create rules for evaluation of highly sensitive automated systems. The companies that use such systems would need to check the algorithms powering the systems to determine whether they are biased or discriminatory. They would also need to determine if the systems pose a threat to privacy or the security of individuals.

The act targets companies that access large amounts of information, making over $50 million per year, hold information on at least 1 million people or devices, or primarily act as data brokers who buy and sell consumer data. The companies would be required to examine a large range of algorithms, including those that could affect a consumer's legal rights, that attempt to predict or analyze consumer behavior, involve large amounts of sensitive information or systematically monitor a large publicly accessible physical place. Theoretically, this would include a huge area of the tech economy. If a report should uncover major risks of discrimination, privacy problems of other issues, the company involved is to address them within a reasonable time.

Facebook sued for bias

The bill comes just a few short weeks after Facebook was sued by the US Department of Housing and Urban Development. The Department had alleged its algorithms resulted in a targeting system which unfairly limited who saw housing ads. The sponsors mentioned this lawsuit in a press release.

The suit is reported in a recent article: The Department of Housing and Urban Development announced Thursday it is suing Facebook for violating the Fair Housing Act by allowing advertisers to limit housing ads based on race, gender and other characteristics. The agency also said Facebook’s ad system discriminates against users even when advertisers did not choose to do so. ProPublica first reported in 2016 that Facebook allowed housing advertisers to exclude users by race. Then in 2017, ProPublica found that—despite Facebook’s promised changes—the company was still letting advertisers exclude users by race, gender, ethnicity, family status, ability and other characteristics."

Bill covers many controversial AI areas

The bill would also include training data that could also result in biased outcomes. An example would be a facial recognition pattern trained on mostly white subjects can as a result misidentify people of other races.

Senator Ron Wyden noted the importance of an audit for bias noting that “computers are increasingly involved in the most important decisions affecting Americans’ lives — whether or not someone can buy a home, get a job or even go to jail. But instead of eliminating bias, too often these algorithms depend on biased assumptions or data that can actually reinforce discrimination against women and people of color.”

As reported in a Digital Journal article some time ago the academic world has already noted the importance of ethics in the field of AI: "The advances of artificial intelligence together with machine learning into greater areas of life requires a review of the ethical and human-facing implications, according to the University of Guelph. A new hub has been launched to address such issues."

Back in October of 2017 the Google AI chief expressed concerns about algorithms: "John Giannandrea AI chief at Google is concerned that bias is being built in to many of the machine-learning algorithms by which the robot makes decisions. At a recent Google conference on the relationships between AI systems and humans Giannandrea said: “The real safety question, if you want to call it that, is that if we give these systems biased data, they will be biased.”"

Previously published in the DIgital Journal

Artificial intelligence is begin used to help save bee colonies

Machine learning and related types of Artificial Intelligence(AI) are used to work to help solve various problems but it still seems surprising that it could be used to help stop the alarming decline in bee populations across the globe.

The Varroa mite

Wikipedia describes the Varrao mite as follows: " The Varroa mite can only reproduce in a honey bee colony. It attaches to the body of the bee and weakens the bee by sucking fat bodies [1]. In this process, RNA viruses such as the deformed wing virus (DWV) spread to bees. A significant mite infestation will lead to the death of a honey bee colony, usually in the late autumn through early spring. The Varroa mite is the parasite with the most pronounced economic impact on the beekeeping industry. Varroa is considered to be one of multiple stress factors[2] contributing to the higher levels of bee losses around the world."

A recent article claims that the mite rarely kills a bee outright but weakens the bee by sucking blood and weakening it making it susceptible to diseases and causes young to be born weak and deformed. In time this can lead to colony collapse. One problem is that you may not even see the mites as they are only a millimeter or so across. An infestation may not be discovered for some time.

As shown on the appended video beekeepers put a flat surface beneath the hive and pull it out to inspect it to find tiny bodies of the mites. It is painstaking and time-consuming work.

How machine learning can help

Machine learning models are good at sorting through data that is "noisy" such as the flat surface with the varroa mites on it but covered in all sorts of other debris. The machine can be taught to identify the shape of the mites, and count them.

Apizoom

Students in Switzerland at the Ecole Polytechnique Federale at Lausanne(EPFL) have created an image recognition device named ApiZoom. When trained on images of mites through a photo, the device can recognize any visible mite bodies in seconds. All a beekeeper has to do is take a photo with a smartphone and upload it to the EPFL system.

The EPFL project was begun back in 2017. The model has been trained with tens of thousands of images that have made it progressively better at its job. The success rate of detection is now about 90 percent about the same as humans achieve. The project now intends to distribute the app as widely as it can.

Alain Burgnon of the EPFL project said: “We envisage two phases: a web solution, then a smartphone solution. These two solutions allow to estimate the rate of infestation of a hive, but if the application is used on a large scale, of a region. By collecting automatic and comprehensive data, it is not impossible to make new findings about a region or atypical practices of a beekeeper, and also possible mutations of the Varroa mites.”

This kind of systematic data collection would be a major help for coordinating response to infestations at a national level. No doubt ApiZoom could be used globally not just in Switzerland. Apizoom is being spun off as a separate company by Bugnon.

There are many ways of dealing with an infestation as described in the Wikipedia article on the virus. Some bee types are resistant to the mite and perhaps more of those bees will be used to produce honey. However, the Apizoom app will no doubt help out to control the mite.

Previously published in Digital Journal

More Americans approve of AI rather than disapprove according to survey

Artificial Intelligence is likely to be one of the defining technologies of this century. It will no doubt affect everything from warfare to health care to jobs. A recent survey gauges attitudes within the U.S. public to AI.

	1 of 2

While more Americans were in favor of AI than opposed, opinions were mixed and varied from one group to another.

The survey

The report describes itself as follows: "This report is based on findings from a nationally representative survey conducted by the Center for the Governance of AI, housed at the Future of Humanity Institute, University of Oxford, using the survey firm YouGov. The survey was conducted between June 6 and 14, 2018, with a total of 2,000 American adults (18+) completing the survey. "

The authors claim they were more interested in breadth than depth in the study. Questions touch on numerous issues such as workplace automation, international cooperation, public trust in various actors to develop and regulate AI and many others. While the results give some preliminary insights into the nature of U.S. opinion on AI, the study raises more questions than it answers. More work is needed to gain a deeper understanding of U.S. public opinion to AI, and the authors recommend caution in interpreting the results.

General attitude to AI

The survey described AI as "computer systems that perform tasks or make decision that usually require human intelligence". 41 percent of respondents said they somewhat or strongly supported the development of AI. On the other hand 22 percent said they somewhat or strongly opposed it. However, 28 percent said they had no strong feeling one way or the other.

Relation of attitude to demographics

Young, educated, and male individuals were most likely to favor AI development. 57 percent of college graduates were in favor of AI but in contrast only 29 percent of those with high school education or less were in favor of AI development.

On some issues there was a strong consensus among groups, for example on regulation. 82 percent of those questioned somewhat or strongly agreed with the statement "robots and artificial intelligence are technologies that require careful management." Respondents ranked data privacy as one of the most important governance challenges but also ranked cyberattacks and autonomous weapons as also quite important.

No consensus on who should limit AI development

University researchers were the most trusted with 50 percent of respondents reporting a fair amount of confidence or a great deal of confidence in them. However, the US military came a close second with 49 percent trusting them. Tech companies had a 44 percent trust level. Facebook however had a poor rating with 4 in 10 saying they had no trust in the company to regulate AI.

Other issues

On the question of whether AI should develop high-level machine intelligence with the common sense ability of a human, there were mixed answers. Fully 29 percent of respondents neither supported nor opposed the development. 31 percent said they somewhat or strongly support the development while 27 percent were opposed.

On the relative importance of AI issues as a global risk out of 15 issues, AI was judged to be relatively low. Nuclear issues or recessions were ranked much higher. However failure to act on climate change was also ranked low. Terrorist acts, infectious diseases were also ranked relatively high as possible problems in the next ten years.

Previously published in Digital Journal

Using AI the approximate time of death of patients can be predicted

Artificial Intelligence can use data from health records to predict with considerable accuracy when a patient will die. Such information can have a positive use and end up saving lives.

Developing more accurate predictive models

A recent paper published in Nature by a large number of researchers says that feeding data from electronic health records (EHR) into a deep learning model can substantially improve the accuracy of predicted outcomes.

In trials that used data from two U.S. hospitals, the researchers were able to show that the algorithms in the model improved predictions as to the length of stay and time of discharge — but also the time of death of patients.

The neural network in the study uses an immense amount of data, including information about the patient's medical history and vitals. A new algorithm lines up events in the patient's records into a timeline, and this enables the learning model to predict future outcomes such as length of stay and even death. The predictions are made in record time as well.

Uses of the predictions

The predictions may enable the hospital to prioritize patient care, or adjust treatment plans, and even recognize medical emergencies before they happen. It will also free up the time of healthcare workers who had to sift through all this data in order to make less accurate predictions.

Other recently developed algorithms can diagnose lung cancer and heart disease better than human doctors. Another study has fed retinal images into AI algorithms that are able to predict the chances that a patient could develop one or more of three major eye diseases.

One problem is to amass all the data about patients in one central database as the data is often spread widely through various healthcare systems and government agencies often without the data being shared.

Dangers of centralizing patient data

Some worry that collecting this massive amount of personal health data and putting it all into a single model that is owned by Google, one of the largest private companies in the world, could very well create problems. There would need to be guarantees that the data would not be shared. Many companies would be anxious to use the information for their purposes. As the article notes: " Electronic health records of millions of patients in the hands of a small number of private companies could quickly allow the likes of Google to exploit health industries, and become a monopoly in healthcare."

A case for caution

Recently the U.K. government was concerned that DeepMind Health, owned by Alphabet, Google's parent company, was able to exert monopoly power as described in an article in TechCrunch. There had been earlier concerns that DeepMind Health had broken U.K. laws by collecting data on patients without their consent last year.

A group who reviewed the project said: “There are many examples in the IT arena where companies lock their customers into systems that are difficult to change or replace. Such arrangements are not in the interests of the public. And we do not want to see DeepMind Health putting itself in a position where clients, such as hospitals, find themselves forced to stay with DeepMind Health even if it is no longer financially or clinically sensible to do so; we want DeepMind Health to compete on quality and price, not by entrenching legacy position."

Healthcare professions are already worried that AI could have some negative effects on medicine once it is embedded in the system, without precautions taken to ensure transparency. The American Medical Association (AMA), while admitting the benefits AI can bring to medicine, also notes that AI tools must “strive to meet several key criteria, including being transparent, standards-based, and free from bias.” It is not clear how one makes algorithms transparent. The AMA notes that the US regulatory framework is far behind the advance of technology with legislation being passed over two decades ago.

Algorithms can be biased

Algorithms developed for the AI systems can reflect bias, often unconscious, of those who develop them. Many seem to accept algorithms since they are mathematical as objective and unbiased by their very nature. However, as a recent article points out there are biased algorithms everywhere and refers to a statement by the author of the book "Weapons of Math Destruction": "Cathy O’Neil, a mathematician and the author of Weapons of Math Destruction, a book that highlights the risk of algorithmic bias in many contexts, says people are often too willing to trust in mathematical models because they believe it will remove human bias. “[Algorithms] replace human processes, but they’re not held to the same standards,” she says. “People trust them too much.”

Previously published in Digital Journal

Microsoft reorganizes to concentrate upon Artificial Intelligence

Microsoft announced another large reorganization today as the company will unify its Artificial Intelligence (AI) and core Windows Operating System into just one team. Terry Myerson, head of the Windows team and a 21-year company veteran is leaving.

However, overall Windows will be split into two teams. One team will be headed by Scott Guthrie who will be in charge of what is called the platform team. The AI platform work will also be part of this team's work. The second team will be headed by Harry Shum who will lead an engineering team AI + Research that will focus on tech advances that could be used in future Microsoft products.

New separate division

The new "Experiences and Devices" division will deal with Windows client releases that are seen on laptops, first party apps, Office 365, and Surface hardware. Rajesh Jha will lead this team. The team wont handle the core parts of the Windows platform but the experiences added to it. This is part of Microsoft's vision of its future: clever modes running on clever hardware.

CEO sees a bright future for Microsoft

Microsoft CEO Satya Nadella claims that Microsoft has a bright future. Nadella appears to be focusing the company core Windows effort, Azure and Internet of Thins (IoT) devices into a team that would embody Artificial Intelligence into them.

Microsoft Azure is described by Wikipedia as follows: "Microsoft Azure (formerly Windows Azure) /ˈæʒər/ is a cloud computing service created by Microsoft for building, testing, deploying, and managing applications and services through a global network of Microsoft-managed data centers. It provides software as a service (SaaS), platform as a service (PaaS) and infrastructure as a service (IaaS) and supports many different programming languages, tools and frameworks, including both Microsoft-specific and third-party software and systems." Azure lists over 600 services that it provides.

Microsoft is also creating a new AI and Ethics in Engineering and Research Committee (AETHER) to handle AI issues in what the company terms " a responsible manner".

The company is to emphasize its future AI and cloud business, enterprise services and AI. The company will need to balance out this new focus on the Cloud with consumer's and enterprises' operating system needs.

CEO Satya Nadella set out the company focus in a memo to employees. Part of the memo reads: "Over the past year, we have shared our vision for how the intelligent cloud and intelligent edge will shape the next phase of innovation. First, computing is more powerful and ubiquitous from the cloud to the edge. Second, AI capabilities are rapidly advancing across perception and cognition fueled by data and knowledge of the world. Third, physical and virtual worlds are coming together to create richer experiences that understand the context surrounding people, the things they use, the places they go, and their activities and relationships."

Intelligent edge

Techopedia explains the intelligent edge as follows: "With an intelligent edge, remote or decentralized nodes of a system are empowered to do different kinds of data handling that may have traditionally been handled at a central point in a system. Specifically with IoT, a classical model of routing all of the many streams of data from IoT-connected devices into a central data warehouse or repository has several distinct disadvantages. It may be inefficient, and, if the data is not encrypted, it can also leave the system inherently more vulnerable.

In an intelligent edge setup, the edge network components or nodes can process the data intelligently, possibly bundling, refining or encrypting it for transit into the data warehouse. This can improve the agility of data-handling systems, as well as their safety. Many cloud providers and other companies knowledgeable about the structure and nature of IoT are recommending the use of an intelligent edge for these reasons."

Intelligent cloud

The intelligent cloud is discussed in great detail in a recent Forbes article. The article claims that the Intelligent Cloud is about getting more insight that will lead to more favorable business outcomes and says: "Line-of-business leaders across all industries want more from their cloud apps than they are getting today. They want the ability to gain greater insights with prescriptive and cognitive analytics. They’re also asking for new apps that give them the flexibility of changing selling behaviors quickly. In short, everyone wants to get to the orchestration layer of the maturity model, and many are stuck staring into a figurative rearview mirror, using just descriptive data to plan future strategies. The future of enterprise cloud computing is all about being able to deliver prescriptive and cognitive intelligence."

Recent updates to Windows 10

There were two big updates to Windows last year that concentrated on Mixed Reality, creating in 3D pain, and pen/touch improvements. Universal Windows Apps were supposed to be the future of Windows with the apps working across many devices. It looks now as if the future lies in its Cloud and AI efforts while the company also tries to adapt to a future in which PCs are much different than now.

Terry Merson's leaving memo to the Windows division

Merson's full memo can be read here. He supports the vision and new orientation of Microsoft under the leadership of Satya Nadella saying: "Satya’s leadership and insight in defining a Microsoft 365 experience, built on top of an intelligent edge/intelligent cloud platform is inspiring. I believe in it, and that these changes are great for Microsoft. Change can be invigorating for us all and I’m grateful I had the opportunity to work with Satya on helping define this new structure. I will be around as we work through this transition, and then I will continue to root on Satya and this team every day."

Microsoft Office 365 is the brand name Microsoft uses for a group of subscriptions that provide productivity software and related services.

Previously published in Digital Journal

Artificial Intelligence being used to study chicken communication

Researchers are using machine learning technology to understand and analyze patterns in chicken speech. The scientists hope to better inform farmers about the health and happiness of their poultry.

New technologies continue to revolutionize the business of farming — from self-driving tractors to IoT sensors monitoring crops. However, scientists at the University of Georgia University and Georgia Institute of Technology are taking a different approach to monitoring the health of farm animals.

Over the last five years, both engineers and poultry scientists have been studying chicken "language" at the two institutions hoping to be able to better use the information gleaned from their studies to help poultry farmers such as Mitchell.

Chickens are chatty creatures

Kevin Mitchell of Wilcox Farms, with properties in both Washington State and Oregon, notes that chickens make sounds that exhibit "patterns of speech" that reveal a considerable amount of information about their well-being.

Mitchell says chickens usually make the most noise in the morning with many clucks, chortles, and caws. When he hears these sounds, Mitchell claims he knows they are reasonably healthy and happy.

In the evening when they’re preparing to roost Mitchell claims, the chickens are much more mellow, cooing softly.

When a hen lays an egg it issues a series of clucks, like drumbeats that culminate in a loud "buck-caw" as heard on one of the appended videos.

When chickens detect an aerial predator approaching they produce a short, high-pitched shriek as also heard on one of the appended videos.

The repetitive clucking often heard when you approach chickens is in fact an alarm call as well indicating a predator may be approaching on the ground.

Studies at the University of Georgia and Georgia Institute of Technology

Georgia Tech research engineer Wayne Daley and his colleagues studied the effects on 6 to 12 broiler chickens of moderately stressful conditions and recorded their vocalizations. The conditions included increased ammonia levels in the air, minor infections and higher temperatures.

The resulting audio data was fed into an AI learning program that enabled it to learn the difference between contented and distressed birds through their vocalizations.

What the AI has learned

The program can now detect when birds are uncomfortable due to heat stress on the basis of the sounds they are making.

It can also detect when the birds have a respiratory infection because of the sound they make when mucus clogs their airways.

Daley notes: “A lot of poultry farmers we have worked with say they can hear when something is wrong with a flock, but they can’t tell us exactly how they know that. There’s a lot of subtlety. We’re learning that there are changes in the frequency of the sounds and the levels—the amplitude or loudness—that the machines can pick up on.”

Chicken "language" may be more complex than previously thought

Carolynn Smith a biologist at Macquarie University in Australia, a leading expert on chicken vocalizations, together with other scientists have discovered that chicken communications appear to be more complicated and nuanced than previously thought.

For example, roosters do not automatically shriek every time they see an aerial predator. After all, the shriek makes them more vulnerable in that the predator can hear it. If there are females around they usually do sound the alarm. However, if they are alone or surrounded by other males they often stay quiet. They also shriek more often when there is cover nearby where they can be safe.

Technology in commercial chicken farms a challenge for AI

One of the problems facing researchers is that large scale commercial chicken farms are very noisy as they have large fans meant to circulate air as well as large heater fans.

Given the background noise it becomes difficult for the AI software to simply concentrate on the sounds of the chickens. The challenge will be to have the software filter out the loud background noises.

When this is solved, the AI software can simply be added to the existing advanced technology which includes the ability to monitor and modify lighting, temperature, and ventilation. The farmer can also activate automatic feeding systems from their phones or computers. The audio AI data would just be one more tool to monitor the conditions of the fowl.

Previously published in Digital Journal

Many have doubts about positive effects of development of AI

Charles Modeas, EU Research Commissioner claims that the development of AI and thinking machines pose a threat to our existence risk confusing science with science fiction.

AI development predominate

In a speech to the European Parliament's Science and Technology Options Assessment Group Modeas said:“If you do any research on artificial intelligence these days, the results are astonishingly pessimistic. Nine articles out of ten on AI are negative. Not just negative. Alarmist and panicked, sometimes even hysterical. For me, a techno-optimist, it's shocking. And very disappointing."

Commissioner Modeas is betting that AI research will be a positive force even though he admits that public fear of the technology appears to be deep. He thinks the public fears what is the most exciting new technology for our generation. To deny the amazing benefits it can bring is not the answer he claims.

Possible negative effects of AI development

Warnings about AI development have come from notables such as Bill Gates, Stephen Hawking, and Elon Musk who argue that it could evolve to a point where it is beyond human control. Many critics also worry that robots with AI will take over jobs creating unemployment.

Elon Musk's worries

Musk who himself has been a prime contributor to technology both in electric vehicles and space rockets worries that competition for AI technology could lead to war as governments compete for superiority in weaponry using AI. However, it has surely always been the case that countries compete for technology especially technology that can be triumphant in warfare. This would be so whether AI developed or not. However, the development of AI will make this competition more dangerous..

Technology in the form of chemical weapons and nuclear bombs already show that it is imperative that we do everything we can to ensure that new AI technology is controlled. Musk's warning are very much based on reality.

Will AI technology result in lost jobs?

Another of Musk's worries was the loss of jobs.

Economists Daron Acemoglu and Pascual Restrepo of the National Bureau of Economic Research looked at the historical effects of robots on employment in the US between 1990 and 2007 controlling for the influence of other factors.

Their study showed that each new robot led to the loss of between 3 and 5.6 jobs in the local area. For each new robot added for 1,000 workers wages would also decline between 0.25 and 0.5 percent.

The two researchers write: “Predictably, the major categories experiencing substantial declines are routine manual occupations, blue-collar workers, operators and assembly workers, and machinists and transport workers.”

Steven Mnuchin, the US treasury secretary said that he was not worried about the effects of AI and automation on employment. Mnuchin said: “Quite frankly, I'm optimistic. I mean, that's what creates productivity." The productivity is increased because more is produced by each worker in conjunction with AI. However, the remaining workers will actually see their wages decline. The AI makes the firm more productive and usually more profitable as well.

As more and more AI is used in production and elsewhere its negative effects on employment will likely be greater.

Hawking's fears

Stephen Hawking is a famous theoretical physicist and cosmologist. He suffers from a slow progressing but early onset type of ALS sometimes called Lou Gehrig's disease. Wikipedia notes: "Hawking has a rare early-onset, slow-progressing form of amyotrophic lateral sclerosis (ALS) that has gradually paralysed him over the decades. He now communicates using a single cheek muscle attached to a speech-generating device."

Hawking is not averse to advanced technology. He uses such technology to communicate and could not talk without it.

Hawking's warning is dramatic. He has even said that the development of full AI could result in the end of the human race.

The new AI used by robots is based on biology as a model rather than mathematics. New robots learn from their experiences just as human beings do. The older model of robots simply programmed them to do specific tasks and their actions were always predictable if properly programmed. With the new robots that learn from experiences they are less predictable.

Hawkins warns that robots are being developed that can match or surpass human abilities and some could take off on their own and redesign themselves quickly. The computer HAL in Kubrick's film 2001 perhaps provides a good science fiction version of what Hawking sees as possible.

Humans Hawking feels are limited by slow biological evolution whereas AI robots could develop much more quickly. Humans would not be able to compete with them. It is not clear how the robots would keep humans from shutting them down. Even HAL in Kubrick's film eventually gets shut down.

Moedas' worries

Moedas does not take these concerns to be pressing issues. As AI develops there will be plenty of time to address these issues he argues.

What Moedas worries about is "fake news". He notes a recent story of two AI programmes at a Facebook research center that began talking to each other in their own language and as a result the researchers shut them down.

Moedas notes that people were saying that this was the beginning of the end that robots would soon take over the human race. But Moedas said that the shutting down of the programmes had little to do with fear of unleashing an uncontrollable force but with not fulfilling requirements.

As many news reports misrepresented the facts he calls it fake news. While some of the reasons for the facts are misrepresented the two programmes did develop their own language, one which was a type of shorthand code that the programmes used in trading. as described in this Telegraph article.

It would take some effort no doubt to interpret the shorthand used by the programs. The shortcoming of the research was that they had not thought of rewarding the programme traders for using English. They actually thought that the programme aims had already been fulfilled. Given that the programmes were no longer using English they were shut down.

A Snopes' article discusses the issue in detail including emails with the head researcher. They did later research that rewarded the programmes for using English.

The lead author Michael Lewis said that he was not worried that the programmes developed their own language saying: “While it is often the case that modern AI systems solve problems in ways that are hard for people to interpret, they are always trying to achieve the goals that were given to them by people.”

Controlling the flow of information

Moedas notes that if complex scientific debates are distorted in the media the result can be immeasurable effects on our lives and families for years to come.

Moedas' concern is being reflected in upcoming studies by the EU. The EU is considering what steps it can take to stem the tide of fake news on all topics.

There are huge debates about issues even within the scientific community. When one considers all topics disagreement is even more prevalent. What is fake news for Donald Trump is probably not fake to many mainstream media analysts. Where is there an objective definition of fake news?

Frans Timmermans, Commission's vice-president said a week ago "“The flow of information and misinformation has become almost overwhelming. That is why we need to give our citizens the tools to identify fake news, improve trust online, and manage the information they receive.”

The Commission would like to control the flow of information as well as evaluating and censuring some. It wants to set up an expert panel of academics to assess who should play what role in tackling misinformation that is flowing through platforms such as Google, Facebook, You Tube, Instagram, and Twitter. No doubt AI programmes using biased algorithms will be established to filter content and feed in positive sponsored content.

Already there is a new law in Germany that is targeting social media platforms such as Facebook and Twitter. The law requires companies to remove malicious content such as hate speech within 24 hours or face fines up to 30 million Euros. Who decides what is malicious content?

The ideal no doubt would be for citizens to only read officially approved news or alternatively have citizens trained only to trust officially sanctioned news. Only official fake news will be effective.

Previously published in Digital Journal

Over next 3 years Chinese giant Alibaba will invest $15 billion in new technology

The Chinese giant firm Alibaba has announced that it will invest $15 billion over the next three years into cutting edge technology including artificial intelligence (AI) and quantum computing.

Alibaba has called its new initiative "DAMO Academy" standing for Discovery, Adventure, Momentum, and Outlook. Research labs will be created not only in the Chinese cities of Beijing and Hangzhou but in San Mateo and Bellevue in the United States. Other cities in which there will be new facilities include Moscow in Russia, Tel Aviv in Israel, and Singapore, Malaysia. All the facilities will be overseen by an advisory board of the best researchers and professors from leading global universities including the MIT in the US. The new units will explore developing technologies relating to data intelligence, Internet of Things (IoT), quantum computing, and human-machine interacting. Jeff Zang, Alibaba's chief technology officer said: "We aim to discover breakthrough technologies that will enable greater efficiency, network security and ecosystem synergy for end-users and businesses everywhere."

Facebook CEO Mark Zuckerberg, is a big supporter of AI and claims it could be used to build safer cars and better detect diseases. However others such as Elon Musk of Tesla suggest that government regulation of research is needed to ensure that research does not threaten humanity. Other critics such as John Giannandrea AI chief at Google are concerned about biased algorithms used in AI as discussed in a recent Digital Journal article.

The DAMO Academy project is just part of a much broader plan that would see Alibaba serve 2 billion customers and create 100 million jobs in 20 years the company maintains. The company is busy trying to entice US small businesses to sell their goods to its huge Chinese market of 443 million customers.

As of last April, Alibaba was the world's largest retailer surpassing Walmart. It operates in more than 200 countries. It is also one of the largest internet companies. Since 2015 its online sales and profits have surpassed those of Walmart, Amazon, and eBay combined. It is expanding quickly into the media industry. As of August this year, Alibaba has over 529 million mobile users across its platforms. It also hosts a digital distribution service 9Apps that hosts a huge amount of content plus apps for downloads. As of this October, the market cap of Alibaba was $463 billion making it one of the top ten most valuable companies in the world.

Google Aritificial Intelligence chief claims biased algorithms are a big danger

Social concerns are at the forefront of Artificial Intelligence conversations right now. John Giannandrea AI chief at Google is concerned that bias is being built in to many of the machine-learning algorithms by which the robot makes decisions.

At a recent Google conference on the relationships between AI systems and humans Giannandrea said: “The real safety question, if you want to call it that, is that if we give these systems biased data, they will be biased.”

This problem is likely to become more widespread as AI technology spreads to areas such as law, and medicine and as more people use the technology who do not have a deep understanding of the technical problems with the algorithm's being used. Indeed many who use the technology stress that it eliminates human bias, as is boasted in the following ad:

Two flavours of AI

In the early stages of the development of AI there were two main models. There was a logic, rules-based model and a competing biological model. The logic model would have developed AI in a manner that anyone who cared to could determine exactly how a robot or other device made decisions but the biological model took a quite different approach:

Instead of a programmer writing the commands to solve a problem, the program generates its own algorithm based on example data and a desired output. The machine-learning techniques that would later evolve into today’s most powerful AI systems followed the latter path: the machine essentially programs itself.

Tommi Jaakkola a professor at MIT who works on applications of machine learning said: “It is a problem that is already relevant, and it’s going to be much more relevant in the future. Whether it’s an investment decision, a medical decision, or maybe a military decision, you don’t want to just rely on a ‘black box’ method.”

Giannandrea elaborated on the problem: “It’s important that we be transparent about the training data that we are using, and are looking for hidden biases in it, otherwise we are building biased systems. If someone is trying to sell you a black box system for medical decision support, and you don’t know how it works or what data was used to train it, then I wouldn’t trust it.”

COMPAS

There have already been complaints of bias in AI algorithm programs. A system called COMPAS predicts defendant's likelihood of re-offending and is even used by some judges in determining whether an inmate is to be granted parole. Although the workings of COMPAS are kept secret an investigation by ProPublica has argued that there is evidence that the model used may involve bias against minorities.

The issue is complicated and Northpointe, the company that created COMPAS, has issued rebuttals to ProPublica's arguments and there has been a continuing back and forth debate that is discussed in detail in this article. The Wisconsin Supreme Court has upheld the use of COMPAS in sentencing.

The COMPAS program assigns defendants scores from 1 to 10 based on how likely they are to re-offend based on more than 100 factors none of which directly include race. The scores have been found to be highly predictive of re-offending. However, ProPublica questions whether they are fair.

Northpointe argued that the assessments were fair in that the scores meant the same regardless of race. However, ProPublica pointed out that among defendants who ultimately did not re-offend, blacks were more than twice as likely as whites to be classified as high risk, 42 percent versus 22 percent. Hence, even though these black defendants did not go on to commit a crime they were subjected to harsher treatment by the courts.

Complex intelligence

Another issue that is not being adequately faced is that the new deep learning techniques are so complex and opaque that they often cannot be understood. However, researchers are attempting to create ways that give them at least some approximation of how they work to engineers and end users.

Karrie Karahalios, a professor of computer science at the University of Illinois showed that even commonplace algorithms, such as those used by Facebook to filter the posts they see in their news feed are not usually understood by users and thus it is difficult to detect bias. The appended video discusses some of the controversy over Facebook filters. Facebook showed clear evidence of censoring Conservative media as shown on the appended video. Google has been accused of censorship by left wing sources in the ranking algorithm it uses for its search engine.