Daily Archives: November 25, 2019

AI is already a significant part of journalism today, but as it stands now the technology is unevenly distributed. Newsroom activities based on technology including search, complex algorithms drawing upon deep learning to create text or videos, all fall under the innovative umbrella of AI.

AI in journalism is developing at a fast pace while bringing about radical changes in media production and its business models. Although its futuristic impact is uncertain yet it has the potential for wide-ranging and profound influence on how journalism is done and spread.

The technology is a part of this sector since the newsrooms have gone digital. The adoption of social media as a source, production tool, and as a distribution and engagement medium is one of the major drivers of this digital trend. Moreover, AI possesses the potential to accelerate journalism throughout the process in significant ways which may have creative structural effects in the coming years. However, it has been discovered in a report, that the use of AI in this field is as additional, supplementary and catalytic factor rather than completely being a transformational driver.

In such an advanced world, there is a range of approaches to AI adoption. Artificial Intelligence should be aimed at providing more direct, meaningful, and engaging experiences in its main services. It should be employed to empower journalists in their news reporting and storytelling. The industry needs to create new methods and tools to better understand journalism and the world around it. AI can be employed to increase the capability for investigations by looking at big data sets, to be faster in finding news and to improve fact-checking / fight disinformation.

The media industry today not only sees AI as a tech-benefactor but also consider it beneficial in the economic context. The rising new organizations are fighting for attention and revenue with everything else online. Here AI is placed as a potential catalyst for renewal and to avoid being left behind or staying irrelevant in the news market.

As the industry is undergoing a crisis, AI facilitates every measure that might provide a competitive advantage to various organizations. In todays digital era, it is difficult for news to survive without technology.

As technology has transformed many industries, in the same way, spreadsheets, databases, and mapping programs have ventured into the newsroom. Also, ML methods are becoming more accessible to motivated journalists and enabling them to use that power for their reporting into stories that would otherwise be difficult or even impossible. This significantly marks the upsurge of next-gen data journalism.

Below are the 10 ways in which AI can reshape the newsroom and journalism as a whole:

Several media professionals strongly believe that AI is there to make journalists more efficient and to increase capacity for firstly, to free up journalists to work, with or without AI, on creating better journalism at a time when the news industry is struggling for economic sustainability and public trust and relevance; secondly, to help the public deal with a world of news overload and misinformation and to connect them in a convenient way to credible source and content that is relevant, useful and stimulating for their lives.

It is high time that media organizations should adopt AI strategy, change their workflows, systems and recruitment process. They cant go alone in this race. They need to embrace AI as the technology is shaping the information ecosystem for new generations of citizens. The news-organizations need to find ways to tap into the resources and expertise of others while encouraging the healthy development of this media industry with this technology.

Share This ArticleDo the sharing thingy

About AuthorMore info about author

Smriti is a Content Analyst at Analytics Insight. She writes Tech/Business articles for Analytics Insight. Her creative work can be confirmed @analyticsinsight.net. She adores crushing over books, crafts, creative works and people, movies and music from eternity!!

See the article here:

Relevance of AI in Accelerating Journalism and Newsroom Workflow - Analytics Insight

AUSTIN, Texas--(BUSINESS WIRE)--ClearDATA, the leader in healthcare public cloud security and compliance, today announced a new agreement with Independence Health Group, Inc. (Independence) in Philadelphia. Independence desires to transform and improve member engagement by leveraging Amazon Web Services AI Cloud Services technology, such as Amazon Comprehend Medical and Amazon Transcribe. Under the agreement, Independence receives not only the Amazon Web Services technology, but also ClearDATAs enhanced security offerings, including ClearDATAs Business Associate Agreement (BAA). This arrangement supports Independences goal of delivering improved quality of service to members, via a short-term analysis by AI Cloud Services, while at the same time utilizing ClearDATAs additional monitoring services that enhance the security protections offered for these products. This solution will support the Independence service team in improving first call resolution in their call center, increasing engagement, and identifying trends or issues members might experience with the Independence part of their health care journey.

With the increasing importance of analytics and data management, a cloud-enabled mentality allows health insurers to streamline operations and optimize costs while delivering innovative approaches to improve member interaction. According to a 2018 survey by research firm IDC, consumer engagement and the move toward optimal health is driving cloud adoption by health insurers. Yet many payers remain unsure of how to appropriately monitor security risks and configure their environment to best protect their data assets on the cloud.

Over the last few years, health insurance companies of all sizes have been under increasing pressure to meet the expectations of todays digital consumer, but they are challenged by regulations, rising costs and lack of security preparedness for mobile apps and cloud based technologies, said Scott Whyte, Chief Strategy Officer at ClearDATA. Our agreement with Independence demonstrates healthcares desire for solutions that make it easier to limit risk, manage compliance and still use powerful, native cloud-based products to improve member services.

By utilizing cloud services and leveraging ClearDATAs deep compliance and cloud technology expertise, Independence joins the ranks of health insurance organizations including multiple Blue Cross Blue Shield licensees across the nation leveraging public cloud solutions through ClearDATA. The continued growth of ClearDATAs payer customer base is further proof of the industrys desire to use cloud products for increased efficiency and enhanced interoperability allowing for improved member experience, while managing industry privacy, security and compliance needs as well.

ClearDATA will be in attendance at AWS re:Invent 2019 in Las Vegas from Dec. 2-6. To schedule onsite interviews with Scott Whyte for further discussion of healthcare innovation in the cloud and the influence of emerging technologies like machine learning on payers, contact media representative Sanah Sadaruddin at sanah@growswyft.com. To learn more about how ClearDATA is making healthcare better every single day, visit https://www.cleardata.com.

About ClearDATA

Healthcare professionals across the globe trust the ClearDATA HITRUST-certified cloud to safeguard their sensitive data and power their critical applications available across the major public cloud platforms. For healthcare organizations, customers receive one of the most comprehensive Business Associate Agreements (BAA) in the industry, combined with market-leading healthcare-exclusive security and compliance solutions, and multi-cloud expertise. ClearDATAs innovative platform of solutions and services helps customers to protect against data privacy risks, improve their data management, and scale their healthcare IT infrastructure, enabling the industry to focus on making healthcare better by improving healthcare delivery, every single day.

Here is the original post:

Independence Health Group, Inc. Sources Advanced AI Cloud Services through ClearDATA to Bolster Healthcare Security Compliance - Business Wire

Humans have enlisted nearly 100 AI-powered robots in North American to come to the rescue for something humans are terrible at: recycling.

Even when we try to do it right, we're often making things worse; About one out of every four of the things people throw into the recycling bin aren't recyclable at all.

All those misplaced greasy pizza boxes (not recyclable) and clamshell containers tossed in with the plastics, have imperiled an industry that was never really that effective in the first place.

Only a small fraction of the over 2.1 billion tons of the garbage the world produces each year gets recycled about 16%.

And even that small sliver has gotten smaller over the past year.

For decades, the US sold more than half of its recyclables to China mostly plastics to be melted into pellets, the raw material for making more plastic.

But in March of 2018, China said, "No More."

"They started shipping more and more stuff to China, often contaminated dirty plastics or mixed too many mixed goods," said Kate O'Neill, a UC Berkeley professor and author of "Waste."

Around a quarter of the shipments China received had to be hand-processed, buried in landfills, or incinerated.

So the Chinese government declared that bales could contain only up to half a percent of things that contaminated them, like food wrappers or a dirty jar of peanut butter. US consumers and recycling centers couldn't keep up.

"I think people in the wealthy countries had gotten complacent, never bothering to build more recycling facilities domestically," O'Neill added.

Today, a handful of start-ups are testing out new technology to make recycling sustainable.

AMP Robotics is an artificial intelligence and robotics company that aims to change the way we recycle.

Founder of AMP Robotics, Matanya Horowitz said "the situation with the Chinese export markets have actually been good for [the company]."

Robots use artificial intelligence to sort through recyclables. BHS

AMP Robotics is rolling out its latest model: a "Cortex Robot" that uses optical sensors to take in what rolls by, and a "brain" to figure out what his "hands" should do with something even if it looks different to anything he's seen before.

"A lot of these recycling facilities are structured with the primary task of basically dealing with contamination that's not supposed to be there," said Horotwiz. ""What we see is a lot of recycling facilities are investing in automation to help improve their operations."

At least four companies are rolling out similar models, in the hopes of turning a profit from the US' growing piles of hard-to-sort recyclables.

And investors are taking notice. In November 2019, AMP Robotics announced a $16 million Series A investment from Sequoia Capital.

But what about helping humans get better at choosing what to put in their recycling bins in the first place?

New policies in Shanghai are one of the first steps in China's push to solve its waste problems.

This past summer, citizens will face fines and what are called "social penalties" if they don't sort things properly.

One trash sorting volunteer said, Shanghai started the test run on June 24. "It was very hard for us at the beginning. Everyone was busy, people didn't know how to sort," the volunteer who requested to be unidentified said.

"At first we had some hard times," said Shanghai citizen Zhaoju Zhang. "The most difficult part was how to differentiate between dry and wet trash. It was so complicated that we all got confused."

Almost immediately, hundreds of AI-enabled apps sprouted up in order to assist everyday sorting.

"If it's something that is confusing whether it's dry or wet trash, we can just scan the item and get the answer," Zhang said.

Shanghai citizens are now required to sort recyclables properly from their trash. Yuan Ye

But not everyone has access to AI to help parse the new rules, and many complain that complying is tough, and punishments are too harsh.

Kate O'Neill said the new laws are having a "massive cultural impact" and there are "some concerns about how draconian it is, but it's too early to really tell the results. But it certainly has seems to be a massive culture shift."

This kind of cultural shift in how we throw things away would be challenging in the US, where the average person produces twice as much trash as a Chinese citizen.

But experts warn that rethinking the way we deal with garbage is essential, and AI technology offers a promising way forward.

It's even possible for it to identify who created a piece of trash in the first place.

Horowitz explained that robots are able to learn the features of materials. They are able to sparse whether a material is cloudy or opaque. AI robots may even be able to identify symbols of specific brands. All of these abilities help the robots like Max narrow down the source of contamination and what to do with it.

Last year, over 250 companies signed a MacArthur Foundation agreement pledging that 100% of plastic packaging will be easily and safely reused, recycled, or composted by 2025.

CEO of SC Johnson, Fisk Johnson, said in an interview, "We're a family company, and we have a very long-term view, and business has to be part of the solution."

Whether or not they make good on this pledge, AI will be quietly watching, and gathering data on the packaging these brands continue to use.

Read more:

Most plastic is not getting recycled, and AI robots could be a solution - Business Insider

NEW YORK, Nov. 25, 2019 /PRNewswire/ --

Growing number of cross-industry collaborations and partnerships and the need to control drug discovery & development costs and reduce the overall time taken in this process are the key factors driving the AI in the drug discovery market.

Read the full report: https://www.reportlinker.com/p05828730/?utm_source=PRN

The global AI in the drug discovery market is projected to reach USD 1,434 million by 2024 from USD 259 million in 2019, at a CAGR of 40.8% during the forecast period. Growth in this market is mainly driven by growing number of cross-industry collaborations and partnerships, the need to control drug discovery & development costs and reduce the overall time taken in this process, the rising adoption of cloud-based applications & services, and the impending patent expiry of blockbuster drugs. On the other hand, a lack of data sets in the field of drug discovery and the inadequate availability of skilled labor are some of the factors challenging the growth of the market.

The immuno-oncology segment accounted for the largest share in 2019.Based on application, the artificial intelligence in the drug discovery market is segmented into neurodegenerative diseases, immuno-oncology, cardiovascular disease, metabolic diseases, and other applications. The immuno-oncology segment accounted for the largest share of 44.6% of the AI in the drug discovery market in 2018, owing to the increasing demand for effective cancer drugs. Neurodegenerative diseases form the fastest-growing application segment, with a CAGR of 42.9% during the forecast period. The ability of AI to discover drugs for complex diseases and the emphasis of market players on providing AI-based platforms for neurological diseases are responsible for the fast growth of this application segment.

The Research centers and academic & government institutes segment to register the highest growth rate in the forecast period.Based on end-user, the AI in the drug discovery market is segmented into pharmaceutical & biotechnology companies, contract research organizations, and research centers and academic, & government institutes.In 2018, the pharmaceutical & biotechnology companies segment accounted for the largest share in the AI in the drug discovery market.

AI and machine learning to allow pharmaceutical companies to operate more efficiently and substantially improve success rates at the early stages of drug development.This is one of the major factors driving the growth of this market.

Research centers and academic & government institutes are expected to show the highest growth of 43.6%.during the forecast period. The growth of the CROs segment is tied to that of pharmaceutical & biotechnology companies, as the rise in research and production activity will ensure sustained demand for contract services.

North America to be the largest and the fastest-growing regional market.

North America, which comprises the US, Canada, and Mexico, forms the largest market for AI in drug discovery.These countries have been early adopters of AI technology in drug discovery and development.

In the North American market, the US is a significant contributor.Also, prominent AI technology providers, such as IBM, Google, Microsoft, NVIDIA, and Intel, are headquartered in the US; their strong presence is a key contributor to market growth.

Other drivers include the well-established pharmaceutical industry, high focus on R&D & substantial investment, and the presence of globally leading pharmaceutical companies. These are some of the major factors responsible for the large share and high growth rate of this market.

The primary interviews conducted for this report can be categorized as follows: By Company Type: Tier 1 (28%), Tier 2 (42%), and Tier 3 (30%) By Designation: C-level (30%), D-level (34%), and Others (36%) By Region: North America (46%), Europe (25%), Asia (18%), and the RoW (11%)

List of Companies Profiled in the Report IBM Corporation (US) Microsoft (US), Google (US) NVIDIA Corporation (US) Atomwise, Inc. (US) Deep Genomics (Canada) Cloud Pharmaceuticals (US) Insilico Medicine (US) Benevolent AI (UK) Exscientia (UK) Cyclica (Canada) BIOAGE (US) Numerate (US) Numedii (US) Envisagenics (US) twoXAR (US) OWKIN, Inc. (US) XtalPi (US) Verge Genomics (US) Berg LLC (US)

Research Coverage:This report provides a study of the AI in the drug discovery market.It aims at estimating the size and future growth potential of the market across different segments, such as offering, technology, application, end-user, and region.

The report also includes an in-depth competitive analysis of the key market players, along with their company profiles, recent developments, and key market strategies.

Key Benefits of Buying the Report:The report will help market leaders/new entrants by providing them with the closest approximations of revenue numbers for the overall AI in drug discovery market and its subsegments.This report will also help stakeholders understand the competitive landscape, and gain insights to better position their business and make suitable go-to-market strategies.

It will also enable stakeholders to understand the pulse of the market and provide them with information on the key market drivers, challenges, and opportunities.

Read the full report: https://www.reportlinker.com/p05828730/?utm_source=PRN

About Reportlinker ReportLinker is an award-winning market research solution. Reportlinker finds and organizes the latest industry data so you get all the market research you need - instantly, in one place.

__________________________ Contact Clare: clare@reportlinker.com US: (339)-368-6001 Intl: +1 339-368-6001

SOURCE Reportlinker

http://www.reportlinker.com

Read the rest here:

The global AI in the drug discovery market is projected to reach USD 1,434 million by 2024 from USD 259 million in 2019, at a CAGR of 40.8% -...

Data security and automation are the top IT projects for 2020, while artificial intelligence projects are not in the top 10 for IT professionals, according to Netwrix.

The online survey asked 1045 IT professionals worldwide to name their top five IT projects for the next year; they could pick from a predefined list or specify their own descriptions. The survey found no dramatic difference in IT priorities among organizations based on size or vertical.

Not surprisingly, data security is a top priority for the majority of organizations. There are several factors that go into a successful data security process. The first is automating current processes to free up time for data security projects. Another is to research and deploy a data security solution.

Be sure your solution offers automated data classification, because it is the optimal way to enhance data security and reduce your attack surface without additional effort by the IT team, said Steve Dickson, CEO of Netwrix.

Infrastructure, operations, networking and security are the foundation upon which the technology-enabled world is built. It must deliver value at each layer as it ultimately supports people.

See the original post here:

Data security and automation top IT projects for 2020, AI not a priority - Help Net Security

Steph is used to starting over. She and her mother have been on the run from her father for as long as she can remember, moving from town to town and school to school, always leaving when it seems like there's a chance he could find them. The only thing that remains consistent in her life is the close connection she's made with a chat group on a website called CatNet. They upload photos of cute animals, they talk about being teenagers, they encourage each other all from a safe, anonymous distance.

But one member of CatNet has a secret. CheshireCat isn't a teenager at all, but a sentient AI with a love for cat photos and an intense interest in the inner workings of humans. That's why they created CatNet in the first place. But they didn't consider what would happen if they came to care for the humans they interacted with, let alone what they would do if one of those humans was in danger.

Danger does indeed come for Steph. As she tries to settle into yet another new school, she chafes against the restraints of her life and begins to take the kinds of risks that would make her mother load the van and speed out of town. It seems almost like a game, but one choice compounds the next, secrets begin to unravel, and soon Steph finds herself in the kind of trouble that will require a miracle to get out of or the help of an all-powerful AI.

This book is an engaging blend of tech thriller, mystery, and teen drama that kept me up reading way later into the night than was strictly wise.

This book is an engaging blend of tech thriller, mystery, and teen drama that kept me up reading way later into the night than was strictly wise. But I almost didn't read it. The title, Catfishing on CatNet, was nearly a deal-breaker for me. It sounded goofy and awkwardly trendy in a try-hard sort of way. But author Naomi Kritzer won the Hugo and Locus awards for her short story, "Cat Pictures Please," which is based on the same premise, so I decided to take a risk. Now I'm here to tell you that if the title has the same effect on you as it did on me, you should get over it and try it anyway.

Why? Well, the teen drama aspect is heartfelt and relatable, the mystery has enjoyable and sometimes shocking twists and turns, and the trajectory of the thriller plot is quite frankly bonkers. Just when you think it's about to slow down or pull back, instead it goes there and then keeps on going. It doesn't always make the most sense, but when you're talking about impromptu armies of hijacked robots, who even cares? It's bold and absurd and a whole lot of fun.

Steph's life is the stuff of made-for-TV drama, but despite that, she feels deeply relatable and accessible as a character. We meet her at a moment when she realizes that she should be asking more questions about her life and begins throwing rocks at the fence that surrounds her, testing its strength. We also get occasional passages from CheshireCat's point of view, and they manage to be simultaneously alarming and affable, acting with a shocking boldness and then wringing their virtual hands, wondering if they've done the right thing. This story heralds a coming of age for both its human and AI protagonists, and the parallels and differences are illuminating.

Catfishing on CatNet taps lightly on the concept of personhood and the ethics of artificial intelligence, but it doesn't trouble itself profoundly. There's never any question of whether or not we're on board with CheshireCat being a person. They just are.

That said, it does traffic deeply in surveillance and cyber security, and as someone who tries really hard to not think about that aspect of modern life too much, I did find myself genuinely creeped out by the amount of spying and hacking various characters are able to do. I know it's science fiction, but I also know it isn't reaching. And I think that's a message for our current times that is important for readers of all ages to keep in mind: You don't really know who you're talking to on the internet, and you don't know what you're telling them without meaning to. This gritty truth runs beneath Catfishing on CatNet's wild, rollicking ride, and it leaves it mark.

Caitlyn Paxson is a writer and performer. She is a regular reviewer for NPR Books and Quill & Quire.

Read the original here:

Review: Further Adventures Of A Cat-Loving AI In 'Catfishing On CatNet' - NPR

NASHVILLE, Tenn., Nov. 25, 2019 /PRNewswire/ -- HCXI, a blockchain and AI based risk solutions company focused on managing alternative and emerging risks as new asset classes, has appointed former Swiss Re and AIG executive David Bassi to its advisory board. Mr. Bassi will advise HCXI on risk transfer and mitigation product innovation bringing considerable experience in insurance risk management and innovation to the role.

Mr. Bassi joins the HCXI Advisory Board with over 25 years of experience in the global insurance industry having held executive roles in innovation, risk management, and modeling. David most recently served as a Managing Director at Ernst & Young in Boston where he focused on insurance strategy and innovation. Prior to this role, he served as the Head of Innovation and Risk Consulting, Casualty, at AIG and held various risk management roles at leading companies including Swiss Re and General Electric. David is a speaker at industry events on topics such as emerging risks, big data, blockchain and the insurance cycle. He is a contributing author to Insurance Thought Leadership having written on the transformative potential of technology and analytics for the insurance industry.

HCXI is currently developing a smart-contract based risk mitigation and transfer product, vizSaver, to address the negative health outcomes and financial costs ($500+ billion) associated with missed patient appointments. vizSaver aligns the interests of patients, health systems, medical cost payers, and providers addressing delivery inefficiencies, social and financial determinants, and other patient, provider, and health system circumstances.

"We are pleased and excited to welcome David to our Advisory Board at this important stage of HCXI's evolution," stated Cyrus Maaghul, Founder and Chief Executive Officer. "As a highly respected and seasoned international insurance executive with valuable risk management, innovation, and data analytics experience, we look forward to David's advice and insight in guiding our objectives to bring breakthrough risk solutions to industry."

About HCXI

HCXI is an early stage RiskTech innovator developing smart-contracts leveraging mobile technology and machine learning as the basis for new risk mitigation, transfer and financing solutions for industries including healthcare, emerging technologies (AI, Blockchain), and others.

https://www.hcxi.co

Media Contact:Cyrus Maaghul615.310.7944229784@email4pr.com

SOURCE HCXI

https://www.hcxi.co

Read more from the original source:

Blockchain and AI Risk Management Technology Company, HCXI, Announces the Appointment of Former Swiss Re and AIG Executive to Its Advisory Board -...

Headlines about machine learning promise godlike predictive power. Here are four examples:

With articles like these, the press will have you believe that machine learning can reliably predict whether you're gay, whether you'll develop psychosis, whether youll have a heart attack and whether you're a criminalas well as other ambitious predictions such as when you'll die and whether your unpublished book will be a bestseller.

It's all a lie. Machine learning cant confidently tell such things about each individual. In most cases, these things are simply too difficult to predict with certainty.

Here's how the lie works. Researchers report high "accuracy," but then later revealburied within the details of a technical paperthat they were actually misusing the word "accuracy" to mean another measure of performance related to accuracy but in actuality not nearly as impressive.

But the press runs with it. Time and again, this scheme succeeds in hoodwinking the media and generating flagrant publicity stunts that mislead.

Now, don't get me wrong; machine learning does deserve high praise. The ability to predict better than random guessing, even if not with high confidence for most cases, serves to improve all kinds of business and health care processes. That's pay dirt. And, in certain limited areas, machine learning can deliver strikingly high performance, such as for recognizing objects like traffic lights within photographs or recognizing the presence of certain diseases from medical images.

But, in other cases, researchers are falsely advertising high performance. Take Stanford University's infamous "gaydar" study. In its opening summary, the 2018 report claims its predictive model achieves 91 percent accuracy distinguishing gay and straight males from facial images. This inspired journalists to broadcast gross exaggerations. The Newsweek article highlighted above kicked off with "Artificial intelligence can now tell whether you are gay or straight simply by analyzing a picture of your face."

This deceptive media coverage is to be expected. The researchers opening claim has tacitly conveyedto lay readers, nontechnical journalists and even casual technical readersthat the system can tell who's gay and who isn't and usually be correct about it.

That assertion is false. The model can't confidently "tell" for any given photograph. Rather, what Stanford's model can actually do 91 percent of the time is much less remarkable: It can identify which of a pair of two males are gay when it's already been established that one is and one is not.

This "pairing test" tells a seductive story, but it's a deceptive one. It translates to low performance outside the research lab, where there's no contrived scenario presenting such pairings. Employing the model in the real world would require a tough trade-off. You could tune the model to correctly identify, say, two thirds of all gay individuals, but that would come at a price: When it predicted someone to be gay, it would be wrong more than half of the timea high false positive rate. And if you configure its settings so that it correctly identifies even more than two thirds, the model will exhibit an even higher false positive rate.

The reason for this is that one of the two categories is infrequentin this case, gay individuals, which amount to about 7 percent of males (according to the Stanford report). When one category is in the minority, that intrinsically makes it more challenging to reliably predict.

Now, the researchers did report on a viable measure of performance, called AUCalbeit mislabeled in their report as "accuracy." AUC (Area Under the receiver operating characteristic Curve) indicates the extent of performance trade-offs available. The higher the AUC, the better the trade-off options offered by the predictive model.

In the field of machine learning, accuracy means something simpler: How often the predictive model is correctthe percent of cases it gets right. When researchers use the word to mean anything else, they're at best adopting willful ignorance and at worst consciously laying a trap to ensnare the media.

But researchers face two publicity challenges: How can you make something as technical as AUC sexy and at the same time sell your predictive models performance? No problem. As it turns out, the AUC is mathematically equal to the result you get running the pairing test. And so, a 91 percent AUC can be explained with a story about distinguishing between pairs that sounds to many journalists like "high accuracy"especially when the researchers commit the cardinal sin of just baldlyand falselycalling it "accuracy." Voila! Both the journalists and their readers believe the model can "tell" whether you're gay.

This accuracy fallacy scheme is applied far and wide, with overblown claims about machine learning accurately predicting, among other things, psychosis, criminality, death, suicide, bestselling books, fraudulent dating profiles, banana crop diseases and various medical conditions. For an addendum to this article that covers 20 more examples, click here.

In some of these cases, researchers perpetrate a variation on the accuracy fallacy scheme: they report the accuracy you would get if half the cases were positivethat is, if the common and rare categories took place equally often. Mathematically, this usually inflates the reported "accuracy" a bit less than AUC, but it's a similar maneuver and overstates performance in much the same way.

In popular culture, "gaydar" refers to an unattainable form of human clairvoyance. We shouldnt expect machine learning to attain supernatural abilities either. Many human behaviors defy reliable prediction. Its like predicting the weather many weeks in advance. There's no achieving high certainty. There's no magic crystal ball. Readers at large must hone a certain vigilance: Be wary about claims of "high accuracy" in machine learning. If it sounds too good to be true, it probably is.

Original post:

The Media's Coverage of AI is Bogus - Scientific American

Executive Summary

Any examination of bias in AI needs to recognize the fact that these biases mainly stem from humans inherent biases. The models and systems we create and train are a reflection of ourselves. So its no surprise to find that AI is learning gender bias from humans. For instance, natural language processing (NLP), a critical ingredient of common AI systems like Amazons Alexa and Apples Siri, among others, has been found to show gender biasesand this is not a standalone incident. There have been several high profile cases of gender bias, including computer vision systems for gender recognition that reported higher error rates for recognizing women, specifically those with darker skin tones. In order to produce technology that is more fair, there must be a concerted effort from researchers and machine learning teams across the industry to correct this imbalance. We have an obligation to create technology that is effective and fair for everyone.

Any examination of bias in AI needs to recognize the fact that these biases mainly stem from humans inherent biases. The models and systems we create and train are a reflection of ourselves.

So its no surprise to find that AI is learning gender bias from humans. For instance, natural language processing (NLP), a critical ingredient of common AI systems like Amazons Alexa and Apples Siri, among others, has been found to show gender biasesand this is not a standalone incident. There have been several high profile cases of gender bias, including computer vision systems for gender recognition that reported higher error rates for recognizing women, specifically those with darker skin tones. In order to produce technology that is more fair, there must be a concerted effort from researchers and machine learning teams across the industry to correct this imbalance. Fortunately, we are starting to see new work that looks at exactly how that can be accomplished.

Building fair and equitable machine learning systems.

Of particular note is the bias research being carried out with respect to word-embeddings, which is when words are converted to numerical representations, which are then used as inputs in natural language processing models. Word-embeddings represent words as a sequence, or a vector of numbers. If two words have similar meanings, their associated embeddings will be close to each other in a mathematical sense. The embeddings encode this information by assessing the context in which a word occurs. For example, AI has the ability to objectively fill in the word queen in the sentence Man is to king, as woman is to X. The underlying issue arises in cases where AI fills in sentences like Father is to doctor as mother is to nurse. The inherent gender bias in the remark reflects an outdated perception of women in our society that is not based in fact or equality.

Few studies have assessed the effects of gender bias in speech with respect to emotion and emotion AI is starting to play a more prominent role in the future of work, marketing, and almost every industry you can think of. In humans, bias occurs when a person misinterprets the emotions of one demographic category more often than another for instance, mistakenly thinking that one gender category is angry more often than another. This same bias is now being observed in machines and how they misclassify information related to emotions. To understand why this is, and how we can fix it, its important to first look at the causes of AI bias.

What Causes AI Bias?

In the context of machine learning, bias can mean that theres a greater level of error for certain demographic categories. Because there is no one root cause of this type of bias, there are numerous variables that researchers must take into account when developing and training machine-learning models, with factors that include:

Four Best Practices for Machine-Learning Teams to Avoid Gender Bias

Like many things in life, the causes and solutions of AI bias are not black and white. Even fairness itself must be quantified to help mitigate the effects of unwanted bias. For executives who are interested in tapping into the power of AI, but are concerned about bias, its important to ensure that the following happens on your machine-learning teams:

Although examining these causes and solutions is an important first step, there are still many open questions to be answered. Beyond machine-learning training, the industry needs to develop more holistic approaches that address the three main causes of bias, as outlined above. Additionally, future research should consider data with a broader representation of gender variants, such as transgender, non-binary, etc., to help expand our understanding of how to handle expanding diversity.

We have an obligation to create technology that is effective and fair for everyone. I believe the benefits of AI will outweigh the risks if we can address them collectively. Its up to all practitioners and leaders in the field to collaborate, research, and develop solutions that reduce bias in AI for all.

Originally posted here:

4 Ways to Address Gender Bias in AI - Harvard Business Review