Category: Uncategorized

Artificial intelligence gets due praise for its power to improve the way we work. Yet that same potential can be used to improve the way we live, even addressing some of our most intractable social issues. Recent developments in AI and other smart technologies are rare good news in one of America’s most urgent and emotionally fraught conversations: gun violence. 

The February 2018 shooting at Marjory Stoneman Douglas High School was the most high-profile incident in recent months, but school shootings have averaged one per week since the beginning of the year. Fear is prevalent, with one Gallup poll finding that 4 in 10 Americans fear becoming a victim of a mass shooting. In the wake of these tragedies, people are desperate for answers and reassurance. Everyone agrees something must be done, but they can’t agree on what. Yet the discussion around gun control and the prevention of mass shootings is highly politicized and emotionally charged. 

Technology, however, is dispassionate. Computer algorithms aren’t prone to uncertainty or emotion and can analyze data, lock weapons remotely, or identify guns before they’re drawn. Because of this, they can provide a steadying neutral influence as we strive to mitigate gun violence. 

While human decisionmaking and intervention are crucial to solving an issue as complex as gun violence, technologies like artificial intelligence and high-powered sensors are being designed to help stem the tide of mass shootings in the U.S. Here is a look at several innovative technologies being employed to combat the gun violence problem. 

From smartphones to smart guns 

After the 2015 San Bernardino shooting spree, former President Barack Obama publicly questioned why we can secure our smartphones using fingerprint biometrics yet hadn’t done the same for guns. It was a high-profile call for smart gun technology that could help prevent mass shootings. 

Theoretically, gun manufacturers could equip weapons with time-stamping and remote control capabilities that allow authorities to track when they’ve been used and shut them down when they’re being used in an attack. Biometric smart locks could inhibit criminal gun use because the devices would be all but useless unless a would-be shooter could fake the owner’s fingerprint or other biological identifiers. 

One company is already developing a smart lock that requires a registered fingerprint or physical key to unlock the gun’s trigger. Importantly, the locking system doesn’t link to an app or other computer software, a design decision intended to mitigate fears that the gun could be hacked. Increased research into this area could lead to the production of guns that are safer for their owners and for the general public. 

Can AI reduce gun violence? 

Smart guns, metal detectors, and panic buttons all have been discussed for catching potential shooters before they have a chance to act and for alerting emergency services immediately after a shooting incident begins. Software known as panic apps are designed to decrease the amount of time it takes for police and emergency workers to get to a school. Digitizing school blueprints can contribute to helping authorities reach victims faster, reducing the death and injury tolls during mass shootings. 

During a shooting, sensors and motion detectors can help law enforcement officers identify where a shooter is in the building. Although the technology is still in its early stages, smart video surveillance has the potential to assist in prevention as well. Researchers are currently studying how to train artificial intelligence algorithms that identify guns on video feeds. They’re feeding the program with movies like “Mission Impossible” and installments of the “James Bond” franchise so it can learn to detect guns not just on high-resolution screens but on grainier images as well. 

AI could play a supporting role as well. Advocates of gun insurance say that AI underwriting algorithms can keep guns safe and affordable for lawful, capable owners while keeping them out of the hands of high-risk individuals. The theory is that AI algorithms could draw on vast data sources to assess a person’s risk before they’re allowed to buy a gun. These data points might include criminal background checks, medical records, employment status, and even online behavioral data to assess their risk profiles. 

People who have clean records and other indicators of trustworthiness and lawfulness would receive lower insurance premiums. Those who have more erratic histories would be required to pay higher premiums, with the highest risk individuals potentially being priced out of the market. For this to work, gun insurance would need to be mandatory in the same way that auto insurance is required. But if implemented, it might bridge the gap between people on either side of the debate and help forge collective progress toward reducing gun violence.  

Preventing mass shootings is a complicated but essential task, and there are many factors that need to be addressed. Fortunately, we live in the most technologically advanced age in history, and we have access to more data every day. Artificial intelligence can help us make sense of that information by identifying trends we can use to create effective solutions. On gun violence and so many other urgent issues, we can leverage technology to find solutions to the most challenging issues of our time. 

The finance and insurance sector represents 7.3% of U.S. GDP, or about $1.4 trillion. The industry’s reach extends into practically every corner of every commercial market, as well as most U.S. households via checking accounts, 401k and IRA investments, and home and auto insurance policies. As the 2008 financial crisis demonstrated, what happens in this sector directly impacts the rest of the economy – and the world. 

For all of that scale, the financial industry has long struggled with challenges like rising regulatory compliance costs and ever-worsening credit card fraud. Yet these perennial problems are neither inevitable nor permanent. For today’s most forward-thinking banks and financial institutions (FIs), advanced artificial intelligence and machine learning solutions have reached a state of maturity where historic dynamics of compliance and risk can finally be turned around in the banks’ favor. 

The costs of compliance 

In the years following the 2008 financial crisis, the U.S. government ratcheted up regulatory requirements. The ostensible goal was to prevent another economic catastrophe, but bolstering the regulatory environment had some unintended negative effects. The Dodd-Frank bill, for example, was designed to create more stability and transparency within the financial system. However, the Dodd-Frank regulations became such a burden for small banks, many of them closed or were absorbed by larger institutions. This had negative repercussions for the small towns and communities many of them served. 

Increased compliance requirements have traditionally been met with increased hiring of compliance personnel. In just a single quarter during 2013, the finance industry added 10,000 jobs in compliance, risk, or regulatory-related functions. Ten thousand new jobs would usually be a cause for celebration, but compliance costs have become a huge operational and competitive burden for banks and other financial institutions. Like in any other regulated industry, such costs are often passed along to customers. 

There’s no end in sight, either. Accenture found that 89% of financial institutions expect their compliance investments to rise over the next two years. As evidence of their strategic importance, 66% of these institutions’ compliance groups report directly to the CEO or Board of Directors. Increasing compliance spending often forces banks to increase prices or to cancel new product development to minimize expenses and avoid potential losses. 

The link between data breaches and credit card fraud 

Compliance costs aren’t the only thing keeping financial executives up at night. Data hacks such as last year’s Equifax breach revealed the personally identifiable information (PII) of at least 145 million Americans. That’s grave news for banks and credit card issuers, because PII is a golden ticket for cybercriminals who want to commit credit card fraud. These records contain people’s names, addresses, birthdates, and even Social Security numbers in some cases. The exact type of data one needs to open a false account. 

Banks and FIs must also be on alert for synthetic identity fraud, a newer scheme that involves using some legitimate PII with fake data to create an account. It’s all the more insidious because oftentimes, cybercriminals will use a child’s PII for this form of identity theft. The logic is that adults might monitor their own credit accounts closely, but what parent thinks to check their 12-year-old’s credit score? Unfortunately, by the time the family catches on, the fraudster may have seriously jeopardized the child’s future financial standing. 

U.S. banks urgently need a solution to fraud mitigation. By one industry estimate, the U.S. sees 39.5% of the world’s payment card fraud losses. Within just the next three years, annual global payment card losses are projected to climb from $22.8 billion to $32.9 billion. 

Advanced AI can mean cost reduction and fraud prevention 

The financial industry’s monetary challenges are substantial on both the compliance and credit card fraud fronts. Credit card use increased by 2.6% this year, and more accounts means risk prevention and fraud detection will be all the more important. But the sheer volume of activity in both areas makes relying solely on human workers a losing strategy. The financial sector can augment their teams with artificial intelligence to reduce costs and lower their losses related to fraud.  

AI-powered automation can alleviate significant operating costs in functions such as customer verification and anti-money laundering (AML) strategies. The banking industry spends an estimated $270 billion a year on compliance in large part due to personnel costs. Rather than frantically hiring more workers, companies can implement technology that works around the clock – and that’s important, because the need for vigilance is constant. Smart processing systems can also scale up as necessary. No matter the volume of accounts created, computer programs can continue analyzing them at the same speed. 

McKinsey predicts that by 2020 AI-optimized fraud detection will grow to a $3 billion market. Artificial intelligence systems can detect suspicious behaviors within milliseconds. A program can learn a cardholder’s typical spending patterns and flag any purchases that fall outside of those. Machine learning systems continuously gather data on users, which enables them to become more and more accurate in identifying potential crimes. 

Banks and FIs should be taking this technology seriously for a number of reasons. The first is that compliance standards aren’t going away. As the industry grows, so does the regulatory scrutiny, and hiring people to meet tough regulation guidelines simply isn’t sustainable. However, training a machine learning system to flag potential fraud and refer it to trained analysts is a workable solution. It’s also worth noting that in addition to losses directly related to fraud, banks also face fines if their processes aren’t stringent enough to catch criminal activity. The federal government fined one U.S. bank $613 million for failing to implement strong AML protocols. 

But banks and FIs have a duty to their customers as well. Those that implement smart compliance and fraud detection systems position themselves to earn customer loyalty and increase their profits. By lowering compliance expenses, they’re able to invest in innovative concepts and improve their existing products. They also have more flexibility in their pricing structures, allowing them to make more competitive offers. 

Sixty-six percent of Americans fear becoming identity theft victims, and people want reassurance that their data and money is safe. Banks that deploy technologies like AI, machine learning, and advanced cybersecurity stand to enhance their value propositions with their customers and boost their competitive positions over the long run. 

New technology enables on-the-fly encryption and privacy controls that persist for the life of a message

PALO ALTO, Calif. June 13, 2018. Entefy Inc. has been issued a patent by the U.S. Patent and Trademark Office (USPTO). Patent No. 9,985,933 describes a “System and method of sending and receiving secret message content over a network.” The newly patented technology enables novel ways of embedding user-controlled privacy protection into messages and their attachments.

This latest Entefy patent is related to the company’s Adaptive Privacy Controls (APC) solution, a game-changer for individuals and organizations requiring message-by-message control over their digital communications and shareable assets. APC works independently of the underlying messaging protocol (e.g. email or text) and persists through forwarding, separating, or even copying file attachments from the original message.

“Email and digital messaging in general have become ubiquitous technologies despite never having solved fundamental issues of privacy, security, and data ownership,” said Entefy’s CEO, Alston Ghafourifar. “Entefy’s latest patent is a game-changer that allows message senders to transmit confidential or otherwise private information with the assurance that only designated recipients will have access. That’s even if a message is forwarded to a new recipient or its attachments are saved locally. It puts you in charge of the information you share with others.”

Entefy recently revealed its suite of machine intelligence solutions in conjunction with the launch of its new website. Today’s news is the latest in a series of patent issuance announcements, including an Entefy patent that enhances intelligent message delivery and a patent covering encrypted group messaging simultaneously across multiple protocols. To date, the company has a combined 48 filed and issued patents in AI, communication, search, blockchain, data privacy, and cybersecurity.

ABOUT ENTEFY

Entefy is a machine intelligence company developing advanced technologies in contextual cognition, computer vision, natural language, audio, time series, and other data intelligence. Entefy’s SaaS and on-premise solutions deliver transformative AI, communication, search, cybersecurity, IoT, and blockchain capabilities—helping people and organizations Discover & Do more in less time. Get started at www.entefy.com.

There is already enough food to feed everyone on Earth, with agricultural producers yielding 17% more food now than they did just three decades ago. Yet 925 million people worldwide suffer from a lack of food security, including 42.2 million in the United States alone. Several factors contribute to the problem, including poor storage and sanitation systems, low crop yields, and political upheaval. As the global population marches ever higher, the Food and Agriculture Organization predicts that food production will need to increase 70% to feed the world by 2050. 

How will the agriculture industry keep pace? Through cutting-edge technologies like artificial intelligence, IoT, and blockchain, of course.

IoT in the field

When people hear the term Internet of Things (IoT), they often think of smart homes and self-starting appliances. Or perhaps they think about how at present, the Internet of Things is still wildly insecure. But from an agricultural production perspective, the Internet of Things holds real potential for increasing crop yields and reducing losses.

IoT sensors installed in the field enable agricultural companies to monitor their crops in real time. The sensors capture data on a range of metrics and send back information that enables producers to optimize their growing processes. They can monitor nutrient and water levels in the soil and adjust them as needed to get the greatest number of healthy, saleable crops. Imaging programs allow them to see exactly what’s happening across different fields and intervene immediately if they identify diseased plants. 

One company predicts that by 2050, farms will produce approximately 4.1 million data points each day, and every one of them can be used to grow more and better food. Already, smart farming devices are shaping the future of agriculture by helping farmers increase yields and lower their production expenses. On average, the farms included in one series of studies saw a 1.75% jump in yields and a decrease in energy expenses of up to $13 per acre. 

Smart robots and predictive analytics 

Artificial intelligence (AI) complements agricultural IoT devices and further improves the growing and selling processes via predictive analytics. These programs can help farmers determine which crops to grow and anticipate potential threats by combining historical information about weather patterns and crop performance with real-time data. The more information that’s collected, the more precise the insights producers will be able to glean about what’s happening with their crops and where they can optimize conditions. 

Smart robots will likely come into play as well. Already, companies are developing robots that can analyze crops in the field to not only identify disease indicators but to prune away weak plants to give strong ones the space and resources needed to thrive. One U.S.-based company expects that smart systems, such as the artificial intelligence programs it uses to capture images of tomatoes grown in its greenhouses, will ultimately boost its yields by 20%. 

Blockchain equals better business 

Blockchain technology is revolutionary because of its implications for creating secure, transparent records. This ability will be incredibly valuable for the agriculture industry, where it can be used to establish smart contracts and track food from its origins to grocery stores. An American company recently conducted a massive sale of 60,000 pounds of soybeans to a Chinese buyer all on blockchain. Agricultural deals of this scale often require a great deal of back-and-forth, with multiple agents and decisionmakers involved in the trade. When such sales are conducted using pen and paper, the risk of mistakes and logistical delays is high. 

Blockchain simplifies the process because everyone has access to the same information, such as receipts, letters of credit, and necessary certificates, all of which are stored clearly and accurately for every party to see. The businesses involved in the deal reported a fivefold reduction in time spent on the logistics of the agricultural commodity trade, which they claimed was the first to be completed via blockchain. 

Blockchain may also help curb future food safety crises, such as the e.coli outbreak that caused people in 25 states to become sick after eating contaminated lettuce. Scientists struggled for at least two months to identify where the diseased crops came from, highlighting the need for better supply chain documentation and management throughout the industry. If agricultural growers and their supply chain partners tracked crops on a blockchain, they could easily identify where different crops originated and who was involved in growing and transporting them. 

A major American food seller found that blockchain programs allowed it to trace a piece of the produce’s origins in just over two seconds, compared to the nearly seven days it takes via more manual processes. Not only does the instantaneous nature of blockchain appeal to consumers who are increasingly concerned about where their food comes from, it has the potential to minimize and even halt widespread contamination outbreaks. From a business perspective, this means pinpointing the source and acting precisely to root out the problem, as opposed to fumbling through disorganized data and taking losses on food that was unnecessarily discarded or unsold as a result of the outbreak. 

The future of agriculture 

Agriculture companies have proven that they’re up to the task of producing food for most of the world. Growers and their partners must do even better if they’re to reduce food insecurity and keep pace with population growth. The time to invest in smart technologies and precision agriculture has arrived. Companies that use IoT devices, AI systems, and blockchain stand to improve their yields, reduce costs, and enhance their profits. The business case for tech-enabled agriculture is clear; now it’s a matter of broader adoption and implementation. 

Productivity, efficiency, and growth all measure different aspects of something fairly straightforward: the transformation of inputs (like time or resources) into outputs (ideas, goods, services, solutions). A simple concept with profound implications for individuals and organizations alike.

Digital technology has an important role in personal and organizational productivity. From instantaneous messaging between colleagues and friends to automated Human Resources systems, digital technologies are central to our lives, at home or at work. But there is room for tremendous improvement in the effectiveness of many of these technologies.

In this quick video, Entefy’s Co-Founder Brienne Ghafourifar presents an overview of 4 key challenges faced by individuals and organizations alike in their quest for productivity, efficiency, and growth.

In the U.S., 98% of medical students take a pledge commonly referred to as the Hippocratic oath. The specific pledges vary by medical school and bear little resemblance to the 2,500-year-old oath attributed to the Greek physician Hippocrates. Modern pledges recognize the unique role doctors play in their patients’ lives and delineate a code of ethics to guide physicians’ actions. One widely used modern oath states:

“I will not permit considerations of age, disease or disability, creed, ethnic origin, gender, nationality, political affiliation, race, sexual orientation, social standing, or any other factor to intervene between my duty and my patient.”

This clause is striking for its relevance to a different set of fields that today are still in their infancy: AI, machine learning, and data science. Data scientists are technical professionals who use machine learning and other techniques to extract knowledge from datasets. With AI systems already at work in practically every area of life, from medicine to criminal justice to surveillance, data scientists are key gatekeepers to the data powering the systems and solutions shaping daily life.

So it’s perhaps not surprising that members of the data science community have proposed an algorithm-focused version of a Hippocratic oath. “We have to empower the people working on technology to say ‘Hold on, this isn’t right,’” said DJ Patil, the U.S. chief data scientist under President Obama. The group’s 20 core principles include ideas like “Bias will exist. Measure it. Plan for it.” and “Exercise ethical imagination.” The full oath is posted to GitHub.

The need for professional responsibility in the field of data science can be seen in some very high-profile cases of algorithms exhibiting biased behavior resulting from the data used in their training. The examples that follow here add more weight to the argument that ethical AI systems are not just beneficial, but essential.

1.     Data challenges in predictive policing

AI-powered predictive policing systems are already in use in cities including Atlanta and Los Angeles. These systems leverage historic demographic, economic, and crime data to predict specific locations where crime is likely to occur. So far so good. The ethical challenges of these systems became clear in a study of one popular crime prediction tool. PredPol, the predictive policing system developed by the Los Angeles police department in conjunction with university researchers, was shown to worsen the already problematic feedback loop present in policing and arrests in certain neighborhoods. “If predictive policing means some individuals are going to have more police involvement in their life, there needs to be a minimum of transparency. Until they do that, the public should have no confidence that the inputs and algorithms are a sound basis to predict anything,” said one attorney from the Electronic Frontier Foundation.

2.     Unfair credit scoring and lending

Operating on the premise that “all data is credit data,” the financial services industry is designing machine learning systems that can determine creditworthiness using not only traditional credit-worthiness data, but social media profiles, browsing behaviors, and purchase histories. The goal on the part of a bank or other lender is to reduce risk by identifying individuals or businesses most likely to default. Research into the results of these systems has identified cases of bias like, for example, two businesses of similar creditworthiness will receive different scores due to the neighborhood the business is located in.

3.     Biases introduced into natural language AI

The artificial intelligence technologies of natural language processing and computer vision are what give computer systems digital eyes, ears, and voices. Keeping human bias out of those systems is proving to be challenging. One Princeton study into AI systems that leverage information found online demonstrated that the same biases people exhibit make their way into AI algorithms via the systems’ use of Internet content. The researchers observed, “Here we show for the first time that human-like semantic biases result from the application of standard machine learning to ordinary language—the same sort of language humans are exposed to every day.” This is significant because these are the same datasets often used to train machine learning systems used in other products and systems.

4.     Limited effectiveness of healthcare diagnosis

There is limitless potential for AI-powered healthcare systems to improve patients’ lives. Entefy has written extensively on the topic, including this analysis of 9 paths to AI-powered affordable healthcare. The ethical AI considerations in the healthcare industry emerge from the data that’s available to train machine learning systems. That data has a legacy of biases tied to variability in the general population’s access to and quality of healthcare. Data from past clinical trials, for instance, is likely to be far less diverse than the face of today’s patient population. Said one researcher, “At its core, this is not a problem with AI, but a broader problem with medical research and healthcare inequalities as a whole. But if these biases aren’t accounted for in future technological models, we will continue to build an even more uneven healthcare system than what we have today.”

5.     Impaired judgement in criminal justice sentencing

Advanced artificial intelligence systems are at work in courtrooms performing tasks like supporting judges in bail hearings and sentencing. One study of algorithmic risk assessment in criminal sentencing revealed how much more work is needed to remove bias from some of the systems supporting the wheels of justice. Examining the risk scores of more than 7,000 people arrested in Broward County, Florida, the study concluded that the system was not only inaccurate but plagued with biases. For instance, it was only 20% accurate in predicting future violent crimes and twice as likely to inaccurately flag African-American defendants as likely to commit future crimes. Yet these systems contribute to sentencing and parole decisions.

Entefy previously examined specific action steps for developing ethical AI that companies can use to help ensure the creation of unbiased automation. 

With a technology as impactful as machine learning, it can be difficult to avoid hyperbole. Sure, billions of dollars in investment are pouring into ML projects. Yes, machine learning is a centerpiece of digital transformation strategies. And, to be certain, machine learning is often what people are talking about when they use the umbrella term “AI.” So it’s worth taking the time to look at real-world ML capabilities being developed and deployed at digitally nimble companies around the globe.

Entefy published two previous articles covering the machine learning revolution, which you can access here and here. We continue this global survey below with a look at how 8 more industries are making use of machine learning technology.

Pharmaceuticals & Life Sciences

Wherever you fall on the death disruption debate, we can all agree that aging is a challenging experience. Even if you don’t aspire to immortality, you likely recognize that increased joint pain and susceptibility to illness and injury will erode your quality of life. But deep learning may be able to slow the aging process. Scientists are now using the technology to identify biomarkers associated with aging. Soon enough, a simple blood test could tell you which parts of your body are showing signs of wear and tear, and your doctor could help you mitigate, and perhaps reverse, those affects through lifestyle recommendations and medication.

Food

Up to 40% of a grocer’s revenue comes from sales of fresh produce. So, to say that maintaining product quality is important is something of an understatement. But doing so is easier said than done. Grocers are at the whims of their supply chains and consumer fickleness. Keeping their shelves stocked and their products fresh is a delicate balancing act.

But grocers are discovering that machine learning is the secret to smarter fresh-food replenishment. They can train machine learning programs on historical datasets and input data about promotions and store hours as well, then use the analyses to gauge how much of each product to order and display. Machine learning systems can also collect information about weather forecasts, public holidays, order quantity parameters, and other contextual information. They then issue a recommended order every 24 hours so that the grocer always has the appropriate products in the appropriate amounts in stock. 

Businesses that have implemented machine learning in their replenishment workflows reduce their out-of-stock rates by up to 80%, along with up to 9% in gross-margin increases.

Media & Entertainment

Machine learning allows media companies to make their content more accessible to consumers through automatic captioning systems. Since implementing an automatic captioning program, YouTube has enabled 1,000,000 functionally-deaf Americans and 8,000,000 hearing impaired to watch and enjoy its videos. As of 2017, its machine learning programs had become sophisticated enough to include captions for common non-speech audio, such as laughter and music, creating an even more complete experience.

Information Technology

Although machine learning is generating unprecedented business insights, many organizations have failed to invest adequately in AI systems. For instance, McKinsey found that “the EU public sector and healthcare have captured less than 30 percent of the potential value” of Big Data and analytics. Organizations that want to avoid a similar mistake will need to ramp up their data science abilities – but so will workers who want to stay competitive. By 2020, there will be more than 2.7 million data science jobs, and the demand for workers who understand and can work with machine learning technology will only grow from there.

Law

Deep learning applications are especially impressive in the legal sector due to the nature of the language these programs must parse. Legal phrasing can be complex and difficult to decipher, yet deep learning systems are already capable of analyzing tens of thousands of vital documents. When legal teams needed to dissect contract clauses that upset their or their client’s business and invoicing processes, they once had to manually review stacks of rigorously prepared documents. Now, they can feed them into a program that works far faster than any lawyer and can pick out important phrases for further analysis by the legal team.

Insurance

Improving risk prediction and underwriting is in everyone’s interest, which is why machine learning is such a gift to the insurance industry. In auto insurance, for instance, machine learning algorithms can use customer profiles and real-time driving data to estimate their risk levels. They can then formulate personalized rates based on that information, potentially creating savings for both consumers and insurance companies.

This process may be enhanced by even more in-depth analyses, in which machine learning programs pull in seemingly unrelated social media data to create a more precise profile. The insurance industry could use artificial intelligence to identify which policyholders are gainfully employed and which seem to be in good health. Theoretically, someone who is responsible in those areas of their lives will be a responsible driver as well.

Education

Intelligent tutoring systems (ITS) hold enormous potential for disrupting the classroom and helping students learn. These AI programs serve as virtual tutors, and they adapt their digital lessons based on each child’s strengths and weaknesses. Each time the student completes a task or quiz, a machine learning program processes that information to customize future materials.

By “learning” a user’s unique needs and identifying which types of lessons are most effective for them, the ITS helps the student overcome learning challenges and retain more knowledge. Research indicates that students who use intelligent tutoring systems perform better on tests than their peers who learn via large-group instruction.

For an overview to key concepts in artificial intelligence, check out Entefy’s article Essential AI: A brief introduction to terms and applications.

On May 24th, Entefy CEO Alston Ghafourifar spoke at the 2018 MIT Sloan CIO Symposium in Cambridge, Massachusetts on the topic of: “Building the Intelligent Enterprise using AI, ML, Mobility, and Cloud Services.” The annual event attracts more than 800 CIOs and other senior executives from global organizations of all sizes. 

During a wide-ranging discussion about digital transformation strategies, Alston presented his views on the importance of a multi-modal approach when implementing AI solutions to build intelligence into enterprise systems. While unimodal AI can deliver promising results when applied to direct use cases, Alston argued, they often fail to capture much of the overall value that multi-modal AI systems can harness and apply to a broader set of business processes. Alston shared an analogy with cloud technologies. An enterprise that transitions only its CRM platform to the cloud can capture some business value; but that value may be a fraction of what can be harnessed when applying digital transformation to a broader spectrum of systems. Alston argued that, when it comes to extracting optimal value out of an organization’s lake of data, multi-modal AI systems that leverage multiple domains of AI such as computer vision, natural language, audio, and other data intelligence can significantly outperform unimodal systems. It truly is a case of “greater than the sum of its parts.” 

Over the course of the program, multiple speakers presented their thoughts on the broad trends impacting digital transformation, including AI and machine learning, cybersecurity and data privacy in the post-GDPR world, infrastructure and interconnection strategies, and workload-to-cloud projects using multiple IoT and mobile interfaces. 

On behalf of Alston and everyone here at Entefy, we offer our sincere thanks to the event organizers, symposium attendees, and Ryan Mallory of Equinix, who moderated the discussion. 

You click through to a dated-looking website. A video begins to play with the mysterious title “Real-time Facial Reenactment.” In the video, a tabletop TV shows cable news network footage of President George W. Bush. Beside the TV is a young man seated in front of a webcam, stretching his mouth and arching his eyebrows dramatically like a cartoon character. And then it clicks. The President is making the exact same facial movements as the young man, perfectly mirroring his exaggerated expressions. The seated figure is, to all appearances, controlling a President of the United States. Or at least his image on a video screen. It’s worth checking out the video to see this digital puppetry in action, with 4 million views and counting.

Welcome to the 21st century version of “seeing is believing.” The video in question is the product of a group of university researchers working in a specialty field of artificial intelligence called computer vision. Computer vision algorithms can do things like digitally map one face onto another in real-time, allowing anyone with a webcam to project their own facial expressions onto a video image of a second person. No special media or expensive technology is required. The researchers used an off-the-shelf webcam and publicly available YouTube videos of American presidents and celebrities.

AI-powered computer vision technologies that manipulate video can be impressively realistic, often eerie, and just the sort of thing that should ignite discussions of authenticity in the age of “fake news.”

Photoshop, the verb

The photo editing software Photoshop has been around so long at this point that we use it as a verb. But AI computer vision technologies with the power to manipulate what we see are something else entirely. After all, it’s one thing to airbrush a model in an advertisement and quite another to manipulate what’s said on the nightly newscast.

To see why, consider a few facts. Billions of people worldwide, including 2 out of 3 Americans, rely on social media and other digital sources for news. And in doing so, they overwhelmingly choose video as their preferred format. These informational ecosystems are already struggling with “fake news,” the term for deliberate disinformation masquerading as legitimate reporting.

If AI computer vision enables low-cost, highly convincing tools for manipulating video—for making anyone appear to say anything someone might want them to say—those tools have an extremely high potential for mayhem, or worse. After all, once we see something, we can’t unsee it. The brain won’t let us. Even if later we learn that what we saw was fake, that first impression remains.

AI raises some big questions, and they’re important ones. What happens when we can no longer rely on “seeing is believing?” How can we ensure reliable authenticity in the digital age? Technology is moving fast, and new AI tools remind us that sometimes we need to put aside the cool factor and think instead about how specific technologies impact our lives. Because digital technology is at its best when it improves lives and empowers people.

Putting the dead to work

Part of the reason that AI computer vision technology is so remarkable is that it compresses hours or days of professional work into something that happens in the moment.

The technology has a clear application in entertainment. The production of the immersive worlds and intricate characters we see in film and video games is quite expensive. CGI (computer-generated imagery) is lauded for its ability to make the unreal appear as close to real as possible, to turn imagination into reality. But there is a lot of effort behind these transformations.

CGI may fill theaters, but we’ve also seen that people place limits on the type of unreality they’ll accept. Look no further than Rogue One, the 2016 film set in George Lucas’ Star Wars universe. The film’s story takes place before the events of the original Star Wars film from 1977. The Rogue One filmmakers wanted to feature one of the characters from the original film but ran into a small problem: the actor, Peter Cushing, died in the 1990’s.

CGI to the rescue. The producers employed computer graphics to generate a digital version of Cushing, then brought him to simulated life using motion capture gear and another actor, Guy Henry, to voice Cushing’s character. The result was a rather convincing resurrection of Cushing, which simulated everything from his facial tics to the differences in lighting between films.

If you’ve seen the film, you probably remember this character. Because watching a dead actor brought back to life…isn’t quite right. And not surprisingly the film attracted critics who raised ethical concerns about the dignity of the dead and the right to use a deceased actor’s likeness. It was exactly the feeling of unease we talk about when we talk about the “uncanny valley.”

Rogue One wasn’t the first digital resurrection to raise these concerns. Other reanimations have met with mixed responses from the public. There was Tupac’s holographic performance at Coachella. And Marilyn Monroe, John Wayne, and Steve McQueen posthumously pitching products in commercials. Despite the public’s unease, more than $3 billion is spent annually on marketing and licensing deceased celebrities for advertisements. But none of this suggests that we’re ready to give up “seeing is believing” just yet.

Authenticity is timeless

The invention of Photoshop didn’t cause people to completely distrust every digital photo. We still happily share pictures from vacations and selfies with celebrities without worry that our friends will doubt their authenticity.

The challenge with AI computer vision tech isn’t how it might be misused—after all, practically any technology can be misused. It’s that it joins a growing list of technologies that are developing so quickly that people haven’t had enough time to collectively decide how we want them to be used. This is as true about some forms of AI as it is about robotics and gene editing.

But if we did come together to have this discussion, what we’re likely to find is a lot of common ground around the idea that the best way to use all of these revolutionary technologies is to make life better for people, solve global problems, and empower individuals. After all, for every potential misuse of computer vision AI, there are hundreds of positive and impactful applications like real-time monitoring of crime, improved disaster response, automated medical diagnosis, and on and on.

Fake news is a problem worth solving. But until we’ve successfully leveraged advanced technology in support of truth and authenticity, let’s not abandon “seeing is believing” just yet.

AI augmentation is expected to generate $2.9 trillion in value by 2021, freeing up 6.2 billion hours of worker productivity in the process. Within the AI category, machine learning attracted about 60% of the estimated $8 to $12 billion in external investment in AI capabilities during 2016. Machine learning’s outsized role comes in large part from its usefulness in enabling other capabilities, like robotics, process automation, and speech recognition. Not surprisingly, machine learning sits squarely at the center of digital transformation strategies the world over.

Entefy recently shared a look at ML innovation in 8 different industries. Below we continue our survey of noteworthy machine learning projects, this time in verticals as diverse as banking, nonprofits, energy, and government.

Banking

Machine learning analytics programs offer banks the potential to increase profits substantially. One report found that companies that invested in advanced analytics saw an average profit increase of approximately $369,000,000 a year. They achieve this by making better use of their customer data.

By analyzing client behavior, one institution learned when to intervene before customers disengaged from the company, leading to a 15% reduction in churn. Other organizations used the insights gleaned through machine learning programs to identify new segments within its existing customer base and phase out costly, unnecessary discount practices. Machine learning programs “discover” hidden trends and insights that companies can use to strategize more effectively.

Energy

Deep and machine learning will prove invaluable to combatting climate change and achieving more efficient energy usage. Companies are already using machine learning to sell solar panels in more cost-efficient ways, while researchers predict that cloud-based monitoring systems will be able to optimize energy usage in real time.

Machine learning could also help tech companies reduce their carbon footprints, optimizing energy usage according to a variety of conditions and ultimately decreasing their power draws by up to 15%. This is will be of growing importance as rising demands for computing power raise serious environmental concerns.

Healthcare

With the global medical community facing a shortfall of more than 4 million medical providers, it’s clear that the healthcare system needs help. As Entefy wrote last year, artificial intelligence is here to help doctors. Researchers are exploring the use of machine learning to diagnose disease, a breakthrough that could lead to faster, more accurate patient treatments. 

One group of researchers used machine learning to predict whether patients would be hospitalized due to heart disease. They achieved an 82% accuracy rate, which was 26% percent higher than the average rate using one of the most common existing prediction models. Identifying a patient’s risk for heart disease and hospitalization could allow doctors to make life-saving recommendations before the condition reaches a critical stage.

Retail

Machine learning makes for a more dynamic and interactive retail experience. Instead of wandering store aisles for hours searching for the right tool or outfit or home appliance, home makeover businesses are embracing artificial intelligence to create more personalized – and convenient – brand encounters.

One company used machine learning to create a tool for better home decorating. Customers will soon be able upload photos of their homes and see realistic simulations of what different shades of paint, pieces of furniture, and light fixtures would look like in their living rooms. No more guesswork, no more aggravating trips to the store because you bought the wrong shade of paint. Thanks to machine learning, you could be sure to get it right the first time. 

Government

Machine learning could make the country safer and more equitable, and that’s not just idealism talking. A 2017 study found that a machine learning algorithm was more adept than judges at predicting which defendants were flight risks while awaiting trials. The program assigned risk scores based on details such as defendants’ ages, their rap sheets, the offenses for which they were awaiting trial, and where they had been arrested.

Researchers determined that the program could decrease the number of defendants in jail while awaiting trial by 40% without risking an increase in crime rates. Widespread use of such algorithms would alleviate strains on the criminal justice systems and could even prevent future crimes. The program’s accuracy would serve as a safeguard against judges’ erroneous decisions, as the defendants they release sometimes commit additional crimes before their trials or fail to appear for their court dates.

Nonprofit

Nonprofits are using machine learning to identify trends related to mental health crises, such as indicators for suicide. Importantly, machine learning can make connections that humans might not see, and that information enables crisis counselors to reach people who are in urgent need. In one example, a machine learning program identified that the term “ibuprofen” was a likelier indicator of an imminent threat than the term “suicide.” Therefore, the computer program prioritizes users who have mentioned ibuprofen in their communications to ensure they reach a counselor as quickly as possible.

If you’re just getting up to speed on artificial intelligence, be sure to check out Entefy’s article Essential AI: A brief introduction to terms and applications.