Amazon Web Services Inc. today debuted a suite of Contact Center Intelligence services that customers can use to add more intelligence to their contact center operations.

AWS CCI is a combination of services that enables companies to bring machine learning capabilities including text-to-speech, language comprehension, translation, enterprise search, business intelligence and chatbots to their contact centers. These capabilities can be used to help with tasks such as self-service, live-call analytics and agent assist, and post-call analytics, Amazon said.

In a blog post, AWS developer advocate Alejandra Quetzalli said the AWS CCI services are being made available through partners such as Genesys Telecommunications Laboratories, Inc., Vonage Holdings Corp. and UiPath Inc. and integrated with existing enterprise contact center systems.

The services are designed to aid companies in various aspects of their contact center workflows, Quetzalli said.

For example, the Self-Service solution is meant to help companies create chatbots and interactive voice response systems that can handle some of the most common queries contact centers receive, enabling human operators to focus on other tasks. Quetzalli said customers will need to employ this capability with the company’s Amazon Lex service for building chatbots, and Amazon Kendra, an enterprise search service the chatbots can use to find the answers to people’s queries.

“The novelty of this solution is that Lex + Kendra not only fulfills transactional queries (i.e. book a hotel room or reset my password), but also addresses the long tail of customers questions whose answers live in enterprises knowledge systems,” Quetzalli said.

Meanwhile, Live Call Analytics & Agent Assist enables contact center staff to get a better understanding of what a caller is asking for. The service relies on Amazon Transcribe for real-time speech transcription and Amazon Comprehend to analyze conversations to detect caller’s sentiments and any key phrases that might be pertinent. In addition, Live Call Analytics & Agent Assist can tap into Amazon Translate if the caller is unable to speak the operator’s language.

Finally, the Post-Call Analytics tool carries out automatic analysis of call center conversations and provides insights that can be used to improve the service they offer. It also leverages Amazon Transcribe and Amazon Comprehend, as well as Amazon Kendra.

The post AWS announces AI tools to assist contact center workflows appeared first on Artificial Intelligence.

Organizations failing to adopt AI risk falling behind in their markets. Even without the budget for it, advancing the AI know-how in your organization must include democratizing AI so it can become the province of everyday operations, and not just data scientists.

“‘Democratization’ can be a dirty word,” wrote blogger Chrissy Kidd in 2019. “Some folks hear it and conjure up long conference calls or work meetings abuzz with words like synchronicity and alignment. On the much more positive side, however, democratization is a good thing, where a resource that is useful can become more readily available to the masses.” 

Kidd said many companies are unsure how to begin. “Lots of other companies are behind. They may not know how to start using AI, and they may not have the resources to create their own AI.” 

Not exactly knowing what AI is, what it can and can’t do, and how to expand its use to different areas of the organization is precisely where most companies are. In the midst of the COVID-19 crisis, companies are also concerned about operating budgets and how much they can invest. In this environment, a discretionary expense such as internal staff training is one of the first items in the budget to get scaled back.

One way to keep AI momentum moving forward while also conserving budgets is by taking advantage of free AI educational resources that can help employees at all levels of the organization learn.

Here are two great, free ways to do this.

AI courses

Yes, you can take free AI courses to begin your education or sharpen your AI skills.

• Google offers a free online Machine Learning Crash Course. It’s 15 hours of work consisting of 25 lessons and 40 exercises. There’s also a free four-hour introductory course to machine learning. 
• Free courses are available for getting started with IBM Watson as well, and also for education on Watson’s cognitive learning functions. 
• Microsoft offers a free AI bootcamp for business executives.

AI tools

The open-source community is rich with free AI tools that enable employees to get involved with AI. Many of these tools provide on-ramps into AI application development methods that can be plugged directly into the company’s IT and data science work with AI.

Here is small sampling of some of the free AI tools that are are available as open source:

• IBM Watson Studio Desktop is available to academic and educational institutions as free, open-source AI software on the cloud. The tool covers AI and subsets of the discipline such as machine learning. Watson is an established platform that students and staff can experiment with as they develop their AI skills. Watson also offers code-free analysis tools, making it easier for non-IT and data science staff to experiment.
• Apache Mahout is a free AI tool for data mining in Hadoop, Facebook, Foursquare, Twitter, LinkedIn, and Yahoo. It is especially useful in online sales and marketing, given its ability to use purchase recommender engines, along with data filtering that can target users and customers.
• OpenNN provides rapid AI data processing speeds and free neural network libraries. Its sweet spot is dealing with machine learning, advanced algorithms, and the design of neural networks. It has delivered impactive results in the utilities and infrastructure sectors. OpenNN focuses on AI, machine learning, and predictive analytics. It is an open-source tool best used by IT and data science professionals.

There are free AI resources and tools out there that can help you and your organization enrich your AI knowledge. At the same time, you can be mindful of the financial bottom line of your company as you continue to move forward with your AI.

The post How to get free AI training and tools appeared first on Artificial Intelligence.

Public and private sector enterprises need to consider expanding their use of AI-augmented cybersecurity tools to better defend their networks and assets, say experts in a new podcast.

As the range of cyberthreats continues to expand, and organizations remain hard-pressed to hire enough talent to keep up, cyber experts recommend that executives explore AI tools to help assess and automate their security posture.

Security veteran Irfan Saif says that AI represents a range of different concepts such as intelligent automation, analytics and conversational AI as well as more sophisticated capabilities that start to approach what may be considered human intelligence, he says.

Saif, a principle and board member at Deloitte, also urges enterprise executives to think about AI in the context of machines helping humans, which he sees as “a much more viable, sustainable and scalable approach rather than thinking about AI in the context of human replacement,” he explains.

Adding to the discussion, Deborah Golden, lead for Deloitte’s U.S. Cyber Risk Services Practice says that this idea of partnership between people and AI-enabled technology will help organizations address the shortage of cybersecurity talent.

Golden and Saif share recommendations for executives leading public and private sector organizations on ways AI can combat new cyberthreats in the latest episode of the “Cyber Everywhere” podcast series produced by CyberScoop and underwritten by Deloitte:

Changes occurring in the cyberthreat landscape

“Bad actors — particularly those on the more sophisticated end of the spectrum — tend to adopt and adapt to changes in the technology landscape a bit faster than those that they are trying to attack,” says Saif.

He cautions that AI is being used against enterprises, noting instances where AI has been used to mimic the activity of legitimate users and bypass various detection measures, he says.

Business case for AI-enabled tools

Golden says CIOs need to consider adopting AI-enabled tools to help available cyber talent achieve greater efficiencies at scale.

As the cyberthreat landscape continues to grow at exponentially, enterprises will need to keep investing in “structured and unstructured machine learning in a way that perhaps we’ve never looked at before,” just to keep pace, she says.

Developing strategies and governance for AI

Saif says that the notion of “trustworthy AI” is gaining currency among security experts. The goal is to build a common language and framework to govern AI as a strategy and as a program “from the boardroom down to the server room.”

“That is effectively taking critical principles of trust — whether that’s ethics, whether that’s explainability — all the sorts of things that people really want to understand when it comes to how to apply AI to business problems, how to manage and govern the data, and the inputs, the outputs and the use of that information,” Saif says.

Irfan Saif currently co-leads Deloitte’s U.S. artificial intelligence and cognitive advisory offering. He has more than 20 years of IT consulting experience, specializing in cybersecurity and risk management.

Deborah Golden has more than 25 years of IT experience spanning numerous industries, including government, life sciences, health care and financial services. She specializes in cybersecurity, technology transformation and privacy and governance initiatives.

Listen to the podcast for the full conversation on AI-augmented cybersecurity. You can hear more coverage of “Cyber Everywhere” on our CyberScoop radio channels on Apple Podcasts, Spotify, Google Play, Stitcher and TuneIn.

This podcast was produced by CyberScoop and underwritten by Deloitte.

The post Cyber Everywhere: Expand security capabilities with AI tools appeared first on Artificial Intelligence.

n an effort to assist lenders expedite the processing of functions for the U.S. Small Enterprise Administration’s (SBA) Paycheck Safety Program, which goals to maintain staff employed throughout the coronavirus pandemic, Google developed an AI resolution known as PPP Lending AI that integrates with present doc ingestion instruments. It’s accessible to eligible lending establishments by June 30.

As Google explains in a whitepaper, AI can automate the dealing with of volumes of mortgage functions by figuring out patterns that will take a human employee longer to identify. Particularly, PPP Lending AI can classify and extract information in essential paperwork earlier than readying paperwork for submission to the SBA.

PPP Lending AI, which Google says takes solely days to implement, is an answer in three elements.

The primary is the Mortgage Processing Portal, a web-based app that serves as a consumer interface and self-servicing middle. Along with offering administration views for mortgage officers and mortgage processors, it permits finish customers and mortgage candidates to create, submit, and look at the standing of their PPP mortgage.VB Transform 2020 Online – July 15-17, 2020: Be part of main AI executives at VentureBeat’s AI occasion of the yr. Register today and save 30% off digital entry passes.

The second piece of PPP Lending AI is the Doc AI PPP Parser, which permits lenders to make use of AI to extract structured data from mortgage paperwork submitted by the mortgage candidates. It’s constructed atop Google Cloud‘s Doc AI, a service that leverages optical character recognition, type parsing, and pure language processing to seize and enrich unstructured information.

The third is Mortgage Analytics, which lets servicers or lenders onboard structured historic mortgage information, carry out de-identification anonymization on delicate data, retailer it securely with fine-grained information entry management, and carry out queries on the info.

“Leveraging synthetic intelligence, we’ve created an end-to-end resolution that accelerates the time-to-decision on loans and helps inform lenders’ liquidity evaluation — from the preliminary utility submission to the underwriting course of and SBA validation,” wrote Google Cloud world monetary providers and options lead Christin Brown in a blog post. “The answer can also be geared up with Google’s safety capabilities, enabling lenders to satisfy coverage necessities and shield essential belongings.”

Google says lenders can converse with a Google Cloud account supervisor for extra data.

PPP Lending AI seems to skirt round a newly imposed U.S. Treasury and SBA rule prohibiting the submission of PPA loans ready by robotic course of automation (RPA), or AI programs that carry out repetitive, monotonous duties at scale with larger velocity and accuracy than people. The companies blamed RPA for overburdening E-Tran, the SBA’s digital mortgage servicing portal, and lowering its capabilities.

On Monday, E-Tran crashed minutes after the opening of $310 billion in further PPP funding. The funds had been accredited final week following the primary $349 billion spherical, which ran out in early April. That’s regardless of the truth that the SBA restricted utility submissions to 350 per hour and allowed banks with a minimal of 5,000 loans to bulk-file their functions.

PPP loans can be found to small companies that had been in operation as of February 15 with 500 or fewer staff, together with not-for-profits, veterans’ organizations, tribal issues, self-employed people, sole proprietorships, and unbiased contractors. Companies with greater than 500 staff in sure industries may apply for loans, in accordance with the SBA and Treasury.

The post Google releases AI tool for processing Paycheck Protection Program loans appeared first on Artificial Intelligence.

Members of the Google Brain team and Google AI this week open-sourced EfficientDet, an AI tool that achieves state-of-the-art object detection while using less compute. Creators of the system say it also achieves faster performance when used with CPUs or GPUs than other popular objection detection models like YOLO or AmoebaNet.

When tasked with semantic segmentation, another task related to object detection, EfficientDet also achieves exceptional performance. Semantic segmentation experiments were conducted with the PASCAL visual object challenge data set.

EfficientDet is the next-generation version of EfficientNet, a family of advanced object detection models made available last year for Coral boards. Google engineers Mingxing Tan, Google Ruoming Pang, and Quoc Le detailed EfficientDet in a paper first published last fall, but revised and updated it on Sunday to include code.

“Aiming at optimizing both accuracy and efficiency, we would like to develop a family of models that can meet a wide spectrum of resource constraints,” the paper, which examines neural network architecture design for object detection, reads.

Authors say existing methods of scaling object detection often sacrifice accuracy or can be resource intensive. EfficientDet achieves its less expensive and resource-hungry way to deploy object detection on the edge or in the cloud with a method that “uniformly scales the resolution, depth, and width for all backbone, feature network, and box/class prediction networks at the same time.”

“The large model sizes and expensive computation costs deter their deployment in many real-world applications such as robotics and self-driving cars where model size and latency are highly constrained,” the paper reads. “Given these real-world resource constraints, model efficiency becomes increasingly important for object detection.”

Optimizations for EfficientDet takes inspiration from Tan and Le’s original work on EfficientNet. and proposes joint compound scaling for backbone and feature networks. In EfficientDet, a bidirectional feature pyramid network (BiFPN) acts as a feature network, and an ImageNet pretrained EfficientNet acts as the backbone network.

EfficientDet optimizes for cross-scale connections in part by removing nodes that only have one input edge to create a simpler bidirectional network. It also relies on the one-stage detector paradigm, an object detector known for efficiency and simplicity.

“We propose to add an additional weight for each input during feature fusion, and let the network to learn the importance of each input feature,” the paper reads.

This is the latest object detection news from Google, whose Google Cloud Vision system for object detection recently removed male and female label options for its publicly available API.

The post Google AI open-sources EfficientDet for state-of-the-art object detection appeared first on Artificial Intelligence.

Google is working on an AI tool for mammograms that researchers hope will one day be more accurate than human radiologists. The tech giant paid for a study, the results of which were published last week (Jan.1) in Nature.

Its findings, at first glance, look promising. But experts caution that AI has a long way to go before it can replace a trained human—especially when it comes to accurately spotting breast cancer in diverse racial and ethnic populations.

Google first trained its AI on mammograms from 76,000 women in the UK and 15,000 women in the US, all who had received a positive diagnosis of breast cancer. It then tested its tool against mammograms of 28,000 other women, only some of whom had breast cancer, and compared its assessment to that of human radiologists. On scans from the US, Google’s AI cut false negatives by 9.4% and false positives by 5.7%; on scans from Britain, the AI cut down those errors by 2.7% and 1.7%, respectively.

But the Google study doesn’t account for the racial makeup of the women studied, which has given researchers pause. A Google spokesperson told Quartz that while information on race was originally collected by healthcare providers, its research team didn’t have access to such identifying details in the data made available to them.

“We already know that there are widespread racial inequalities in the healthcare system that have life or death consequences for patients,” says Sarah Myers West, a postdoctoral researcher at the AI Now Institute at New York University. Before AI tools like the one created by Google are deployed more widely, West says, it’s critical that they don’t reflect or even amplify those inequalities.

In England, for example, black women are more than twice as likely to be diagnosed with advanced breast cancer as their white counterparts, according to a 2016 analysis from Cancer Research UK and Public Health England. Although black women are less likely to suffer from breast cancer as a whole, they have lower survival rates, according to data from the NHS Cancer Intelligence Network.

Researchers are still exploring why black women with breast cancer are more likely to get aggressive tumors than white women; black women in the UK are less likely to get breast cancer screenings, and public health campaigns neglect to target them specifically. Breast cancer rates for South Asian women in the UK have also risen in recent years, putting the group at an 8% higher rate of breast than white women, according to a University of Sheffield study.

In the US, the racial disparity is similar. White women in the US have a slightly greater risk of contracting breast cancer than their black counterparts, but lower mortality rates, according to data from the Kaiser Family Foundation. This trend isn’t reflected in Hispanic, Asian, and Native women, who experience the lowest diagnosis and mortality rates.

For AI to accurately predict breast cancer for all women, it’s crucial that the data used to train the system is reflective of the wider population. But these underlying disparities in the treatment of breast cancer mean that preexisting data sets don’t represent all populations.

“Artificial intelligence systems work by identifying patterns in large pools of data—in this case, looking for patterns in mammogram images and linking them to diagnoses—in order to make predictions about future data,” said West. Both the US and the UK, despite their diverse populations, are still countries where the majority identify as white. Training an AI model on a data set from these nations, without weighing race, stands to perpetuate existing racial biases.

There’s been some work on remedying racial bias in AI. Last year, MIT created a AI model that it claimed could predict breast cancer effectively in black and white women. But MIT’s model still relied on a relatively small number of non-white participants. Of the women whose mammograms were used to train the model, 81% were white, 5% were black, 4% were Asian, and 8% were either marked as “other” or “unknown.”

Even if these issues were addressed, AI diagnosis wouldn’t be right for everyone. The value of mammograms—whether they’re evaluated by a human or a machine—is fiercely debated. Recent studies suggest that routine screenings can lead to overdiagnosis of breast cancer, identifying growths that don’t necessarily develop into life-threatening cancers. In those cases, treatments including radiation and surgery may prove more risky than the growths themselves.

Fortunately, Google’s model isn’t ready for prime time yet. Google plans to work with additional partners around the world to build its data set before making it available to hospitals. “We will continue to explore and build upon our model, working with additional partners across the world, before considering bringing it into clinical practice,” said Shravya Shetty, a Google researcher who co-authored the paper, in an email to Quartz. Google didn’t clarify how future models would address racial diversity.

The post Google’s AI for mammograms doesn’t account for racial differences appeared first on Artificial Intelligence.

Emerging technologies such as big data and AI can be daunting for the small business owner, but there’s also significant pressure to stay ahead of the curve. Small businesses are always competing to keep costs and overheads low, as they’re less able to sustain themselves through the lean times. However, McKinsey research has shown that implementing big data strategies can significantly impact the bottom line. For example, a retailer could increase operating margins up to 60 percent.

Implementing big data and AI platforms doesn’t need to involve hiring teams of data scientists or analysts. At this point in 2019, small businesses have a wealth of out-of-the-box solutions available to them, covering a variety of use cases.

Chatbots are helping to reduce operational costs by providing a “human” element to first-line customer services and general inquiries. Juniper Research estimates that chatbots will save businesses around $8bn by 2022. Bots can also help with lead generation and conversion, by providing a direct route from the company website to a sales representative for those who choose to engage with the bot online.

Now in 2019, many chatbots operate cross-channel, integrating seamlessly with Facebook Messenger, WhatsApp, Instagram, and more. Liveperson is one example, with a chatbot builder feature that makes it easy for anyone in the organization to build and optimize bots using industry-specific templates. Using Liveperson, a business can automate up to 70 percent of messaging conversations across various platforms.

Predictive Analytics

Predictive analytics platforms have a broad range of applications in companies, including reducing employee turnover and decreasing risks such as cyberattacks. However, the implications for sales and marketing are particularly significant. By understanding which demographics are likely to buy a product, sales and marketing teams can make more efficient use of their budgets.

Endor is one company with a predictive analytics platform that’s as easy to use as a Google search.

The user only needs to type in their question, which could be something as straightforward as “who is most likely to buy x product?” Endor’s data science methodology relies on the “social physics” discipline, to deliver fast and accurate answers based on crowd wisdom. It was developed by a team from MIT, who pioneered the concept of social physics in an academic setting and are now applying it in across a range of industry sectors including retail and finance.

Business Intelligence

Internal business systems are often fragmented, with different software for managing sales, customer services, human resources, and accounting. However, together they generate a vast amount of data that’s greater than the sum of its parts and can be converted into actionable business intelligence. 

This is the goal of Insight Squared. It takes historical data from internal company systems and analyzes it in aggregate to generate recommendations for sales, marketing, and staffing. Users have access to multi-dimensional reports and analytics to help manage pipelines, gain more accurate forecasts of sales or product usage, and clearer visibility into marketing-generated demand.

Recruitment

Recruiting new staff is a considerable cost for any business, taking up a vast amount of management time. Among a 2017 Wasp Barcode survey of 1,100 small business owners, fifty percent stated that their top challenge was hiring new employees.

Developments in AI mean that hiring managers can significantly cut down on the legwork involved in the recruiting process, particularly in the earliest stages of sifting through dozens of applications. Ideal offers a suite of AI-powered tools that will integrate with existing HR software to enable data-backed hiring decisions and make the recruiting process more efficient. Ideal can pre-screen candidates, engage with them online via a chatbot, and automate tedious tasks such as sending out interview requests and rejection letters. 

Visual Analytics

For a bricks-and-mortar business, visual analytics can provide powerful insights. Most stores these days are equipped with security cameras both front and back of house, usually to deter thieves. However, platforms such as Prism can put these cameras to work harvesting all kinds of data to help make better business decisions.

For example, Prism can produce heat maps showing how customers move around a store, helping to ensure optimal placing of merchandise for maximizing revenues. It can also analyze footfall to tell a retailer when the busiest periods will happen so they can arrange adequate staffing. Behind the scenes, the software can also assist with inventory management and help to spot theft.

These are just a few examples of the available AI and analytics tools and platforms on the market today, and none of them require extensive technical expertise to operate. Each of them illustrates how small businesses can start to seamlessly incorporate AI and big data to help reduce business overheads and increase profitability. The point is not to implement technology for its own sake but to find methods of incorporating technology into your business in such a way that it enhances your edge over the competition and helps secure the future of your business

The post Big Data Platforms and AI Tools for Your Small Business appeared first on Artificial Intelligence.

Experts around the world are getting increasingly worried about new AI tools that make it easier than ever to edit images and videos — especially with social media’s power to share shocking content quickly and without fact-checking. Some of those tools are being developed by Adobe, but the company is also working on an antidote of sorts by researching how machine learning can be used to automatically spot edited pictures.

The company’s latest work, showcased this month at the CVPR computer vision conference, demonstrates how digital forensics done by humans can be automated by machines in much less time. The research paper does not represent a breakthrough in the field, and it’s not yet available as a commercial product, but it’s interesting to see Adobe — a name synonymous with image editing — take an interest in this line of work.

Speaking to The Verge, a spokesperson for the company said that this was an “early-stage research project,” but in the future, the company wants to play a role in “developing technology that helps monitor and verify authenticity of digital media.” Exactly what this might mean isn’t clear, since Adobe has never before released software designed to spot fake images. But, the company points to its work with law enforcement (using digital forensics to help find missing children, for example) as evidence of its responsible attitude toward its technology.

The new research paper shows how machine learning can be used to identify three common types of image manipulation: splicing, where two parts of different images are combined; cloning, where objects within an image are copy and pasted; and removal, when an object is edited out altogether.

To spot this sort of tampering, digital forensics experts typically look for clues in hidden layers of the image. When these sorts of edits are made, they leave behind digital artifacts, like inconsistencies in the random variations in color and brightness created by image sensors (also known as image noise). When you splice together two different images, for example, or copy and paste an object from one part of an image to another, this background noise doesn’t match, like a stain on a wall covered with a slightly different paint color.

As with many other machine learning systems, Adobe’s was taught using a large dataset of edited images. From this, it learned to spot the common patterns that indicate tampering. It scored higher in some tests than similar systems built by other teams, but not dramatically so. However, the research has no direct application in spotting deepfakes, a new breed of edited videos created using artificial intelligence.

“The benefit of these new ML approaches is that they hold the potential to discover artifacts that are not obvious and not previously known,” digital forensics expert Hany Farid told The Verge. “The drawback of these approaches is that they are only as good as the training data fed into the networks, and are, for now at least, less likely to learn higher-level artifacts like inconsistencies in the geometry of shadows and reflections.”

These caveats aside, it’s good to see more research being done that can help us spot digital fakes. If those sounding the alarm are right and we’re headed to some sort of post-truth world, we’re going to need all the tools we can get to sort fact from fiction. AI can hurt, but it can help as well.

The post Adobe is using machine learning to make it easier to spot Photoshopped images appeared first on Artificial Intelligence.

A few weeks ago, the Kenyan minister of technology was promoting it and talking it up to his government. The U.S. government has increased the funding for it by forty percent since 2015. It is a State priority in China, where it is being used for various government processes including catching law breakers. It is the buzzword in all of technology today. It is Artificial Intelligence (AI). This column in Daily Trust has written a few articles on the facial recognition component of AI, including those on 17 July 2017 and 19 February 2018, but AI is much more expansive and it is surely getting bigger by the day. I’ll dig deeper into this technology paradigm in today’s article; as they say: “where attention flows, results show.”

Artificial intelligence, also sometimes referred to as Machine Learning (ML), is computer-based. It enables computers to see inconspicuous patterns in a dataset to improve performance (“learning”) progressively without being explicitly programmed. In this way the computer is made to dig up stuff that would otherwise remain hidden inside of large amounts of data, and provide you with information that you can act on. There is some historical perspective to AI, for example, as recently described by Michael Kratsios, deputy assistant to the president of the U.S. on technology: “In the summer of 1956, a dozen American scientists gathered on Dartmouth’s campus with the goal to, ‘find how to make machines solve the kinds of problems now reserved for humans. Now, nearly 62 years later, the age of artificial intelligence is here, and with it the hope of better lives for the American people.”

There are different kinds of AI. One of the most commonly deployed these days is the Facial Recognition technology (FRT), or the technology of recognizing people’s faces. As stated in this column on 17 July 2017, the concept is quite simple: At some point, a picture of your face would have been taken by someone or some machine, and specific geometric/topological features of your face extracted and stored along with your personal data, bio data, or metadata. By searching through such a database and making comparisons, any newly captured facial features can be evaluated to explore a match and present identification. The technology is being used for verifying customer identities in business transactions such as banking. Installing FRT hardware in students’ dormitories can discourage criminals. Facial-recognition technology is also quite useful in tailoring marketing efforts based on some database. Security and surveillance represents a critical area where FRT is also very useful. Here, the identities of people are obtained before they are allowed entry into a building or are matched with some information in a database for the purpose of catching criminals. Also, a few universities are installing FRT to identify ghost exam takers trying to sit exams for other students. Law enforcers love FRT.

There are other types of AI. At its 2018 Build Developers Conference in Seattle, USA, 7-9 May 2018, Microsoft unveiled the prototype of a next generation stenographer, which combines facial and audio recognition technology within the context of artificial intelligence. Caitlin Fairchild describes this as follows in the 10 May 2018 article in Nextgov.com: “The device recognizes employees as they walk into a conference room and microphones capture audio of the meeting as it takes place, and AI tools then transcribe everything being said in real time. The AI is even able to learn about team members’ speech patterns over time to improve transcription accuracy. If you have a meeting between people who speak different languages, the device can provide translations.” The system reportedly has an accessibility tool that helps those with hearing loss better participate in the meeting.

The use of AI by tax auditors is different from the two examples described above. Here, standard business rules and various types of ML are combined to give a result that is better than the sum of its parts. In predictive modeling – also called supervised learning, – tax agencies use all past fraud and audit cases to figure out patterns that are correlated with successful cases. “Unsupervised Learning,” such as “clustering,” has also been deployed, whereby the computer automatically puts all tax returns into groups that have similarities (clusters) and then targets returns falling outside these clusters as outliers that require additional scrutiny.

There are a few important things to note about AI. Its success in various deployments depends on the availability of data, which is required to train the algorithms. For example, even though the technology might have been conceived and developed in the West, a Chinese AI tool, such as SenseTime, may be more accurate than similar tools developed in America because the former has more data for training its software. The other issue is related to bias, such as the racial bias situation of FRT tools alluded to in the 19 February 2018 article in this column in Daily Trust. In addition, the public needs to be informed, to avoid scandals such as the revelation that the city of New Orleans, USA, secretly used the so-called Palantir’s algorithms for predictive policing for six years until this was uncovered. As it were, deploying awesome technology comes with awesome responsibilities.

The post Artificial Intelligence is the new technology frontier appeared first on Artificial Intelligence.

The post Google pitches artificial intelligence to help unplug appeared first on Artificial Intelligence.