aianalytics

[aianalytics]

EU MOVES TOWARDS DEFEATING TOXIC AI ALGORITHMS THAT HARM HUMANITY

Artificial intelligence is currently disrupting and revolutionizing almost every global industry. With the development of several advanced technologies revolving around AI, the technology is gaining the potential to improve several aspects of our lives. But it does not come without risks! As technology progresses, the amounts of AI algorithms are on the rise! Swathes of experts are warning users about the potential dangers that revolve around artificial intelligence. There are several tech fanatics who believe these claims are false and poorly exaggerate the impact of AI on our daily lives, but it seems like the EU takes up these warnings quite seriously, as it moves forward to curbing all AI threats to humanity.

Politicians serving the European Union are planning to introduce its first comprehensive global infrastructure for regulating AI. Since more global institutions are moving towards automating routine and monotonous tasks in an attempt to boost efficiency and cut operational costs, more businesses are falling prey to the harmful AI algorithms that aim to threaten them. The upcoming framework termed the Artificial Intelligence Act, will be levied across all EU borders, like the EU’s General Data Protection Regulation that applies to all areas including UK banks and the ones who serve the EU legislation.

EU officials expect that the extra regulation and monitoring of AI algorithms on the different types of AI models that can be produced and used will definitely adjust the kind of machine-based discrimination that influence the most important decisions in our lives. With this upcoming regulatory infrastructure, experts believe AI users might experience massive benefits, starting from open access to quality, truthful news, and others.

Source

[aianalytics]

TOP HUMAN BRAIN INSPIRED AI PROJECTS TO KNOW IN 2021

Brain-inspired AI projects are helping in the advancements of artificial intelligence

Scientists and researchers are continuously working on artificial intelligence and multiple AI technologies such as neural networks to create advanced products and services for the welfare of society. AI projects are thriving in the tech-driven market from multiple reputed tech companies and research centers to improve and discover new ventures in this domain. Thus, it has led to hundreds of human-inspired AI projects available on the internet to gain sufficient knowledge of smart functionalities of artificial intelligence. Let’s explore some of the top human brain-inspired AI projects to approach in 2021.

Top Human Brain-Inspired AI Projects

OpenNN

OpenNN is one of the top human brain-inspired AI projects that helps to build the most powerful AI models with C++. It is known as an open-source neural network library for machine learning and artificial intelligence available for multiple industries. It consists of sophisticated algorithms to develop artificial intelligence solutions for AI projects.

A(rtificial)Human

a(rtificial)Human is a human-inspired AI project that was started in 2008 to implement human personality with the integration of a computer program. This AI project utilized strong computer science background with artificial intelligence knowledge. AI models help to learn about neurobiology and approaches assisting in building software programs.

Numenta Platform for Intelligent Computing

Numenta Platform for Intelligent Computing is of the top brain-inspired AI projects consisting of a set of learning algorithms. Learning algorithms are known for capturing different layers of neurons for neural networks in artificial intelligence. Visual pattern recognition, NLP, object recognition, and many more can be done by human brains with the help of the neocortex. This AI project helps the machines to approach and take over human-level activities efficiently and effectively.

Neu

Neu is known as a C++ framework with a collection of multiple programming languages as well as multi-purpose software systems. These help in creating artificial intelligence applications, AI models with simulations, and other technical computing for further advancements in AI technologies.

AILEENN

AILEENN is known as Artificial Intelligence Logic Electronic Emulation Neural Network. This is one of the top human brain-inspired AI projects that act as a cloud-based PaaS platform and IaaS infrastructure based on neural networks and fuzzy logic. This AI project helps in the decision-making process in this tech-driven world.

Visual Hierarchical Modular Neural Network

Visual Hierarchical Modular Neural Network is a brain-inspired AI project that visually constructs a human thought process and logic with a flow to generate artificial intelligence automation. It provides a wide range of innovative and user-friendly tools frameworks that can integrate professional human-like decision-making into commercial systems. It also protects users from mathematical associated with neural networks as well as artificial intelligence algorithms.

Source

[aianalytics]

China up for global domination in AI, ML; US stands no chance: Pentagon's former software chief

The United States has already lost the artificial intelligence battle to China which is heading towards global dominance because of its advances in a emerging cyber capabilities, the Pentagon's former software chief told the Financial Times.

Nicolas Chaillan told the FT that he resigned in protest at the slow pace of technological transformation in the US military and because he could not stand to watch China overtake America.

"We have no competing fighting chance against China in 15 to 20 years. Right now, it’s already a done deal; it is already over in my opinion," he told the FT.

Beijing heading for global dominance

Chaillan said that Beijing is heading for global dominance because of its advances in artificial intelligence, machine learning and cyber capabilities.

Emerging technologies

He argued these emerging technologies were far more critical to America’s future than hardware such as big-budget fifth-generation fighter jets such as the F-35.

Kindergarten level defences

Chaillan further added US cyber defences in some government departments were at “kindergarten level".

Blames Google

Chaillan compares US tech giants with their Chinese peers. He blames the reluctance of Google to work with the US defence department on AI, and extensive debates over AI ethics for slowing the US down.

He adds that Chinese companies are obliged to work with Beijing, and were making “massive investment" into AI without regard to ethics.

Source

[aianalytics]

PROFITABLE INVESTMENT: TOP ARTIFICIAL INTELLIGENCE STOCKS TO BUY IN OCTOBER 2021

Analytics Insight describes top artificial intelligence stocks to buy in October

Artificial intelligence is showing its inevitable functions through AI models in multiple industries across the world in these recent years. Tech companies are highly instigated to leverage artificial intelligence to gain a competitive edge in the market with enhanced customer satisfaction and better customer engagement while manufacturing AI models. Investors tend to be in a risky position in the cryptocurrency market due to its volatility with the cryptocurrency prices. But artificial intelligence stocks provide stability to gain higher revenue in the nearby tech-driven future. Thus, let’s explore some of the top AI stocks in October to provide growth in revenue to investors.

Top artificial intelligence stocks to buy in October

Persistent Systems

Market cap: US$271.77 billion

Persistent Systems is one of the popular tech companies in India and artificial intelligence stocks for investors in October 2021. It is known as a trusted digital engineering and enterprise modernization partner. It offers a wide range of services such as digital business strategy, digital product engineering, CX innovation and optimization, data-driven business and intelligence, identity, access, and privacy, and core IT modernization in the tech field on AI models. It provides these services to multiple industries like banking, financial services, and insurance, healthcare, and life sciences, industrial, software, and hi-tech, as well as telecom and media. Recently, this tech company has announced a dedicated payments business unit and expansion in cloud capabilities through strategic acquisitions while leveraging artificial intelligence.

Oracle

Market cap: US$245.32 billion

Oracle is one of the well-known AI stocks in October that has outperformed the tech market on strong trading day. It offers a wide range of products such as Oracle Cloud Infrastructure and Oracle cloud applications. Infrastructure includes software, hardware, and featured products on AI models while applications include cloud applications, industry solutions, NetSuite, and on-premised applications. It provides these services to multiple industries like automotive, communications, construction and engineering, consumer goods, financial services, hospitality, government and education, retail, and many more. Thus, artificial intelligence stock from this tech company is stable to gain higher revenue after buying in October 2021.

Zensar Technologies

Market cap: US$ 108.86 billion

Zensar Technologies is another artificial intelligence stock that investors can buy in October 2021. This AI stock in October is a technology consulting services company to more than 130 leading enterprises. It offers expertise in conceptualizing, designing, engineering, and managing digital products through innovation in artificial intelligence in AI models. It serves multiple industries such as hi-tech, banking and financial services, insurance, healthcare, and many more.

TD SYNNEX

Market cap: US$10.21 billion

TD SYNNEX is a popular AI stock in October for providing business process services by leveraging artificial intelligence. It provides system components, consumer electronics, virtual distribution, online services, cloud services, marketing services, telemarketing campaigns, and many more. The tech company is known for offering technology products, services, and solutions to the world through cutting-edge technologies such as artificial intelligence.

The Trade Desk

Market cap: US$33.71 billion

The Trade Desk is one of the top artificial intelligence stocks that operates a self-service cloud-based platform to allow consumers to create and optimize data-driven digital advertising campaigns. This tech company holds a huge potential for growth in October because investors have observed that it continues to benefit from mobile and desktop advertising rather than radio and print media, as per the digital transformation.

Source

[aianalytics]

HOW CAN AI AND DATA SCIENCE MAKE IPL 2021 MORE INTERESTING?

Watch out how AI and data science are making the current IPL more interesting.

Cricket and technology work hand in hand. Technology helps in tracking things ball speed, camera in the stump, third umpiring, etc. And now this has taken a new form. In this article, we will focus on IPL 2021 and the use of AI and data science.

Without an umpire we cannot imagine cricket, right? But with technology, this won’t be so far. Soon there would be no need for an umpire to judge the game and AI would have completely replaced the requirement. Ball by ball predictions is now being made using some very innovative aspects of AI. Prediction models are in place, where the data is sustained in and matched with previous data that can analyze the outcome.

IoT enabled bats are becoming a reality, where a sensor on a bat can detect everything from the twist of the ball to the speed of the ball to the quality of a batsman’s shot and this data is relayed for analysis. Team coaches can now clearly analyze the capability of the batsman and can give real-time inputs for improvements. Winning and scoring predictions are becoming more and more accurate making sports betting a lucrative industry for many.

The ongoing IPL is also boosted by AI and data science. Let’s see how AI and data science are making the current IPL more interesting.

Automated Sports Journalism

AI-driven platforms have been designed that can translate hard data into narratives using natural language. Multiple smart cameras can now detect hits, misses, boundaries, scores, centuries, etc., and relay the information directly to the network, without the need of an intermediary to collate information.

Broadcasting and AI Targeted Ads

In addition to revolutionizing the world of sports for coaches and players, AI also has a sizable impact on the way the audience experiences sports. AI systems can be used for automatically choosing the right camera angle to display on viewers’ screens, providing subtitles for live events in different languages based on the location of the viewer, and also enabling broadcasters to utilize monetization opportunities through advertisements. Traditionally advertisements that we see are globally published ads by the network that is relaying. However, targeted & localized ads are now being shown on the field side banners by overlapping the original ads.

Data-Driven Analysis and Augmented Coaching

AI continues to have a significant impact on the strategic decisions made by coaches, before, during, and after the game. With the help of wearable sensors and high-speed cameras, AI platforms measure a forward pass, a penalty kick, LBW in cricket, and a lot of similar actions in various sports. This data empowers coaches to prepare players better for competition. This data-driven analysis of players along with the quantitative and qualitative variables helps coaches to develop better training programs for their teams.

Player Performance Improvement

AI is also being used to enhance the performance of players. Apps like HomeCourt make use of computer vision and machine learning to assess basketball players’ skills, giving them a good medium to improve. The recording of these performance metrics of the athletes is not only credible but also helps the players to understand the areas where they have maximum potential to excel and the areas that still need improvement.

Virtual Reality Sports Strengthened by AI

Virtual Reality (VR) has added a different dimension to sports and gamification as now with virtual reality headsets, enthusiasts can compete virtually with each other from around the world. A virtual platform with AI technology provides a realistic experience in a virtual environment that matches the experience of witnessing the game live. Also, with the emergence of 5G, such experiences will get more interactive and the sports industry will be changed forever.

AI in Match Predictions

Machine learning can be used to predict the result of matches. Be it in soccer, or in cricket, where massive data is available, a model outcome can be created to predict the upcoming confrontations. One of the best practical applications of this can be conveyed through the project made by the students at Great Learning on ‘IPL Cricket Match Outcome Prediction using AI Techniques.’

Apart from these, there are some other ways in which the assistive-technology is being used in IPL:

Chatbots

Sports teams are using virtual assistants to respond to fan queries across a wide range of topics including game information, team information, team stats, etc.

Computer Vision

Researchers are training deep learning neural networks to predict beyond human abilities like optimized player’s performance, analyze and give instant inputs in things like batting technique, stance, and shot selection.

Wearable Tech

Companies are using wearable devices to track, measure & predict real-time impacts on a player. Not just while training, but also during the game, which can help the teams put their best player in front.

Data Mining for Selecting Players

Sports committees are using the regular feeds that keep coming from regional/local matches to try and identify the best player, who is currently in the best form.

Teams, players, promoters, advertisers, and viewers are all taking advantage of this advancement in technology through AI and are indeed becoming a game-changer in the field. AI is all set to completely revolutionize the way the sport is played.

[aianalytics]

Cybersecurity can be made agile with zero-shot AI

Artificial intelligence (AI) is being looked upon as the next frontier to make cyber-defence robust and scalable. The key aspects that make AI and particularly machine learning (ML) attractive in cybersecurity is the ability to learn from large volumes of telemetry data and find patterns of abnormal behaviour. ML algorithms can be used to find anomalies in different parts of the enterprise like application logs, network flows, user activities and authentication logs. As enterprises adopt models like zero-trust with strict identity verification and explicit permissions, augmenting these with ML algorithms to monitor user behaviour patterns becomes critical.

Modern security information and event management and intrusion detection systems leverage ML to correlate network features, identify patterns in data and highlight anomalies corresponding to attacks. Security researchers spend many hours understanding these attacks and trying to classify them into known kinds like port sweep, password guess, teardrop, etc. However, due to the constantly changing attack landscape and the emergence of advanced persistent threats (APTs), hackers are continuously finding new ways to attack systems.

A static list of classification of attacks will not be able to adapt to new and novel tactics adopted by adversaries. Also, due to the constant flow of alarms generated by multiple sources in the network, it becomes difficult to distinguish and prioritize particular types of attacks—the classic alarm flooding problem.

A possible solution would be if we had a smart system that could auto-label alarms and categorize them so that the analyst can focus on particular alarm types. We propose a dynamic classification system using a zero-shot classification approach to ML.

What exactly is this? The traditional approach to applying ML is supervised, where labelled data points are used to train models to make predictions. For example, a classifier model may process a record logged by a network monitor and classify it as an attack. While this is useful, these models can only learn from previously known attacks; so, a human would need to annotate the network flow for the attack data and feed it to build the model. The other approach becoming popular is unsupervised, where models learn to observe “normal" behaviour and flag any anomalies. This approach can highlight unknown attack patterns but only provide anomaly information to the security analyst.

One approach to tackle this is an upcoming research area in AI/ML called Explainable AI (XAI). Here, the models are either redesigned or enhanced to provide an explanation along with the prediction. So, when the model predicts an anomaly, it will also mention which feature values made it make that decision.

XAI and zero-shot learning can be applied to different areas of a cybersecurity ecosystem. Let’s take an example of an ML model that monitors network traffic in an office network. Say, it flags a data transmission above 100MB happening from a network computer to a Google drive account as an anomaly—different from normal network flows. If we show the security operation centre analyst additional parameters that made us flag this as anomaly, like size of data files and destination domain, this information can save the analyst valuable time in classifying this as a data exfiltration attack. The system can further take feedback from the analyst and start auto-labelling new such attacks as data exfiltration. Extrapolate this to a network with thousands of nodes and users, explainability and zero-shot learning can save hours of valuable time spent by analysts in searching for the needle in the haystack.

In today’s world, enterprises face APTs, which are well-funded attackers that focus on a high-value target and can stay undetected inside networks for days. Behavioural analytics becomes key to understand patterns and identify tactics, techniques and procedures used by attackers. Zero-shot learning models that decipher tactics like reconnaissance, privilege escalation and exfiltration can be extremely valuable to prevent major damage. Over time, this active learning system can learn from feedback.

Source

[aianalytics]

TOP USE OF ARTIFICIAL INTELLIGENCE AND MACHINE LEARNING IN FINANCIAL SCAMS

Machine learning refers to analytic techniques that “learn” patterns in datasets without being guided by a human analyst. Artificial intelligence refers to the broader application of specific kinds of analytics to accomplish tasks, from driving a car to, yes, identifying a fraudulent transaction. For our purposes, think of machine learning as a way to build analytic models and AI as the use of those models.

Machine learning helps data scientists efficiently determine which transactions are most likely to be fraudulent, while significantly reducing false positives. The techniques are extremely effective in fraud prevention and detection, as they allow for the automated discovery of patterns across large volumes of streaming transactions.

If done properly, machine learning can clearly distinguish legitimate and fraudulent behaviors while adapting over time to new, previously unseen fraud tactics. This can become quite complex as there is a need to interpret patterns in the data and apply data science to continually improve the ability to distinguish normal behavior from abnormal behavior. This requires thousands of computations to be accurately performed in milliseconds.

Preparation for the Worst-Case Scenario

Because organized crime schemes are so sophisticated and quick to adapt, defense strategies based on any single, one-size-fits-all analytic technique will produce sub-par results. Each use case should be supported by expertly crafted anomaly detection techniques that are optimal for the problem at hand. As a result, both supervised and unsupervised models play important roles in fraud detection and must be woven into comprehensive, next-generation fraud strategies.

Familiarity with Behavioral Profiles

Given the sophistication and speed of organized fraud rings, behavioral profiles must be updated with each transaction. This is a key component of helping financial institutions anticipate individual behaviors and execute fraud detection strategies, at scale, which distinguish both legitimate and illicit behavior changes.

Generic Behavior Models are not Enough

In order to maintain a positive consumer experience, specialized fraud analytics must be used to assess the “tough” questions. This is where advanced profiling, fraud-specific predictive characteristics, and adaptive capabilities separate themselves from generic behavior analytics. In a world of real-time payment processing and rapidly changing consumer preferences, generic behavior models are not sufficient for cross-channel, enterprise fraud solutions. After all, when and how someone chooses to transact is not as predictable as his or her likelihood to cancel a fitness club membership.

Source

[aianalytics]

India needs tech moonshots to power $10 tn ai-driven economy

On 25 May 1961, something special happened. President John F. Kennedy addressed a special session of Congress that the US planned to put people on the Moon and return them safely to Earth before the decade’s end.

Eight years after Kennedy’s initial so-called moonshot challenge, two American astronauts took the famous “one giant leap for mankind"—walking on the Moon for the first time. However, the bigger impact of that moonshot was the resultant building of a massive military and industrial innovation complex that propelled the US to the top of the industrial economy. The industrial economy, in simple terms, grew by selling excess production from one place to another by connected pathways.

By 1975, the industrial economy was at its peak, with more than 1 billion places connected by rail, road and airways dominated by multinational companies born in the US, selling technology developed during the moonshot.

The other spinoff of this moonshot was ARPANET, kicked off in 1966, which eventually became the internet. The internet gave rise to the knowledge economy driven by connecting people via the world wide web. One of the other offshoots of this moonshot was a technology named GPS, which is now powering at home pizza delivery to rental cabs via location-driven business models.

The world slowly witnessed the rise of the knowledge economy powered by massive information networks driven by connected human beings exploiting the knowledge deficits across nations and companies. India, too profited from this knowledge economy race by leveraging its talented workforce and grew companies such as Tata Consultancy Services Ltd, Infosys Ltd and Wipro Ltd.

The next 15 years are going to be critical for India’s growth story. We will have an opportunity to combine demographic dividend with artificial intelligence (AI)-driven hyper-innovation, in which 50 billion smart things (machines and devices) will combine with billions of connected humans. Unlike previous innovation cycles, the AI wave is different in which rapid technology innovation (combination of AI, Robotics, 5G and Quantum technologies) will occur together with business model innovation (digital intangibles-driven experience economy).

A case in point is Alexa, which promises to create the experience of owning an intelligent assistant that will provide a personalized digital experience in anything they do—shop, travel, eat, watch movies, etc. Global stock markets are seeing this shift very clearly. This is accelerating rapidly post-covid, with 90% of the perceived value being in digital intangibles—a case in point being the top 5 tech stocks amassing almost $8 trillion in market value.

For India to play a massive role in this upcoming ecosystem, we need to launch ambitious technology moonshots, which will create the innovation ecosystem necessary to dominate the experience economy. Else, China, with its almost-ready experience economy ecosystem, will grab half of the estimated $15.7 trillion of new economic value this AI-driven experience economy might provide.

Source

[aianalytics]

Want to turn low-res photo into high-res? New Google AI is crazy good at it

There are times when we wish a distorted, low-image was not the way it was because the same image would make for a great image had the quality been better and in high resolution. Google notes that super resolution models transform a low-resolution image into a detailed high resolution image. Super resolution can be used to restore old family portraits and to improve medical imaging systems. Google notes that the Diffusion models were originally proposed in 2015 but have seen a revival now due to their training stability and their promising sample quality results on image and audio generation.

Now, Google’s Research team has introduced two new approaches which use machine learning to enhance images. Google has introduced two models including SR3 -- Image Super-Resolution and CDM -- Class-Conditional ImageNet Generation which according to Google “push the boundaries of the image synthesis quality for diffusion models.”

Google notes that the SR3 is a super-resolution diffusion model that takes as input a low-resolution image, and builds a corresponding high-resolution image from pure noise. “The model is trained on an image corruption process in which noise is progressively added to a high-resolution image until only pure noise remains. It then learns to reverse this process, beginning from pure noise and progressively removing noise to reach a target distribution through the guidance of the input low-resolution image,” Google noted in a blog post.

Google noted that SR3 works efficiently when upscaling portraits and natural images. It showed a “confusion rate” of nearly 50 per cent while existing methods only go up to 34 per cent when used to 8x upscale faces.

Google saw a positive result in the SR3 model and introduced the CDM model which further enhances the picture’s resolution. The next model is CDM which Google notes is a “class-conditional diffusion model trained on ImageNet data to generate high-resolution natural images.” Google has posted a set of picture examples that show low-resolution photos upscaled in a cascade. A 32×32 photo can be enhanced to 64×64 and then 256×256. A 64×64 photo can be upscaled to 256×256 and then 1024×1024.

Google will also introduce a new data augmentation technique called conditioning augmentation, that will further improve the sample quality results of CDM. “Conditioning augmentation refers to applying data augmentation to the low-resolution input image of each super-resolution model in the cascading pipeline,” Google explained. These augmentations prevent each super-resolution model from overfitting to its lower resolution conditioning input, eventually leading to better higher resolution sample quality for CDM.

Source

[aianalytics]

Investments by Indian Conglomerates Boost Indigenous Production of Aerial Robots

According to BIS Research, global market intelligence and advisory firm, the global aerial robots market will touch a whopping $28.47 billion this year, of which over 4 per cent will be contributed by India. The country’s aerial robots market is estimated to become a billion-dollar industry in the next decade.

The strengthening of policies and the emergence of several academic institutions can be attributed to such positive anticipation. In addition to this, increasing investments by big-ticket Indian corporations is also one of the major reasons behind the country’s thriving aerial robots market. Speaking of which, Reliance Industries is one of those corporations which have been betting on aerial robots.

In 2019, RIL invested around Rs 23 crore in the Bengaluru-based robotics and artificial intelligence company Asteria Aerospace by acquiring a 52 per cent stake. Earlier, backed by investors like the Luxembourg - based Artificial intelligence and deeptech investment company, Boundary Holding, founded by Rajat Khare, the AI start-up has launched a slew of innovative solutions pertaining to aerial robots. Identifying its business potential, RIL went on to tap into it.

However, RIL is nowhere near stopping from eyeing disruptive start-ups. For the past five years, it has invested in more than 100 companies and will continue doing so in 2021 as well.

Multinational conglomerate Larsen & Toubro has also been vying for investments in aerial robots. The Defence Research and Development Organisation (DRDO) is said to have roped in L&T and Tata, among others, to scale up the production of its technology developed for countering rogue aerial robots. Furthermore, last year, L&T entered into an agreement with ideaForge to offer aerial robots and allied systems.

Today, while several innovative start-ups are emerging, a lack of sufficient funds to scale their operations is one of the major challenges. Result? India is compelled to rely on other countries for the latest technologies. Thus, multinational conglomerates pouring in investments in those firms will definitely boost indigenous production of such new-age technology.

[aianalytics]

IIT Delhi launches AI Lab for judiciary

Recently, IIT Delhi launched the Universal Justice Foundation (UJF) lab facility on artificial intelligence for the judiciary, which was inaugurated by Justice S Ravindra Bhat.

This is the second time that the institute has undertaken an initiative for the legal domain, with the first being the establishment of a Centre of Excellence for Law and Technology.

AI-based tools can also streamline and accelerate the management of case-flows and bridge the existing gaps in the judiciary. The use of AI can serve as a support system for legal personnel, boosting their decision-making processes and operations.

Technology has already made its way into courts and has significantly improved data management. It has also replaced the otherwise time-consuming paper process. The new lab facility at IIT Delhi will make more such innovations.

Rajat Khare, an IIT Delhi alumnus, said, “The legal sector is undergoing a massive transformation, more so amid the pandemic. Thus, embracing new-age technologies like AI and machine learning has become the need of the hour.”

[aianalytics]

ARTIFICIAL INTELLIGENCE: THE NEXT GENERATION ANTI-CORRUPTION TECHNOLOGY

Artificial Intelligence (AI) can be a useful weapon in the fight against corruption. Its capacity to handle huge data is unrivaled, as is its ability to spot abnormalities or trends, such as in financial transaction data. Some of the ways AI is used in society have skeptics, who fear a society that is more monitored, putting privacy and individual freedom in danger. Let’s get into the topic in more detail.

AI as an Anti-corruption Tool

Artificial intelligence (AI) refers to technologies that allow machines to simulate human intelligence in order to tackle complicated issues. On the one hand, there are techniques in which an algorithm, or a “recipe” for dealing with a given set of inputs, directs the computer process that decides or recommends a result. Machine learning (ML) is a subdomain of this area, in which many approaches of varying degrees of complexity are used to tackle diverse issues. Some of these methods require a dataset in order to ‘train’ the algorithm on how to deal with the data. The datasets used to ride the algorithm are often the source of algorithmic bias. Without any supervision, certain systems ‘learn’ how to produce the best possible result. Artificial neural networks are built in the same way as our brain is. Millions of computations are done and communicated between the network’s nodes, resulting in a level of complexity that is difficult to comprehend. The term “black box problem” describes calculations in sophisticated algorithms that are not transparent. Artificial general intelligence (AGI) or superintelligence, which are more sophisticated imitations of human intelligence, remain in the future and are not the subject of this text. We’re also not going to talk about robotics.

High Hopes for The Future

Development organizations express optimism about the benefits of new technologies, as well as some skepticism about the drawbacks.

Some designs include new, digitized procedures that eliminate previously corruptible jobs. Other initiatives use a more “direct” approach to uncovering previously concealed transactions or perpetrators of fraud.

In many situations, the basis on which AI applications are built is digitized interactions between society and its inhabitants. Reconfiguring business or governance processes to allow for automation and AI help may, in some circumstances, minimize the risk of fraudulent activity.

Using AI to Uncover Corruption and Fraud

Artificial intelligence, according to Oxford Insights, is the “next step in anti-corruption,” partially because of its capacity to uncover patterns in datasets that are too vast for people to handle. Humans may focus on specifics and follow up on suspected abuse, fraud, or corruption by using AI to discover components of interest. Mexico is an example of a country where artificial intelligence alone may not be enough to win the war.

The telecommunications industry is one of several segments of the Mexican economy that has witnessed improvement. Telecom was once dominated by a single company, but it is now open to competition. As a result, the cost of connectivity has decreased significantly, and the government is currently preparing for its largest investment ever. By 2024, the objective is to have a 4G mobile connection available to more than 90% of the population. In a society moving toward digital state services, the affordable connection is critical.

The next stage is for the country to establish an AI strategy. The next national AI strategy will include initiatives such as striving toward AI-based solutions to offer government services for less money or introducing AI-driven smart procurement. In brief, Mexico aspires to be one of the world’s first 10 countries to adopt a national AI policy.

Digital Reports on Development Aid

Corruption and fraud in donor holdings are one problem where new technology might help speed up investigations or make suspicious occurrences easier to identify. The International Aid Transparency Initiative (IATIOpenAid)’s idea has been around for a long and has been implemented by a number of countries. Transactions and reporting must be synchronized for AI technologies to be effective. Projects spanning several nations, including different languages, currencies, or reporting methods may require some ‘cleaning’ before an AI program can monitor effectively enough to detect potential anomalies with a satisfactory degree of precision. The following sample comes from a fully digitized donor organization with a well-established structure. Even yet, the reports must be reviewed by humans before being disseminated. To help with this, a machine learning program was built.

Taking up The Problem of Getting Solid Data

For the AI revolution to materialize, digitization is required. Improving the volume and quality of the data from diverse areas of society is one cross-national goal for IBM and the corporation’s more than 24 regional offices in Africa. A major problem is the absence of trustworthy and consistent data, such as from off-grid economies. To assist this digitization initiative, IBM is using resources from its regular business operations.

Conclusion

There are reasons to be concerned about biased outcomes if and when AI is used in governance and decision-making to assist or replace existing services. Adverse side effects of these decision-making systems might be caused by bias in the data for training the AI or in the algorithm’s architecture. The black box problem refers to opaque algorithms and, as a result, opaque decision-making systems. The ability to explain necessitates the development of transparent algorithms or techniques capable of testing or contesting judgments. Several organizations, including the European Union, have created ethical standards for the design, implementation, and promotion of AI trust, emphasizing that a trustworthy AI should be legal, ethical, and resilient. When technology advances faster than regulation, challenges arise because it may function in uncontrolled, global settings.

[aianalytics]

How to use the Google Cardboard virtual reality app with compatible viewers

Google Cardboard is the company's affordable answer to a virtual reality (VR) experience: It's a free app that works alongside other VR apps to let users effectively turn their phones into a VR device. They also offer relatively cheap viewer options on the Google Cardboard site, lowering the bar to entry for VR.

Here's how to find Google Cardboard-compatible VR viewers and use the Cardboard app on your iPhone or Android.

What is Google Cardboard?

The term "Google Cardboard" can be used to refer to two different things:

- The Google Cardboard app: The app, which can be downloaded for iPhone or Android, helps users launch VR experiences on their device.

- The Google Cardboard VR viewer: This cardboard viewer was made to fit around your smartphone and function like VR goggles. Google has discontinued the device, though they do provide instructions for creating your own cardboard viewer.

Note: Even though the Google-made viewer is no longer available, you can still find similar products from other VR companies through the Google Cardboard site. And those can work alongside the Google Cardboard app.

How Google Cardboard works

To use the Google Cardboard app, you just have to do the following:

1. Download the Google Cardboard app from your phone's app store.

2. Give the app permission to use your camera.

3. Find the QR code on your viewer and scan it.

4. Place your phone in your VR viewer device.

5. Load up your desired VR experience, either within the Cardboard app or through a Cardboard-compatible VR app.

How to get and use Google Cardboard

The VR viewers showcased on the Google Cardboard site range from about $8.95 to $39.95, and are made from cardboard, plastic, and nylon ABS. They can fit screen sizes up to seven inches, depending on the model you purchase.

Quick tip: Once you have a viewer, you can scan the QR code into the Cardboard app to connect it up, so the shape of each side of your screen will fit the viewer you're using.

You can also download Cardboard-compatible VR apps, which can be found within the Cardboard app.

To use an app in tandem with Cardboard and your viewer, select the View in Cardboard option.

From there, you should be able to enjoy your VR experience, properly formatted for your viewer.

Source

[aianalytics]

A PEEK AT TOP ARTIFICIAL INTELLIGENCE FUNDING IN JULY AND AUG 2021

The amount of money invested annually into startup companies working in the artificial intelligence (AI) market worldwide has continuously increased. The companies receive funding by showcasing their expertise and their clients’ reviews.

Olive

July 26: Olive deploys the Artificial Intelligence workforce built specifically for healthcare, delivering hospitals and health systems increased revenue, reduced costs, and increased capacity. The objective is to automate all those repetitive, high-volume tasks and workflows, and also to monitor their performance, identify their improvements, and find opportunities for new work as well. As far as the AI funding of this firm is concerned, it is funded by 18 investors. Olive, in over 9 rounds has raised a total funding of $856.3M. Their recent funding was raised on July 1, 2021, from a Series H round. This recent funding amounted to $400 million.

Covariant

July 27: Covariant, a leading AI Robotics company, today announced it has raised $80 million in Series C funding, bringing its total capitalization to $147 million within two years of the company’s public launch. The round was led by returning investors, Index Ventures, with the additional participation of Amplify Partners and Radical Ventures.

Blaize Inc

July 26: Artificial intelligence computing company Blaize Inc has raised $71 million from investors including Temasek and Franklin Templeton, according to people familiar with the matter. The El Dorado Hills, California-based company is also in early-stage talks with special purpose acquisition companies (SPACs) about a potential deal that would make it public, the sources added. Blaize previously raised $65 million in 2018, at a valuation of about $370 million from investors including Denso, Temasek, GGV Capital, and Daimler. The new series D round of funding will be primarily used to scale out the business and products development. Details on Blaize’s latest valuation were not immediately available.

Nektar.ai

August 2: Nektar.ai, a business-to-business (B2B) sales productivity startup, has raised $6 million as part of the second tranche of its seed round, from investors led by B Capital Group, 3One4 Capital, and Nexus Venture Partners. This takes its total funding in the seed round to $8.1 million, making it one of the biggest such rounds for a Software as a Service (SaaS) startup in the region.

Neuron7.ai

August 10: Amid the broad proliferation of devices and data in our homes and businesses, Neuron7.ai, a new cloud-software company focused on the new category of service intelligence, has emerged from stealth mode and announced a seed investment of $4.2 million from Nexus Venture Partners and Battery Ventures. The company, led by repeat entrepreneurs Niken Patel and Vinay Saini, is helping drive the transformation of customer service into a cloud-based AI-powered workflow, particularly for companies managing complex products in technology, manufacturing, and healthcare, where service organizations are required to support hundreds of product models, versions, errors, and issues.

Source

[aianalytics]

OVERVIEW OF TRANSFORMATION IN RADIOLOGY THROUGH ARTIFICIAL INTELLIGENCE

The scope of Artificial Intelligence is not only limited to certain industries and household chores but in all directions. The global market size of AI is increasing at an increasing rate due to digital transformation and globalization. The world has experienced how AI is transforming the healthcare industry, especially in the COVID-19 pandemic. It is known that radiology includes diagnosing as well as treating diseases or injuries through X-rays, Computed Tomography (CT), Magnetic Resonance Imaging (MRI), nuclear medicine, ultrasound, and many more. Radiologists get continuous exposure to radioactive rays that are very harmful to health. Analytics Insight provides an overview of the transformation in radiology through the implementation of Artificial Intelligence in recent years.

Radiologists are concerned whether Artificial Intelligence machines will take over their job opportunities in the nearby future. But it is a myth because the implementation of Artificial Intelligence will optimize the workflow through quantitative radiology by protecting the health of fellow radiologists. Artificial Intelligence is helping radiologists in different ways— detecting early-stage of cancer, auto-segmentation of organs in 3D models, NLP for reports, and many more. AI has raised a significant value to radiologists through accurate pieces of advice or insights regarding patients and their health issues.

Artificial Intelligence and machine learning algorithms are transforming the regular work of radiologists through appropriate communication, coordination, screening examinations like mammography, colonography, and chest CT, standardize reports as well as immediate alerts for critical patients. It is used for inventory and equipment management and maintenance efficiently and effectively. The AI models have started acting as assistants to senior radiologists in analyzing the medical records and real-time data of patients and detect any deterioration or improvement in a particular disease or injury. The Artificial Intelligence algorithms are here to guide radiologists in redefining their purpose and perceive critical data that are impossible for them to notice through naked eyes. Radiologists need to work long shifts daily which makes them tired and distracted. Artificial Intelligence is particularly helpful in this situation to make them notice certain issues in a patient’s body as well as complete their mundane tasks.

Yes, when radiologists across the world are set to utilize smart functionalities of Artificial Intelligence for the utmost care of patients, there are certain challenges faced by them at the same time. One of the major drawbacks of algorithms is these machines do not have sufficient medical knowledge as per a radiologist or physician. Sometimes it is difficult for AI machines to understand the proper workflow in the radiology department. Secondly, these Artificial Intelligence machines are trained with historical medical data of different kinds. But, radiologists find unique and new types of symptoms of a disease that are not registered in the historical data. Thus, it is difficult to solve diseases and predict the curing process on the basis of one modality. Thirdly, AI machines work differently at different hospitals ie. the quality of outcomes may decline or improve, depending on the medical records and systems of hospitals. Hence, radiologists in different hospitals and clinics need to update the training dataset with their existing medical records respectively.

Artificial Intelligence has created a massive impact on radiomics which is a new field in radiology in recent years. Radiomics deals with the extraction of a high number of different features such as size, shape, and texture from medical images of patients. These features include spatial information on pixel or voxel distribution and patterns as well as provide support for the diagnosis of brain, heart, liver, prostate, adrenal gland, pituitary gland, and lung. The blend of Artificial Intelligence with radionics provides a smarter capability of managing enormous datasets efficiently than the traditional systems. It gives computational power in radiology and encourages radiologists to adopt radiomics and AI for better practice.

That being said, it is essential to incorporate Artificial Intelligence in radiology to transform the workflow efficiently, despite having a few barriers. It is useful for gaining trust and loyalty from existing and potential patients for a better cure as soon as possible.

Source

[aianalytics]

A BRIEF UNDERSTANDING OF MACHINE LEARNING

Machine Learning is a sub-branch of Artificial Intelligence (AI) that is one of the fast-evolving fields of computer science. It is simply the study of computer algorithms and related data, which automatically improve through experience and learns to imitate the way a human learns and acts, with high accuracy. The great mind of this field is Arthur Samuel, the researcher who coined the term and is one of the pioneers of AI.

As seen in the above image, the study includes researching computer data and algorithms, using data mining processes and tools, uncovering key insights that shall help in the decision-making process and its implementation in business, and in turn impacting the key growth metrics to get in-hand automated results.

Approaches to Machine Learning

The different approaches to Machine Learning are divided into five categories-

Supervised Learning: It is defined by its use of labeled datasets to train algorithms that classify data or predict outcomes very accurately. It includes active learning, classification and regression.

Unsupervised Learning: It uses machine learning algorithms to analyze and cluster unlabeled datasets. The algorithms, therefore, learn from the test data that has not been labeled or classified.

Semi-supervised Learning: As the name suggests, semi-supervised learning falls in between supervised and unsupervised learning. This type of learning uses the smaller labeled figures to guide classification and feature extraction from larger, unlabeled figures.

Reinforcement Learning: This is a behavioral machine learning model which is similar to supervised learning. This model learns by using trial and error methods. It is used in various disciplines likegame theory, control theory, operations research, information theory, simulation-based optimization, multi-agent systems, swarm intelligence, statistics and genetic algorithms, etc.

Dimensionality Reduction: Dimensionality reduction techniques can be considered as feature elimination or extraction. It is simply the reduction of the number of random variables under the consideration by obtaining a set of principal variables.

Real-life Applications of ML

Machine Learning in your everyday life

Automated Speech Recognition: This computerized speech recognition uses natural language processing (NLP) to process written text formats and convert them into human speech. Many mobile and laptop devices have this in-built feature that makes texting very much easier.

Customer Service: Online chatbots and pre-recorded customer service calls use machine learning to automatically answer FAQs, suggestions, personalized advices, customer engagements, etc.

Search Engine Learning and Recommendation Systems: Websites, applications and search engines constantly use ML to improvise the recommendations and personalization problems. For example- Google, Netflix, Uber, Amazon, etc.

Education: Gamified Learning is a very efficient way of providing education. The algorithm is programmed to display only the correct answer of the user at the very end and the questions that are incorrectly answered will repeat again so that the user shall thoroughly remember the correct answers.

Not only the above mentioned, but ML has many other uses like predicting illness in healthcare cases, checking the credit-worthiness of a bank’s clients, self-driving cars, ranking of social media posts, computer vision in agriculture, targeted e-mails

Source

[aianalytics]

Facebook testing virtual reality ads in Oculus VR

In May, Facebook announced that it would begin testing advertisements in virtual reality. Those tests are now about to go live. “The company revealed it’s going to begin experimenting with the ads in the Oculus Quest title Blaston from Resolution Games. The experiment will also expand to two other unnamed developers in the coming weeks,” said Michael Tan for PCMag.

In-headset ads. The advertisements, deemed “in-headset ads” by Facebook, are part of the company’s exploration of ways for developers to generate revenue: “This is a key part of ensuring we’re creating a self-sustaining platform that can support a variety of business models that unlock new types of content and audiences,” the company said in the announcement blog.

Privacy. Facebook, which develops the Oculus VR headsets, plans to monitor user interaction with the VR ads, but did say that all Oculus ads will still have to follow Facebook’s advertising rules. As such, users will still be able to use “controls to hide specific ads or hide ads from an advertiser completely.”

In the announcement blog, Facebook outlined the privacy policies for Oculus ads:

1. We do not use information processed and stored locally on your headset to target ads. Processing and storing information on the device means it doesn’t leave your headset or reach Facebook servers, so it can’t be used for advertising. 2. We take extra precautions around the use of movement data like minimizing what we need to deliver a safe and immersive VR experience and we have no plans to use movement data to target ads. 3. We do not use the content of your conversations with people on apps like Messenger, Parties, and chats or your voice interactions to target ads. “Still, because the company is trying to require Oculus VR owners to sign in with a Facebook account, the social network can still analyze your personal data to serve up targeted ads,” points out Tan.

Why we care. Privacy is a big issue for users right now and advertisements in virtual reality might be a step too far for many. Facebook has been caught in the middle of this privacy debate, especially as iOS 14 has cracked down on app tracking. However, the opportunity may be perfect for many advertisers to reach a new or specific type of audience if the tests go well. There’s a precipice where ads in VR almost feel like we’re headed toward Ready Player One territory, so it’s a trend worth watching.

Source