Society is experiencing the benefits of AI in this COVID-19 era as hospitals detect infections and treat conditions. The recent AI investments by IBM Watson AI Lab and MIT will accelerate the use of AI for societal benefits.
AI applications in health care will continue with hospitals using artificial intelligence to detect infections and offer treatments. The search for the COVID-19 vaccine² is heating up and AI is at the center of this search.
Blue Dot¹¹ from Canada works directly with the World Health Organization because of its sophisticated tools that detected the first infections in Wuhan, China.
The Industrial Revolution can teach us about the adoption of transformative technologies including artificial intelligence.
Many people still do not understand AI despite its enormous potential¹ for supporting socio-economic development. The investment and harvesting stages of new technology take time and the same applies to AI that is still at the early phase.
Are we using AI as we should?
AI technologies enabled the pandemic to be tracked and information of it to be synthesized. It also assists in diagnosing patients, triaging them, and identifying those in need of intensive care before their condition deteriorates.
Artificial intelligence is searching for drugs to combat Covid19 as well. There are hypes that exist about the use of artificial intelligence (AI) to combat coronavirus and some of it is justified.
The ability of AI to sift through vast amounts of data and recognizing patterns has been of great value. Some of the hype is flannel. As some claim, AI predicted the pandemic³ before humans recognized it.
The use of AI for mass surveillance causes concern. The pandemic has exposed the limitations of AI in some aspects. The algorithms used by online retailers, trained on normal human behavior found that people’s behavior has completely changed when it comes to shopping or travel, and they are often stumped.
The ability to recognize patterns enable AI to triage patients, but the machine has no ethical commitment to patients in the way doctors and nurses do. The best public policy rests on facts but it also requires us to choose between competing demands.
Machines cannot think like humans. Humans should not act like machines.
AI Improving Voice of the Customer Programs
AI is revolutionizing customer experience. Here are the ways how AI can improve the voice of the customer programs
✅Companies are being able to know what causes some customers to churn faster than others using an AI-based analysis of customer journey data helps define new campaigns to keep them.
✅ Companies are using algorithms to perform real-time text mining of every source of textual, unstructured data available to analyze the sentiment levels of customers.
✅The scope of speech analytics⁴ is expanding using AI to include contact center conversations, text-based customer feedback, and operational #data from every customer touch point.
✅@Amazon Connect and other Cloud-based Speech Platforms are relying on AI to remove the roadblocks that get in the way of launching and fine-tuning VoC programs across multiple geographies and languages.
✅AI helps to gain insights allowing call centers to be transformed from being first-line service providers to becoming strategic differentiators that drive significant improvements in customer satisfaction and financial performance.
✅A real-time multidimensional view of the caller and agent-based attitudinal performance is possible using machine learning algorithms.
AI-powered Messenger feature from Facebook
Facebook unveiled an AI-powered Messenger feature that surfaces tips to help younger users spot malicious actors. The Messenger guidelines outline steps for blocking people if necessary and intended to educate users under the age of 18 about interacting with unknown adults.
Users can follow the rollout of limits on chats and a hub that spotlights pandemic resources — both attempts to limit the spread of misinformation. According to a report from global nonprofit Avaaz, misleading content on
Facebook is still shared and viewed hundreds of millions of times despite Facebook’s renewed campaign against false coronavirus information⁵. The U.S. Federal Trade Commission has documented over 20,000 instances of messages offering bogus testing kits, unproven treatments, or predatory loans.
Jay Sullivan, Messenger privacy and safety director, said, the AI powering the safety tips feature rolled out on Android in March and will expand to iOS next week — looks at behavioral signals like an adult sending a large number of or message or friend requests to under 18 users.
The invention allows the AI to work with end-to-end encryption schemes, ensuring it will continue to function after Messenger becomes encrypted by default.
History tells us about the accelerating AI Revolution
Most ordinary people don’t feel particularly optimistic about the future of AI despite the recent advances that sparked much excitement. Three-quarters of Americans expressed serious concerns about AI and automation according to the 2017 Pew Research survey. Serious concerns about the impact of technology is part of a historical pattern.
The Covid19 pandemic likely accelerates the rate and pace of technological change. What can we learn from the Industrial Revolution that can help us better face our emerging AI revolution⁶?
The life cycle of historically transformative technologies consists of two phases: investments and harvesting. It takes longer to reach the harvesting phase for more transformative technologies.
We’re still in the early stages of AI’s deployment. The extensive investments required to embrace a GPT like AI will generally reduce productivity growth.
The transitions of AI will be very challenging while a growing technology-based economy will create a significant number of new occupations according to a 2017 McKinsey study.
Game Design in the Future
Though Nvidia¹² is worldwide recognized for its graphics cards, but the company conducts some serious research into artificial intelligence too. Nvidia researchers taught an AI system to recreate the game of Pac-Man simply by watching it being played for its latest project.
For this software no coding and no pre-rendered images involved. The AI model is simply fed visual data of the game in action along with the accompanying controller inputs and then recreates it frame by frame from this information. Nvidia says the game will be released soon and playable by humans.
By no means, the AI version is a perfect facsimile, and the imagery is blurry and it doesn’t seem like the AI managed to capture the exact behavior of the game’s ghosts, each of which is programmed with a specific personality that dictates its movement. But the basic dynamics of PacMan are all there.
This work shows how artificial intelligence will be used for game design in the future according to Nvidia. Developers can use AI to create variations or maybe design new levels. If an AI that can learn⁷ the rules of a virtual world just by watching it in action also has implications for tasks like programming robots then the AI might watch videos of robotics trolleys navigating a warehouse.
AI’s Transformative Potential for Society
The MIT-IBM Watson AI Lab is funding 10 projects at MIT aimed at advancing AI’s transformative potential for society. The 10 research projects are highlighted below.
✅Early detection of sepsis in Covid19 patients:
Researchers will develop a machine-learning system⁸ to analyze images of patients’ white blood cells for signs of an activated immune response against sepsis.
✅Designing proteins to block SARS-CoV-2:
Researchers will enlist the protein-folding method used in their honeybee-silk discovery to try to defeat the new coronavirus.
✅Saving lives while restarting the U.S. economy:
Researchers will consider how antigen tests and contact tracing apps can further reduce publichealth risks.
✅Which materials make the best face masks?
The researchers will test materials worn alone & together, and in a variety of configurations and environmental conditions.
✅Treating Covid-19 with repurposed drugs:
✅A privacy-first approach to automated contact tracing:
✅Overcoming manufacturing and supply hurdles to provide global access to a corona virus vaccine:
✅Leveraging electronic medical records to find a treatment for Covid-19:
✅Finding better ways to treat Covid-19 patients on ventilators:
✅Returning to normal via targeted lockdowns, personalized treatments & mass testing:
Microsoft’s Build20200′ Developers’ Conference
Microsoft recently hosted the biggest event of the year Build 2020′ its developers’ conference and had plenty of big news for businesses, developers, and business developers. Cloud and AI announcements abounded for good reason in the event but the highlight of the event was where these all overlapped: a supercomputer in the cloud.
Microsoft’s $1 billion investment in OpenAI is bearing fruit to jointly develop new technologies for Azure. OpenAI’s supercomputer collaboration with Microsoft marks its biggest bet yet on AGI. Microsoft explained why you should care about self-supervised learning⁹ and a supercomputer in Azure.
Machinelearning experts have focused largely on relatively small AI models to learn a single task. It has shown lately by AI research community that applying self-supervised learning to build a single massive AI model can perform some of those tasks much better. If trained on code — writing code, larger AI models can learn the language, grammar, knowledge, concepts, and context to the point.
Microsoft is building AI supercomputer as part of Azure to throw more compute at AI. A supercomputer in Azure makes perfect sense. Microsoft is going to make those very large AI models and the infrastructure needed to train them broadly available via Azure.
Explainability Issues of Machine-Learning Algorithms
There is one fundamental truth that the machine-learning model is only as good as its data regardless of the algorithm¹⁰ used to create the machine-learning model.
There are many reasons for which a model performs poorly such as the input data could be riddled with errors or poorly cleansed or the various settings for the model could be set improperly yielding substandard results. Moreover, the data scientists and ML engineers that trained the model selected a subset of available data that includes inherent bias resulting in a skewed model.
Some machine-learning algorithms are unexplainable which has been acknowledged. It is known that when the model has come to some conclusion such as a classification or a regression, there is little visibility into the understanding of how the model came to that conclusion. The current celebutante of machine learning, deep learning neural networks particularly suffers from this problem.
All machine-learning algorithms don’t suffer from the same explainability issues. Decision trees are explainable by nature, although we lose elements of that explainability. when used in ensemble methods such as Random Forests.
The question that needs to ask is about the training data is where is that data from? How was it cleaned? Can I see the training data?
Successful AI Implementations
Do you know? 85% of enterprises are evaluating or using artificial intelligence in production today. Today 55% of all enterprises adopting AI using TensorFlow as their primary development tool. Evaluating AI in production today TensorFlow continues to be the most popular development tool across all enterprises.
Supervised learning is the most popular machine learning technique according to 73% of enterprises with the most advanced AI adoption. Enterprises prefer Human-in-the-loop AI models with advanced AI expertise compared to their peers.
According to a recent MIT Sloan Management Review study, Enterprise’s enthusiasm for AI is growing, with 62% increasing their spending last year. AI is being adopted evenly across enterprises, with R&D leading all departments by a wide margin in 2020.
AI has the potential for redefining enterprises, making them more customer-driven, adaptive, and capable of generating and sharing intelligence faster than ever before. The most advanced enterprises are more likely to include steps during model building to improve fairness, ethics, and limit or control biases while adopting AI.
Senior management provides strong support for the most successful AI implementations. Does your senior management support AI implementations?
Cutting-Edge AI for Enterprises
The most exciting event for the AI and ML ecosystem is The GPU Technology Conference.This conference has something relevant for researchers in academia to product managers at hyper-scale cloud companies to IoT builders and makers.
The keynote delivered by Jensen Huang, the CEO of @NVIDIA made the virtual event an unforgettable one. The keynote of Jensen was set in the kitchen of his California home which became the center stage for the game-changing technologies debuted at GTC. Jensen was at his best when he unveiled the next-generation GPU technology from NVIDIA — Ampere.
Ampere the GPU architecture is the successor to Volta, which powers a variety of AI accelerators in the data center and the cloud. Ampere becomes the cornerstone of NVIDIA’s AI strategy and product portfolio, and the unified and consistent GPU technology platform from the edge to the cloud.
Ampere features the third generation of Tensor Core — a chip that’s purpose-built for acceleratingAI. NVIDIAVolta architecture introduced Tensor Core in 2017 which dramatically changed the time it took to train complex deep learning models.
A100 — The first AI accelerator to be powered by Ampere. DGXA100 — The last thing an enterprise needs for cutting-edge AI.
AI showing a glimpse of how it will help Healthcare
The artificial intelligence platform BlueDot picked up an anomaly on New Year’s Eve of last year. It registered a cluster of unusual pneumonia cases in Wuhan, China. The Toronto, Canada based BlueDot uses natural language processing and machine-learning to track, locate, and report on infectious disease spread.
Health care, government, business, and public health bodies get alerts from it. BlueDot spotted Covid19 nine days before the @World Health Organization released its statement alerting people to the emergence of a novel coronavirus.
The role in spotting the outbreak was an early example of AI intervention. Artificial intelligence has already played a useful role in many aspects to combat coronavirus.
AI has been used for prediction, screening, contact alerts, faster diagnosis, automated deliveries, and laboratory drug discovery in the past months.
Innovative applications of AI have cropped up in many different locations as the pandemic has rolled around the planet. Location-based messaging has been a crucial tool in the battle to reduce the transmission of the disease in South Korea. Alibaba announced in China that an AI algorithm can diagnose suspected cases within 20 seconds with 96 percent accuracy.
AI for Content Moderation is troubling
Tech companies have scrambled to ensure their services are still available to their users, while also transitioning thousands of their employees to teleworking in response to the spread of the novel coronavirus around the world. Social media companies have been unable to transition all of their content moderators to remote work due to privacy and security concerns.
So social media companies have become more reliant on artificial intelligence to make content moderation decisions. Over the last couple of months Facebook and YouTube admitted much in their public announcements, and Twitter appears to be taking a similar tack. The sustained reliance on AI due pandemic is concerning as it has significant and ongoing consequences for the free expression rights of online users.
The comprehensive use of AI for content moderation is troubling because these automated tools have been found to be inaccurate. This is happened due to the lack of diversity in the training samples that algorithmic models are trained on.
Researchers have identified instances in which automated content moderation tools on platforms such as YouTube mistakenly categorized videos posted by NGOs documenting human rights abuses by ISIS in Syria as extremist content and removed them.
¹AI Potential, ²COVID-19 Vaccine, ³Pandemic, ⁴Speech Analytics, ⁵False Coronavirus Information, ⁶AI Revolution, ⁷AI that can Learn, ⁸Machine-Learning System, ⁹Self-Supervised Learning, ¹⁰Algorithm