Ce diaporama a bien été signalé.
Le téléchargement de votre SlideShare est en cours. ×

Ai in Learning Thought Paper

Publicité
Publicité
Publicité
Publicité
Publicité
Publicité
Publicité
Publicité
Publicité
Publicité
Publicité
Publicité
Chargement dans…3
×

Consultez-les par la suite

1 sur 30 Publicité

Ai in Learning Thought Paper

Télécharger pour lire hors ligne

This Thought Paper discusses how Artificial Intelligence could be used in the learning and development space. It provides various inspiration based on examples of solutions from and beyond the industry. The paper was prepared by Shweta Panwar and Debamitra Dasgupta within the Think Tank Center of Excellence of the Talent Development and Learning Practice in Accenture Capability Network.

This Thought Paper discusses how Artificial Intelligence could be used in the learning and development space. It provides various inspiration based on examples of solutions from and beyond the industry. The paper was prepared by Shweta Panwar and Debamitra Dasgupta within the Think Tank Center of Excellence of the Talent Development and Learning Practice in Accenture Capability Network.

Publicité
Publicité

Plus De Contenu Connexe

Diaporamas pour vous (20)

Similaire à Ai in Learning Thought Paper (20)

Publicité

Plus récents (20)

Publicité

Ai in Learning Thought Paper

  1. 1. Center of Excellence Thought Paper Artificial Intelligence in Learning Accenture Capability Nework
  2. 2. Copyright © 2018 Accenture All rights reserved. 2 Unmanned Service Smart Parking Google Waymo Ecommerce: Amazon Grid Healthcare: Watson Home Security Banking and Insurance Unmanned Delivery Google Now Alexa Netflix DominoesCortana Siri P A DAY IN OUR LIVES… Copyright © 2018 Accenture All rights reserved.
  3. 3. Copyright © 2018 Accenture All rights reserved. 3 Copyright © 2018 Accenture All rights reserved. A DAY IN OUR LIVES…WITH AI P Service Robots Robotic / Smart Parking Self Driving Cars: Google Waymo Predictive Ecommerce: Amazon Smart Grid Predictive Healthcare: Watson Home Automation and Security Banking and Insurance Drone Delivery Google Now Alexa Netflix DominoesCortana Siri
  4. 4. Copyright © 2018 Accenture All rights reserved. 4 ARTIFICIAL INTELLIGENCE IS A REALITY You may not realize it, but Artificial Intelligence IS ALL AROUND US
  5. 5. Copyright © 2018 Accenture All rights reserved. 5 Artificial Intelligence is the science of making a computer, a computer-controlled robot, or a software thinking intelligently, in the similar manner the intelligent humans think. WHAT IS ARTIFICIAL INTELLIGENCE? ARTIFICIAL INTELLIGENCE ARTIFICIAL INTELLIGENCE COGNITIVE SKILLS Learning and understanding Applying reason and logic Adapting to situations Problem-solving Decision-making Communication Artificial Intelligence is the capability of a machine to imitate intelligent human COGNITION.
  6. 6. Copyright © 2018 Accenture All rights reserved. 6 The AI market has grown tremendously in last few years. Experts forecast that annual global revenue from AI products and services will grow from $643.7 million in 2016 to $36.5 - $100 billion by 2025. Here’s a look into what lies ahead for AI: AI MARKET GROWTH By 2018, 62% fast growing companies are expected to use AI, while only 38% of the companies used machines in 2017. 62% By 2020, Gartner claims that 85% of all customer interactions would be mechanized minimizing the need for customer service agents. 85% By 2025, the AI software market is estimated to be worth a whopping $89.8 billion globally. $90 bn By 2018, more than 3 million workers globally will be supervised by a “robo-boss.” 3 mn
  7. 7. Copyright © 2018 Accenture All rights reserved. 7 ARTIFICIAL INTELLIGENCE IS A COLLECTION OF MULTIPLE TECHNOLOGIES Natural Language Processing Virtual Agents Robotic Process Automation Machine Learning Deep Learning Predictive Analysis
  8. 8. Copyright © 2018 Accenture All rights reserved. 8 MACHINE LEARNING Machine learning is the ability of computers or algorithms to interpret ambiguous data, such as text, speech, visual, and digital inputs, and analyze them for meaningful insights. Machine learning can help in predicting and making recommendations, detecting anomalies, performing classification, ranking, and decision making. Netflix is using machine learning for its video recommendation engine to personalize as much of Netflix as possible and match consumers to their content. It is estimated that the recommendation system has significantly increased customer retention and resulted in savings of about $1 billion a year for the company.
  9. 9. Copyright © 2018 Accenture All rights reserved. 9 DEEP LEARNING Deep Learning allows a machine to process large datasets and recognize patterns for decision making using methods that are similar to human brain’s neural networks. Deep learning is used by organizations to analyze and predict online behavior, organize information, predict outcomes, and boost the accuracy of other AI technologies, such as image analysis, face recognition, handwriting recognition, and speech recognition. PayPal is currently using deep learning to combat fraud and manage risk. The company gathers huge quantities of data about its customers, including financial data, network information, and machine information and then feeds this data to deep learning algorithms to spot patterns, fix glitches, and detect frauds.
  10. 10. Copyright © 2018 Accenture All rights reserved. 10 VIRTUAL AGENTS Virtual agents are conversational, computer-generated characters or programs that replace the online customer service representation by responding to customer queries and questions as well as updating basic account information in the system. It delivers voice and text information to the users via a web browser, kiosk, or mobile interface for resolving their problems. Amelia is a virtual agent avatar that redefines how enterprises operate by automating and enhancing a wide range of business processes. Amelia interfaces with consumers in astoundingly human terms, parsing questions, analyzing intent, and even sensing emotions to resolve issues more efficiently and effectively than customer service representatives.
  11. 11. Copyright © 2018 Accenture All rights reserved. 11 ROBOTIC PROCESS AUTOMATION Blue Prism is an enterprise RPA software to help organizations eliminate low-return, high-risk, manual data entry and processing work. Blue Prism helps organizations in deploying software robots to: • Build an agile digital workforce • Increase productivity with less resources • Improve cost savings and accuracy • Provide superior customer experience Robotic Process Automation (RPA) uses software to automate manual, time-consuming, or routine processes, which can help in reducing costs, reducing errors, improving efficiency, and increasing employee and customer satisfaction. Once instructed and configured, Robotic Process Automation tools can process transactions, manipulate data, and communicate with external systems in an autonomous way.
  12. 12. Copyright © 2018 Accenture All rights reserved. 12 NATURAL LANGUAGE PROCESSING Natural Language Processing is a method of interacting with intelligent systems, using a natural language such as English, to enable them to perform tasks as per instructions. This technology analyses text to interpret sentence structures, their meaning and intention, using statistical methods and machine learning. Siri is a voice-activated intelligent personal assistant developed by Apple that uses natural language processing to answer questions and perform actions. It is capable of voice interaction, music playback, making phone calls, sending messages, setting alarms and reminders, streaming podcasts, playing audiobooks, scheduling meetings and appointments, and providing weather, traffic, and other real-time information.
  13. 13. Copyright © 2018 Accenture All rights reserved. 13 PREDICTIVE ANALYTICS Predictive analytics is the science of getting computers to predict and provide recommendations based on previous trends without being explicitly programmed. It uses various statistical modelling and machine learning techniques to analyze historical data and predict future outcomes. Amazon uses browsing and purchasing history of customers to make recommendations of products that they may be interested in buying in future. Predictive Analytics helps amazon to: • Know what the customers are most likely to buy in advance • Determine the highest price a customer will pay for the products • Target recommendations and promotions • Improve supply chain management
  14. 14. Copyright © 2018 Accenture All rights reserved. 14 AI USE CASES BY INDUSTRY (1/2) An advertising giant M&C Saatchi collaborated with Clear Channel and Posterscope to develop an intelligent poster campaign. The poster uses a body-tracking technology to analyze the people standing in its vicinity and assesses up to 12 people at a time. It uses a combination of pictures from the “gene pool” to create and access ads that are likely to be successful. Marketing and Advertising Banks are combining AI technology with traditional processes to cut costs and increase service efficiency. AI technologies currently utilized in financial services include data mining, NLP, Machine Learning, and text analytics. The intelligent engine Wallet AI helps customers make informed financial decisions. Finance A law firm international city lawyers BLP uses AI to capture the details of a property from a Land Registry title deed. Major firms employ e-discovery systems to sift through documents, e-mails and records to identify material that could be relevant to a litigation. Law
  15. 15. Copyright © 2018 Accenture All rights reserved. 15 AI USE CASES BY INDUSTRY (2/2) AI is being extensively used in patient care and diagnostic systems. For example, IBM Watson Oncology specializes in developing patient-centric treatment plans. In Boston Children’s Hospital, clinicians feed genomic data into IBM Watson to develop a cognitive system for interpreting a child’s genome sequencing data. Healthcare The AI-powered “Cognitive Tutor” in the education industry can monitor and track learner’s progress and detect the extent of the learner’s understanding on the content. Another example is Cram101, which is an innovative learning solution that converts textbook content into interactive content with chapter summaries, practice tests, and flashcards. EducationRetail The IBM Watson AI provides data analysis and visualization solution for analyzing unstructured data and providing customer insights to retail organizations. Also, an AI platform known as inSTREAM can decipher patterns in unstructured data, which can be applied to customers’ emails for better understanding of their sentiments and improve customer service.
  16. 16. Copyright © 2018 Accenture All rights reserved. 16 HOW WILL AI BE PLACED IN LEARNING? As organizations are now transitioning from traditional learning to new age digital learning solutions, it is critical for the L&D function to adopt AI in developing learning strategies. AI can provide insights based on the data it gathers and analyzes, which will enable organizations to develop a better understanding of learner behaviors and predict their needs to recommend content. It can also automate various processes, and provide algorithm-driven design and development.
  17. 17. Copyright © 2018 Accenture All rights reserved. 17 AI CAN PROVIDE THREE LEVELS OF INTELLIGENCE TO LEARNING SYSTEMS Assisted Intelligence AI assists humans to perform some of their tasks faster and without errors. Assisted Intelligence aids in improving existing manual tasks, requires constant human intervention, and works well with defined inputs and outputs. Example: Automated grading for almost all types of assessments, automated learning content authoring, LMS uploading, and publishing of courses reduce the time and effort required. Augmented Intelligence AI enhances human intelligence rather than simulating independent intelligence. Augmented Intelligence can help employees to perform their tasks faster and better by enabling them to access knowledge resources and collaborate with fellow workers, experts and databases using AI techniques such as Machine Learning, Natural Language Processing, Image Recognition etc. Example: Machine-assisted solutions – eg. voice-driven and image-recognition- based performance support tools for servicemen in their field tasks. AI is developed for the future generation. Autonomous Intelligence establishes machines that perform in place of humans. Example: This field is still being researched upon and applications of Autonomous AI in learning is a future view. Autonomous Intelligence
  18. 18. Copyright © 2018 Accenture All rights reserved. 18 Intelligent Automation Use of AI to combine cognitive capabilities such as learn, reason, solve problems, and make decisions with automation technologies for streamlining complex and/or routine processes and deriving solutions. Enhanced Judgement Use of AI to support human intelligence by analyzing, evaluating, and recommending options for consideration and informed decision making. Leveraging AI to bring together people, devices, and services. It provides a personalized user experience by analyzing their needs and assisting them by providing real- time information or performing tasks. Intelligent Products Use of AI to develop new and innovative products and services that can sense, comprehend, act, and learn based on situations to accelerate growth of an organization. Enhanced Trust Use of AI to drive compliance, transparency, and trust within and outside an organization. It ensures that employees, customers, and other stakeholders believe in the benefits of the AI implementation. ENTERPRISE VALUE CREATION LEVERS OF AI Enhanced Interaction i
  19. 19. Copyright © 2018 Accenture All rights reserved. 19 ENTERPRISE VALUE CREATION LEVERS IN LEARNING Auto- authoring Curriculum Calibration Intelli-Device Mentoring Gamification Learner Profile Modelling AI can enhance the effectiveness of Corporate Training Programs through:
  20. 20. Copyright © 2018 Accenture All rights reserved. 20 INTELLIGENT AUTOMATION – AUTO-AUTHORING AI technologies can enable machines to apply cognitive skills, such as learning, reasoning, and decision-making for automating content design and development. Algorithm-driven design and analysis engines will help in gathering, analyzing, building, and reviewing content faster, thereby reducing the cost of development by 80-90%. How does it happen? What does the future look like? AI machines can curate and create variations of training content and test questions to cater to specific learning needs of different roles using learning materials repository. Another application of AI in learning is the automatic translation of content in other languages. Example: IBM’s Watson Content Analytics can gather and analyze unstructured content in documents, emails, databases, websites, and other enterprise sources to derive unique relationships. Another example is the Wordsmith’s platform that uses AI to convert data into logical coherent text using a template and a structured data set. In the future, AI is most likely to generate content from a set of rules and present it in a logical form. It can develop a draft storyboard from a set of compliance regulations, which can be reviewed and revised as appropriate by an Instructional Designer. With the advancement in AI capabilities, it may be possible to provide content suggestions, fill in content gaps, suggest objectives, create concept maps, and also create content on its own. Organizations will benefit from AI as the content creation will be less expensive, less time consuming, and more accurate. Auto-authoring
  21. 21. Copyright © 2018 Accenture All rights reserved. 21 ENHANCED JUDGEMENT – CURRICULUM CALIBRATION E-learning giants and corporate content experts need solutions that can address the current and future learning requirements of their participants. AI technologies can be used to analyze and correlate large volumes of data, such as user preferences and usage patterns, and draw inferences for developing a comprehensive and effective curriculum. For an organization with a large number of employees, modifying and updating the curriculum on a regular basis to match their learner’s needs is a tedious and time consuming process. It requires a lot of data to be captured and analyzed and is often lesser prioritized as a result. AI teaching software can identify areas where participants are deficient and thereby recommend such areas for curriculum calibration. Example: xAPI makes training activity data of learner actions from desktop browser and mobile space readily accessible. With the use of right AI algorithms, machines will likely be able to analyze emails and voice calls to derive insights such as required competencies or performance gaps in a target audience. This will become a crucial input for curriculum designing. Expert systems will be able to examine training activity data, identify weak spots in a course or curriculum, suggest improvements, as well as make automatic adjustments to ensure effectiveness of a training program. Curriculum Calibration How does it happen? What does the future look like?
  22. 22. Copyright © 2018 Accenture All rights reserved. 22 ENHANCED INTERACTION – INTELLI-DEVICE MENTORING AI has made accessibility and scalability a reality in global classrooms. The human- machine interface of AI has the potential to change the way we access, learn, create, and retain information. AI introduces a two-way relationship between the learning system and the participants by implementing intelligent methods for analyzing and assessing knowledge and abilities as well as conducting surveys and interviews. AI is also used as conversational agents or chatbots to engage participants in discussions amongst themselves or with the instructors and supervisors, which promotes healthy collaboration. AI can also provide just-in-time support to trainers while delivering content. Example: Third Space Learning app uses analyzed audio and written data to identify qualities of a good teacher. The app can provide real- time instructions if the tutor deviates from the topic or speaks too fast. AI is likely to assist tutors in class-room and virtual trainings by answering questions that require high-order thinking and creativity as well as hold a conservation. AI will soon help in presenting content, monitoring responses to evaluation questions, and determining the best path to follow toward the mastery of a subject. Intelli-Device Mentoring How does it happen? What does the future look like?
  23. 23. Copyright © 2018 Accenture All rights reserved. 23 INTELLIGENT PRODUCTS – GAMIFICATION Game-based motivation, including competitions, performance rewards, and recognition increases productivity for 90% of workers. Research proves that learner retention can be up to 9% more using game mechanics. Gamification is the use of game mechanics to simplify and improve learning, develop skills, edify behaviors, or to increase people engagement in an organization. Most businesses have begun identifying opportunities to leverage gamification as an intelligent product for accelerating the growth of the organization. AI plays a crucial role in designing strategic games for learning content as it enables higher retention and recall. Some innovative ways of using AI in gamification design are power balancing, difficulty level adjustments, and tailored feedback mechanisms. Example: Google’s artificial intelligence unit, DeepMind, developed a software program called AlphaGo to take on a human challenger in Go chess game. This program uses a Monte Carlo tree search, which involves sampling the most promising moves and running through different scenarios. The Gamification 2020 report by Gartner predicts that gamification combined with other emerging trends and technologies will have a significant impact on Innovation, design of employee performance, globalization of advanced education, gamification of personal development, and emergence of customer engagement platforms. Gamification How does it happen? What does the future look like?
  24. 24. Copyright © 2018 Accenture All rights reserved. 24 ENHANCED TRUST – LEARNER PROFILE MODELLING It is predicted that in the next 5-7 years, the training systems will use AI to analyze the learner’s strengths and weaknesses, their understanding of the subject, and adapt themselves by devising individualized learning profiles based on their progress. AI helps in implementing adaptive learning by analyzing the abilities, interests, and requirements of different learner groups and assigning them appropriate learning materials automatically. Use of AI in adaptive learning will help in catering the learning materials to the unique needs of the participants by emphasizing on certain topics, repeating other topics not yet mastered, and teaching students at a comfortable pace. Adaptive online learning, as seen in examples such as Khan Academy, could see a bigger role, as computers will be able to tailor instruction and assessments based on student performance. AI will help organizations by preparing a comprehensive learner profile of the employees using their job performance data. AI can be integrated with the organization’s ERP system and Learning and Talent Development Platform to track employees’ daily performance and customize their learning plans accordingly. Based on an individual's learning style and constraints, such as internet connectivity and shift timings, AI will help organizations to map learning materials to individual learners. It will automatically sync learning content to individual’s calendar and send communication. Learner Profile Modelling How does it happen? What does the future look like?
  25. 25. Copyright © 2018 Accenture All rights reserved. 25 AI USE CASES IN LEARNING (1/2) Mika is an AI-based tutoring tool for providing one-to-one attention to students. The app is modelled to suit a student’s unique learning process, track their daily progress, and adapt lessons to overcome their weak areas. Netex Learning allows teachers to design curriculum and customize student materials across various digital platforms and devices in a learning cloud platform. They can provide tools for video conferences, digital discussions, personalized assignments, and learning analytics to show each student’s growth visually. CTI is an AI company that uses deep learning to create customized textbooks. The teachers import syllabi into a CTI engine, which learns the content and uses algorithms to create personalized textbooks and courses based on the core concepts.
  26. 26. Copyright © 2018 Accenture All rights reserved. 26 AI USE CASES IN LEARNING (2/2) BiLAT is a game-based learning environment to practice skills in conducting meetings and negotiations in a specific cultural context. It enables learning via interaction with AI-driven simulated characters, who respond to learner’s actions and provide unsolicited hints about the appropriate action. The ARGUNAUT project combines AI analysis module with the online realtime monitoring module and helps students to stay on topic, elicit contributions from all members of the groups, and generally guide the students toward fruitful discussion and collaboration. Thinkster Math is a mathematics tutoring app with a personalized teaching style. The app presents each student with problems appropriate to their skill set and then analyzes how they’ve derived an answer. It improves each students’ logic process by providing video assistance as well as immediate and personalized feedback.
  27. 27. Copyright © 2018 Accenture All rights reserved. 27 WHAT AI OFFERS TO LEARNING? AI driven game-based training can make learning more fun and interactive. It can also help increase retention of content and improve performance of the learners. Inclusive It can provide trainers and learners the tools that allow them to respond not only to what is being learnt, but also to how it is being learnt, and how the learners feel. Personalization Flexibility It can help learners develop the knowledge and skills that employers are seeking, and it can help trainers create more sophisticated learning environments. AI has the potential to make learning accessible to more people. In places, where trainers are unavailable, a robust AI system can be used to teach learners with minimal or no engagement from a human trainer. Engaging AI offers the possibility of learning that is more personalized, flexible, inclusive, and engaging.
  28. 28. Copyright © 2018 Accenture All rights reserved. 28 FUTURE OF AI IN LEARNING 2018 By 2018, 20% of all business content will be authored by machines according to Gartner Incorporation. 2021 According to Research and Markets, the use of AI in the education sector will grow to 47.5% through 2021. 2035 AI’s economic output in 16 industries will be $14 trillion and corporate profitability will increase by an average of 38%.
  29. 29. Copyright © 2018 Accenture All rights reserved. 29 SUMMING UP, THE FUTURE HAS ALREADY ARRIVED! With the fast pace advancements that we’ve already witnessed, AI in the future is expected to: Provide personalized tutorial to the learners based on their existing knowledge, experience, strengths, and weaknesses. Create new forms of intelligent e-learning platform allowing students to explore, collaborate, and as test their knowledge in multiple scenarios Free trainers from monotonous and time consuming tasks, thereby helping them to empathetically attend to the learners by.
  30. 30. 30 THANK YOU! This Thought Paper was created in the Think Tank Center of Excellence operating in Talent Development and Learning Practice of Accenture Capability Network. If you would like to discuss about Artificial Intelligence in Learning, feel free to reach out to us. Developers: Shweta Panwar shweta.panwar@accenture.com Debamitra Dasgupta debamitra.dasgupta@accenture.com Sponsors: Devanshi Sahay devanshi.sahay@accenture.com Marek Hyla marek.hyla@accenture.com

×