The first edition of HerStory’s Women on a Mission Summit 2019, aimed at championing women entrepreneurs, innovators, and leaders, was held at The Lalit Ashok, Bengaluru, with over 700 people in attendance.
One of the highlights of the day was the Microsoft AI Masterclass: Envisioning AI For Good, a workshop that outlined the opportunity that Artificial Intelligence (AI) brings to developers, entrepreneurs and business leaders to contribute and accelerate India’s technology landscape in a responsible and ethical manner.
In India, Microsoft is democratising AI to empower individuals, organisations and societies towards fulfilling the promise of holistic growth by building on four key pillars:
1. Enabling transformation across industries: Leveraging AI to facilitate innovation and accelerate digital transformation, with a focus on industries critical to economic growth.
2. Forging coalitions for innovation: Engaging with government, and other key stakeholders to create a collective approach to AI.
3. Building a future-ready workforce: Empowering individuals, across all demographics, with skills for an AI-enabled world, and the creation of a future-ready talent pipelines.
4. Creating sustained societal impact: Applying AI technologies to address environmental sustainability, accessibility, humanitarian action and other crucial challenges.
The Masterclass was conducted by Lathika Pai - Country Head, Microsoft for Startups, MENA & SAARC; and Sandeep Alur - Director – Partner Engagements, Microsoft India. Introducing the speakers, Dipti Nair, YourStory’s Editor at Large said that the workshop would also highlight upon the need for a diverse outlook when dealing with data in order to drive inclusion across products and services.
The power of AI
Pai kicked off the Masterclass session, and shared AI’s power to bring technology to people anywhere in the world. She explained that it was “important that the advancements that have been made are extremely valuable not just to the elite amongst us, but to everyone.” Pai spoke about Microsoft’s AI for Accessibility program, which harnesses AI and inclusive design principles to amplify human capabilities for people with disabilities and help make the workplace more inclusive. A five-year, USD 25 million programme, AI for Accessibility aims to accelerate the development of accessible, intelligent AI solutions that build on recent advancements in Microsoft Azure cognitive services. Developers can win grants of Azure compute credits to create intelligent apps that understand and better interpret people’s needs under the categories of employment, daily life, and communication & connection.
Pai also touched upon the Seeing AI app, which harnesses the power of AI to provide cues on people, text, and objects to low vision and blind individuals by scanning and analysing their surroundings in real time. In addition to recognising and describing people, the app also has the power to identify currency notes, read snippets of text, and describe surrounding objects.
As well as highlighting AI’s potential to equalise societies and benefit the differently abled, Pai addressed common risks and fears associated with the development of AI. She explained that the most common area of concern is the potential for abuse of information, with modern consumers becoming increasingly protective of their personal data. Pai put attendees at ease, however, by elaborating on Microsoft’s ethical decision framework that guides the cross-disciplinary development and use of Artificial Intelligence. Driving all AI-related development at Microsoft, the framework consists of six principles: fairness, reliability & safety, privacy & security, inclusiveness, transparency, and accountability. The company firmly believes that every individual, organisation and society must engage in a reliable and disciplined approach to AI in order to realise technology’s full potential.
AI in practice
Following Pai’s portion of the Masterclass, Sandeep Alur stepped up to demonstrate some of Microsoft’s AI innovations, and their applications in the real world. He began with a demo of Microsoft Translator's live translation capabilities, highlighting the technology’s potential to bridge communication gaps between users. The translator has live captioning capabilities, designed for the hard-of-hearing, as well as live translation support for over 60 languages. With an immense scope of application, this technology has already been implemented in classrooms and businesses to promote inclusivity and accessibility.
Alur also pointed out that while remarkable advances have already been made, we are still in the genesis of AI technology, with large skilling gaps and a notable lack of affordable development infrastructure to train ML models. He elaborated on the difference between Artificial Intelligence, Machine Learning, and Deep Learning - three of today’s most common tech buzzwords - and explained that the successful emulation of biological neural networks has helped researchers and developers achieve human parity. In short, this means that Microsoft’s AI applications have matched or exceeded human accuracy in vision, speech, language translation, and reading comprehension - a feat that brings us closer than ever to implementing AI solutions in our daily lives.
Alur touched on the importance of data scientists for further AI developments, because as he put it, “people are amassing data, but what next?” He explained that, as the amalgamation of statistics and development, data science is the key to making sense of the vast quantities of data required to skill ML models. Alur’s session ended with a snapshot of Microsoft Azure’s AI solutions, including frameworks to distribute AI solutions to remote geographies, custom domain-specific models, and GPU support to develop and test ML models.
The Masterclass left attendees with ample food for thought on the social implications of AI and its still-untapped potential, and the session was peppered with insightful audience questions. Audience members asked about specific applications of Microsoft’s AI solutions, how governments can promote AI solutions for society, Microsoft’s solution to inbuilt bias in AI solutions, and how AI technologies can work in a variety of digital environments. Fielding each query, Pai and Alur left attendees with a deeper understanding of Microsoft’s commitment to crafting an inclusive, AI/ML-driven future that can solve emerging problems and tackle the challenges faced by diverse economies like India.