Cloud Hosting

Closing the Imagination Gap for GirlsWhy this is a critical step for creating an equitable future

The term artificial intelligence (AI) was coined 64 years ago at a scholarly conference. The AI field hasnt remained the theoretical province of computer scientists and mathematicians; it now is a pervasive part of everyday life. With a technology this powerful, it is critical to include the perspectives of all women, including those from underrepresented communities.

AI applications based on algorithms are found in robotics, machine learning, natural language processing, machine vision, speech recognition and more. These applications are found in homes, vehicles and myriad other aspects of daily life. Examples include facial recognition; robots helping older people live more independently at home; autonomous vehicles; smart watches; and drone safety systems.

AI applications must be able to reach conclusions and offer information. Some require the capacity to sense emotions in order to relate to people.

Today, women are making their way into AI and leading the way for more girls to enter AI careers. Theyre helping this burgeoning industry progress and innovate in ways that otherwise might not be possible. In essence, adding women to the teams creating components of AI fundamentally changes the suitability and functionality of a product or service by eliminating biases and better reflecting the needs of a wider group of users.

Taniya Mishra is director of Artificial Intelligence Research and lead speech scientist at Affectiva, which originated at MIT. The companys technology calibrates peoples speech patterns to recognize emotions.

Mishra offers some concrete examples of machine learning algorithms.

Algorithms are a set of rules logic or a set of instructions that you can give to a machine in order to get it to accomplish a goal to make it behave like a human being, Mishra says. It could be any goal. It could be lifting a block from one place to another. It could be understanding human emotion. All of these could be the goals for designing a machine learning algorithm.

The basic algorithm recipe tells the computer when to do x, then when to do y and then z. For this process to work right, the programmer must give the right instructions. For it to be inclusive, the programmer must think of the entire humanity of users, Mishra notes.

When it comes to diversity, AI benefits from including women and other underrepresented people. These voices must be included when writing instructions or algorithms to power machine learning or other elements of AI. The data gathered to support AI must also come from diverse groups of people, if the resulting algorithm is going to fully meet its potential.

For example, a small homogenous group designing a facial recognition program for a large heterogeneous group will miss the target if data about a variety of faces from the larger group is not represented. In other words, the algorithm is only as bias-free as the sources of data and the data sets.

To be effective, creators of AI-related applications need to be as diverse as the people using them.

Eighteen-year-old Betelhem Dessie is founder and chief executive officer of iCog-Anyone Can Code in Ethiopia. She also co-founded Solve IT, which provides technical resources to develop local solutions for community problems.

As different AI tools were being developed, I observed a lack of contributions from people of color and women, Dessie notes. The solution, I thought, was having early childhood tech education but also inspiring girls who are already in the workforce to pursue these types of career paths. The most rewarding part of my work is inspiring others particularly women and girls to pursue careers in technology.

But gender and diversity issues remain.

A 2019 article written by Kari Paul for The Guardian states the lack of diversity in the AI field has reached a moment of reckoning, according to findings by a New York University research center. The survey of more than 150 studies and reports, published by AI Now Institute, found that diversity disaster has contributed to flawed systems that perpetuate gender and racial biases, Paul writes.

One remedy is educating girls including girls of color sooner and more widely about the field and making appropriate educational opportunities and career guidance accessible to them early on.

Mastery of complex subjects is required, so girls must continue building on their basic math and science education, and intensify their focus as early as seventh grade. High school and certainly college may be too late to capture their interest so they can acquire the needed foundation.

Girls interested in AI will need to write code, algorithms and source data sets. Beyond that, they will need to understand and eliminate bias in data sets, as well as in applications designed to serve humanity.

Along with a rigorous early academic foundation, girls must develop social and emotional learning skills to help fuel their careers. These skills will prove beneficial whether they are leading a team or a company or programming soft skills into a robot.

A proven method for inspiring girls is to bring female role models working in AI into your classroom. Give girls a chance to ask these experts questions about their careers and personal stories. One way to start your search for experts is to inquire at universities and businesses from your local community; network with those professionals to build your sources.

Girls visions for the future are boosted when theyre introduced to female role models who demonstrate rewarding careers in the AI field and show that girls can excel in this arena.

As women enter the profession and assume leadership roles, society is seeing the advantages of perspectives they bring to AI systems.

For example, Mishra builds new systems that enhance peoples lives and give them a positive experience of interacting with technology. AI is ingrained into every aspect of our lives now and will be even more so in the future, says Mishra. Her advice to girls is to dream big: ambition is attractive and inspires those around you.

See original here:
Getting Girls Into the Artificial Intelligence Pipeline - Medium