This community is in archive. Visit community.xprize.org for the current XPRIZE Community.
About the Gender Data Gap Prize Design
XPRIZE
Posts: 193 admin
XPRIZE is excited to officially begin the Gender Data Gap Prize Design. Sponsored by The XPRIZE Guardian Circle, this ambitious initiative strives to exponentially improve gender equity by incentivizing and accelerating innovations that close the gender data gap.
By Gender Data Gap, we are referring to the reality that the vast majority of data collected perpetually excludes information about the perspective and experience of women. This type of biased data is used to inform design across all facets of the world in which we engage.
To prevent a cycle of algorithmic bias and enable an equitable future for humanity, XPRIZE is designing a series of challenges to incentivize the crowd to create, analyze, use robust data sets that will then be fed into AI and machine learning technologies which will produce unbiased insights that tell the story of all humanity —not just a subset of it.
We are calling on gender, data and technology experts to join us in this Prize Design, which will provide the outline of what the winning team must accomplish to be awarded the prize.
The Missing Data Problem
Data is the basic building block of the modern world. It informs our institutions, shapes our cities, determines how our products are made, and guides policymakers and political systems. For centuries, men have designed the world in their own image: using their worldview, physiology, habits, and priorities as the default. These biases are now embedded in society and the effects that this ‘unthinking’ has on 50 percent of our population is shocking. The world as we know it has been shaped by men, using an average 5’9”, 155 pound male as a substitute for representation for the entire adult human population.
The implications of this assumption range from small inconveniences to matters of life and death. From Google voice recognition software being 70 percent less likely to recognize female voices to 17 percent of women being more at risk of death and 47 percent more likely to be seriously injured in a car crash, it’s clear that designing a world tailored for the average man has had extreme consequences.
Though these design flaws and their consequences were unintended, correcting them is essential for the creation of a more equitable world. As artificial intelligence (AI) and machine learning replace manual design and human oversight, society gains efficiency. However, the inputs used to determine these machine-made choices rely on flawed data that lacks critical information about female preferences, behaviors, and needs. Incomplete or skewed training datasets, bias in labels used for training, and bias introduced in the features and modeling techniques themselves contribute to inaccurate gender data. AI is learning gender bias from humans. Biased data leads to biased insights, biased algorithms, biased solutions and policies, and ultimately a biased world that is designed only to meet the needs of the few.
The Goal
The XPRIZE Gender Data Gap Prize design seeks to expose biases in data and design and to incentivize the collection of complete and comprehensive data sets reflective of all genders, disrupting the destructive gender bias status-quo and building a future in which the needs and preferences of all humanity are met.
Your Role
We know your time is precious, and we greatly appreciate your participation and input. In addition to contributing to transformative breakthroughs, community participation also allows you to do the following:
XPRIZE Team
This Prize Design is led by @Kathleen_Hamrick, @Aaron_Denham and @Karan. @Shashi is the community manager.
Need Help?
If you have questions or need help with the community, leave a comment here or contact the community manager, Shashi Rai at shashi.rai@xprize.org.
By Gender Data Gap, we are referring to the reality that the vast majority of data collected perpetually excludes information about the perspective and experience of women. This type of biased data is used to inform design across all facets of the world in which we engage.
To prevent a cycle of algorithmic bias and enable an equitable future for humanity, XPRIZE is designing a series of challenges to incentivize the crowd to create, analyze, use robust data sets that will then be fed into AI and machine learning technologies which will produce unbiased insights that tell the story of all humanity —not just a subset of it.
We are calling on gender, data and technology experts to join us in this Prize Design, which will provide the outline of what the winning team must accomplish to be awarded the prize.
The Missing Data Problem
Data is the basic building block of the modern world. It informs our institutions, shapes our cities, determines how our products are made, and guides policymakers and political systems. For centuries, men have designed the world in their own image: using their worldview, physiology, habits, and priorities as the default. These biases are now embedded in society and the effects that this ‘unthinking’ has on 50 percent of our population is shocking. The world as we know it has been shaped by men, using an average 5’9”, 155 pound male as a substitute for representation for the entire adult human population.
The implications of this assumption range from small inconveniences to matters of life and death. From Google voice recognition software being 70 percent less likely to recognize female voices to 17 percent of women being more at risk of death and 47 percent more likely to be seriously injured in a car crash, it’s clear that designing a world tailored for the average man has had extreme consequences.
Though these design flaws and their consequences were unintended, correcting them is essential for the creation of a more equitable world. As artificial intelligence (AI) and machine learning replace manual design and human oversight, society gains efficiency. However, the inputs used to determine these machine-made choices rely on flawed data that lacks critical information about female preferences, behaviors, and needs. Incomplete or skewed training datasets, bias in labels used for training, and bias introduced in the features and modeling techniques themselves contribute to inaccurate gender data. AI is learning gender bias from humans. Biased data leads to biased insights, biased algorithms, biased solutions and policies, and ultimately a biased world that is designed only to meet the needs of the few.
The Goal
The XPRIZE Gender Data Gap Prize design seeks to expose biases in data and design and to incentivize the collection of complete and comprehensive data sets reflective of all genders, disrupting the destructive gender bias status-quo and building a future in which the needs and preferences of all humanity are met.
Your Role
We know your time is precious, and we greatly appreciate your participation and input. In addition to contributing to transformative breakthroughs, community participation also allows you to do the following:
- Network with diverse stakeholders
- Brainstorm with top experts
- Promote your work
- Earn rewards, such as online gift cards
- Be considered for blogs and podcasts
XPRIZE Team
This Prize Design is led by @Kathleen_Hamrick, @Aaron_Denham and @Karan. @Shashi is the community manager.
Need Help?
If you have questions or need help with the community, leave a comment here or contact the community manager, Shashi Rai at shashi.rai@xprize.org.
0
Comments
At present we are requesting members to advise us on:
• Which areas of health present the greatest gender data gaps?
• What emerging models and technologies are most promising for closing gender data gaps in health?
We would love to hear your thoughts on these latest discussion topics here.
You can introduce yourself to the rest of the community here.
Thanks.
I think the 115 there is a typo. 155? 165?
That was a good catch. We have rectified the error.
You can introduce yourself to the rest of the community here.
At present we are requesting members to advise us on:
• What are the most significant data needs in adolescent health? How might technology be leveraged to better collect needed data from adolescents (broadly defined as ages 10-19)?
We would love to hear your thoughts on this latest discussion topic.
Click here to learn more about what's next.