AI Insights

How to Solve Ethical Issues in AI Through Collaboration

May 15, 2022


article featured image

How a community of technology changemakers from around the world is on a mission to solve ethical issues in AI development through collaborative and diverse teams. This article describes the founding story of Omdena.

By Rudradeb Mitra and Michael Burkhardt 

Ethical issues in AI

Photo by Pablo García Saldaña on Unsplash

Photo by Pablo García Saldaña on Unsplash

While many initiatives have been formed (such as OpenAI or a recent joint program by Harvard and MIT) many problems remain to be solved.

Building ethical AI solutions means to answer the “hard questions” — Is it moral? Is it safe? What value & positive impact does it bring? And only by answering these questions properly, socially-beneficial AI can be realized.

A recent study by the Nuffield Foundation explored the topic of AI ethics deeper and concluded that despite a shared set of concepts and concerns is evolving, some main gaps remain to be filled; most significantly a lack of clarity or consensus around the meaning of central ethical concepts as well as insufficient attention towards tensions between ideals and values.

How can we get closer to solving these problems?

A community as a mean to unite people and values

There is no power for change greater than a community that discovers what it cares about.

Margaret J. Wheatley

A community is formed first and foremost because its members share common values, interests, and goals.

With regards to AI that means to bring together the right people that associate with a problem and are willing to solve it together. Considering the tremendous impact that most of today’s AI solutions have on people and society, it would be detrimental to built solutions in isolation from the people and social circumstances that make them necessary in the first place.

Therefore, we need to move away from individuals or small AI teams but shift towards communities of people solving a problem they deeply care about

50 AI enthusiasts, two experts, and a common mission

Late in 2018, I worked with a startup in India to build a clean tech solution.

The technical goal was to build a sophisticated Machine Learning model to increase the adoption of rooftop solar panels. 

Instead of relying on a small team, I managed to build a project community of more than 50 AI engineers and enthusiasts.

All members shared the following mission:

Through collaborative work and shared learning, we will reach our goals faster while boosting our knowledge and building a solution that creates a positive impact in the clean tech sector.

Using Machine Learning for Low-Resolution Satellite Images

Using Machine Learning for Low-Resolution Satellite Images

The enthusiasts came from all over India, have never met each other but were united through the power of community and collaboration.

Some of the students — From the top — Jitendra, Abhigyan, Raghav, Devendra, Rasika, Iresh, Jerin Paul, and Shivani.

Some of the students — From the top — Jitendra, Abhigyan, Raghav, Devendra, Rasika, Iresh, Jerin Paul, and Shivani.

Introducing Collaborative Artificial Intelligence

Collaborative AI means to merge the concepts of community and collaboration by involving organizations, experts, and enthusiasts to build solutions that are ethical, trusted, and value-creating; and as a result beneficial for society.

In the words of one of the collaborators:

“Any real-world problem could be best solved if a group of people comes together to put in their dedicated efforts. When it comes to the collective efforts of dedicated individuals, success is bound to occur!” Iresh Mishra, a 4th-year student of Shri Mata Vaishno Devi University, India.

The advantages of Collaborative AI are as follows.

1. Empowering talent globally to acquire real-world skills

With today’s technological advancements, online courses, and available tools, talent is everywhere and can be accessed easily.

In the Solar Machine Learning project, tasks were announced in the community and collaborators took up those tasks according to their skill-set.

The advantages for collaborators:

  • Work on real-world data
  • Improved communication skills through frequent demos of their work
  • Steeper learning curve through shared learning
  • Mentorship by leading AI experts

Abhigyan Das, who helped to gather the data says “I think such a community model should be followed by more organizations because we as students can gain not only first-hand experience about work but can also learn a lot of things which are not available in any course.”

I feel working like this meets both ends; the organization gets the best of enthusiastic people and the members add to their learning curve by working towards delivering their own product while making a real impact. This is only possible when there is mutual trust and respect for each other.”, adds Smriti Bahugana.

2. Building trusted and value-creating solutions

For organizations, Collaborative AI means to harness crowd wisdom, diversity, and inclusion united in one project community.

How organizations benefit:

  • More trust
  • Faster results
  • Access to leading AI experts and domain knowledge
  • More and better data

Especially the trust generated by community-driven development can significantly help to make people more willing to share their data. Something which is receding in products built by large corporations.

3. Getting access to data

Having access to a larger amount of high-quality data through forming project communities stems from two aspects.

First, leveraging crowd wisdom by having more people involved results in innovative approaches to gather and work with data.

Additionally, a large project community compared to a small team of people generates and prepares high-quality data faster and more efficiently.

One of the students, Rasika Joshi, says “I could focus more on building Neural Network and do training over required formatted data set just because I was working with fellows and they provided me with the data in a given time frame.”

4. Democratizing AI and solving pressing ethical issues

There are tens of thousands of AI engineers and data scientists, who find it extremely hard to work on real-world projects. By connecting organizations and impactful problems with the right people, we have the power to make this technology accessible to a broader audience and all contribute to the democratization of AI.

Building an equal opportunity world

AI has the potential to become one of the greatest technologies of today’s and tomorrow’s time, and it is in our hands to make this a reality.

Building an equal opportunity world

Building an equal opportunity world

Imagine a world where no matter where you are born or live if you are talented you get equal access to work and opportunities as anyone else living in any other part of the world.

This is our vision at Omdena of how the future of work and education should be and we are willing to contribute our part.

We believe community and collaboration are two of the main ingredients to realizing this future and we welcome organizations, experts, and enthusiasts to join our community to solve not only ethical issues in AI but also some of the most pressing problems in the world.

Related article: Building the Future Talent for Ethical AI

Ready to test your skills?

If you’re interested in collaborating, apply to join an Omdena project at: https://www.omdena.com/projects

Related Articles

media card
Improving Data Privacy Through Federated Machine Learning
media card
AI in Cybersecurity: Navigating the Future of Digital Defense
media card
Revolutionizing the Oil & Gas Industry with Advanced AI Solutions Empowering the Future of Energy