Artificial intelligence has enormous power to transform society. Current breakthroughs should increase economic opportunity, alleviate human suffering, and expand access to learning and knowledge. The responsible and ethical use of this technology should vastly improve the quality of life and standard of living for people everywhere.

To borrow the adage from a legendary comic book, with great power comes great responsibility. This is why ethics remains at the core of our community, and why we continue to be driven by our shared values and principles as we develop and adopt exciting new applications. 

At Collective[i], we not only prioritize ethics in the development of AI, we believe it is our duty to help both industry and government create the infrastructure needed to support, encourage and accelerate the responsible development of AI. In addition to enforcing our Responsible AI Guidelines and policies in the development, deployment and usage of AI at Collective[i], as an AI company, our business model is designed to accelerate prosperity, democratize access to economic opportunities, and reward ethical outcomes. 

In the spirit of transparency, here are six principles that we consider in the development of our technology and that we believe are worthy of your consideration as a user. 

1. Social benefit: The AI products that are developed and deployed by Collective[i] aim to promote the public interest and lead to greater economic prosperity for everyone. Collective[i]’s business model is designed to help and incentivize people to collaborate, and to produce more profitable outcomes for everyone. Great AI elevates us. It creates more rewarding work and allows us to operate more effectively. It is not meant to replace people or further entrench existing inequalities. For example, Collective[i]’s enterprise application removes menial work (like CRM data entry and sales forecasting) to free up time for sales professionals to build meaningful relationships that lead to revenue. Collective[i]’s Connectors feature surfaces helpful business connections for people who might not normally have access to elite networks. To the extent that Collective[i]’s products change the way people work, we are committed to finding ways to re-train and upskill people for an AI-first world.

2. Fairness and anti-bias: One major pitfall of deep learning AI is the potential for bias to be encoded into the model stemming from biases present in the training data. If a model is trained on a dataset that is predominantly male, for example, it may perform poorly on tasks related to female individuals. We recognize that bias is the norm and we aim to mitigate it by the inclusion of a wide range of data sources and perspectives. Collective[i] will continue to capture data from our diverse community and leverage diverse talent to train and build our AI. We will also regularly audit (both internally and with the help of 3rd parties) our AI system for bias and take steps to mitigate it if detected. Finally, we do everything in our power to ensure our products are designed and developed in a way that makes them accessible to a diverse range of users, including people with disabilities and those from various socioeconomic backgrounds.

3. Transparency and accountability: One concern with deep learning AI is that it's a “black box”, meaning it’s almost impossible to reverse engineer the thinking (and the specific data) that led to an output. Because deep learning models are complex and operate by learning patterns in data, it can be tough to understand how they arrived at a particular decision or prediction. To the extent possible, we will ensure that our clients and other stakeholders understand how Collective[i]’s AI products work and what data it is using. We will be honest about the limitations of our products and ensure that it is not being used to make decisions that it is not equipped to handle. We also encourage our community to share via Discord and other channels areas where they would like to better understand how we developed particular features and/or functionality. 

4. Privacy and security: Because the AI- and machine learning-powered products we offer, including the Collective[i] platform and Intelligence.com, collect and process large amounts of data, we must ensure that individual rights and freedoms are respected, and that our data management and storage practices are secure and comply with applicable regulations. We must also ensure that our clients’ data is not being used for unintended purposes or shared with unauthorized parties. For more information on our privacy program and philosophy of data minimization, please visit our Trust Center. 

5. Interoperability and compatibility: We believe the future of AI includes how foundational models work together, so we foster a collaborative ecosystem and encourage Collective Intelligence that promotes innovation and benefits all stakeholders. We actively participate in and contribute to open standards and protocols that facilitate interoperability and compatibility across AI systems. We engage with industry partners, research institutions, and relevant stakeholders to promote knowledge sharing, collaborative research, and the development of widely-accepted best practices. We also develop AI systems with a focus on modularity and adaptability, ensuring that they can easily integrate with other systems. To ensure people can take advantage of our products, we prioritize having well-documented APIs, reusable components, and clear guidelines for system integration to facilitate seamless collaboration and interaction between AI systems and their users.

6. Human oversight: We recognize the importance of human oversight in AI development. AI should be used to augment human decision-making, not replace it entirely. Through the establishment of an internal Responsible AI Committee made up of engineers, product managers, data scientists and compliance leaders, we will ensure there is human oversight of the development, deployment, adoption and implementation of AI products at Collective[i], including over any automated AI decision-making. In addition to reviewing and assessing the ethical implications of our AI products and ensuring they are developed and deployed in accordance with our core principles, the Committee will be responsible for training and educating colleagues on ethical considerations related to AI technology and will present any findings or concerns to Collective[i]’s Executive Team.

Finally, as with any technology, the ethical use of AI depends on the responsible application of the technology by its users. This includes considering the potential impacts of the technology on society and taking steps to mitigate any negative consequences. The ethical use of AI requires careful consideration of a range of issues, among other things, bias in training data, transparency, privacy, and human oversight. Together, through awareness and our actions, we can ensure that AI is developed and used in a way that benefits our community and society as a whole.