This fast-paced era in which we live, work and thrive is increasingly populated by many types of technologies, including ones that focus on human speech and voice information. In addition to the ease of accessibility, and the economical repositories for storing this data, robust artificial intelligence (AI) driven tools that detect, distill, and translate specific outputs have the potential to enhance the human condition. However, as with most advanced technological systems, it also carries the possibility for misuse.
For the benefits of data technologies to offset the challenges, proper supervision, and fair, yet firm constraints should be engaged. The latter involves the realm of ethics, which at the most fundamental level is, “A system of moral principles relating to human conduct, as concerns the goodness and badness of motives and the ends of such actions.”
Ethical Guidelines
The ethical guidelines of many AI or algorithm-driven voice analysis companies and associations generally address at least one of the following five foci:
- Privacy
- Fairness
- Governance
- Joint-Advantage
- Transparency
The 1948 Universal Declaration of Human Rights describes the moral claim to privacy under Article 12:
Privacy
Generally accepted as an area of significance, the concept of privacy underscores most client-supplier interactions. Binding contracts between both parties typically feature language stating that consumer provided data is owned by the company, who agrees to use the data responsibly while protecting confidentiality.
Fairness
Fairness translates to using information gathered in a manner that respects the human source of that data, by maximizing safety while minimizing bias.
Governance
Governance concerns data management accountability and veracious reporting. It also guarantees the highest standards of measurement and the principled application of computer procedures in the solving of specific problems.
Joint-Advantage
Joint-advantage (also known as shared-benefit) is a concept that recognizes the rights of those who produce data in contributing to its control. Notably, there is no unanimity in the field as to how to engage a mutually beneficial working model of joint-advantage.
Transparency
Finally, transparency signifies openness and candor about the collection and utility strategies of data. However, considering that complete transparency can place the intellectual property of a business in jeopardy, this principle has not been fully endorsed and adopted by the majority of voice tech companies.
Ensuring Proper Safeguards
Thanks to recent data-breach scandals (in other tech fields) that have made headlines, voice analytics companies are increasingly taking safeguards. Across the board, companies are heightening security measures to protect data. They are increasingly vigilant in how datasets are curated, processed, and organized, such that objectivity is maintained, and machine stereotypes do not replace human ones.
At Clearspeed, we take ethics seriously. Multiple layers of precautionary processes are implemented in a system of checks and balances to ensure privacy, fairness, and adequate governance. Additionally, in order to maintain and protect the integrity of data and our technology, the latest and most impermeable of operational security (OPSEC) methodologies are continually updated.
Inclusive of natural language processing, signal recognition, machine learning, data analytics, and cognitive computing, AI carries the potential to change the world to a great and positive extent. Societies that use it will be smarter, healthier, and wealthier.
So long as computer algorithms and AI are (1) used in trustworthy ways by technology companies and (2) augment (rather than replace) human judgment, as was stated at the World Economic Forum in Switzerland in 2019, “There might be a day when AI will be considered as much of a human right as is food, water, or shelter.” That sounds just about right to us.
References:
Arvanitakis, J. (2018, November 28). What are tech companies doing about ethical use of data? Retrieved from https://phys.org/news/2018-11-tech-companies-ethical.html
Beauchamp, T. L., & Childress, J. F. (2019). Principles of biomedical ethics. New York, NY: Oxford University Press.
Hand, D. J. (2018). Aspects of Data Ethics in a Changing World: Where Are We Now? Big Data, 6(3), 176–190. doi: 10.1089/big.2018.0083
Liebert, M. A. (2018, September 24). Why are data ethics so challenging in a changing world? Retrieved from https://phys.org/news/2018-09-ethics-world.html
Moss, E., & Metcalf, J. (2019, November 14). The Ethical Dilemma at the Heart of Big Tech Companies. Retrieved from https://hbr.org/2019/11/the-ethical-dilemma-at-the-heart-of-big-tech-companies
Sejnowski, T. (The Center for Ethics in Science & Technology). (2020, May 6). The Deep Learning Revolution [Video podcast]. https://www.uctv.tv/shows/The-Deep-Learning-Revolution-35462
Sommers, C. & Sommers, F. (2013). Vice & Virtue in Everyday Life: Introductory Readings in Ethics. Belmont: Wadsworth.
The ethical challenges of AI: A leader’s guide (free PDF). (n.d.). Retrieved from https://www.techrepublic.com/resource-library/whitepapers/the-ethical-challenges-of-ai-a-leader-s-guide-free-pdf/