Perceptyx Perspective on Artificial Intelligence, Machine Learning, Natural Language Processing, and Generative AI

Overview of artificial intelligence

As a preliminary matter, we find it helpful to provide an overview of AI generally, including machine learning, natural language processing, generative AI, and narrow (or weak) AI.

Artificial intelligence

Artificial Intelligence is a broad field of computer science focused on creating machines or systems that can perform tasks requiring human intelligence. These tasks include learning, reasoning, problem-solving, perception, and natural language understanding. AI can be categorized into two types: narrow AI, which is designed for a specific task, and general AI, which aims to exhibit human-like intelligence across a wide range of tasks.

Narrow artificial intelligence

Narrow AI, also known as weak AI, includes techniques such as rules and heuristics and refers to artificial intelligence systems that are designed and trained for a specific task. These systems excel in performing the predefined task but lack the general cognitive abilities and understanding that characterize human intelligence. Most of the current AI applications, including many machine learning models, fall under the category of narrow AI as they are specialized for specific functions. Perceptyx uses narrow AI for the following product areas: Activate, Cross-Survey Projects, and Third-Party Comments.

Machine learning

Machine Learning is a subset of AI that involves the development of algorithms and statistical models that allow computers to learn from data. Instead of being explicitly programmed for a task, machine learning systems use patterns and insights from training data to make predictions or decisions. Machine learning is considered narrow AI and is used for the following product areas: Advanced Reporting, Cross-Survey Projects, and Third-Party Comments.

Natural language processing

Natural Language Processing is a branch of AI that focuses on the interaction between computers and human language. Natural language processing enables machines to understand, interpret, and generate human language, allowing for tasks such as language translation, sentiment analysis, and speech recognition. Natural language processing technologies are crucial for enhancing communication between humans and machines, making it possible for computers to process and respond to natural language inputs.

Perceptyx uses natural language processing in the following product areas: Advanced Reporting, Cross-Survey Projects, and Third-Party Comments.

Generative artificial intelligence

Generative AI involves the creation of models that can generate new, original content. This can include text, images, videos, or other forms of data. Generative AI models use advanced algorithms to produce content that closely resembles what humans might create. Applications of generative AI range from creative endeavors, like art and music generation, to practical uses in content creation and simulation. We use generative AI for Comment Copilot and AI coaching, each of which is available for customers upon request.

Perceptyx does not currently use generative AI in any other products. If we were to ever use generative AI in our other products, we would do so in conformance with the highest ethical standards.

Artificial intelligence, privacy, and security

Here's how we integrate AI, machine learning, natural language processing, and generative AI while maintaining our commitment to privacy and data security:

Privacy-focused machine learning

• Our machine learning practices prioritize user privacy by excluding any customer-specific data from the training and refinement of our models.
• We adhere to strict anonymization and security protocols, ensuring that the development of our machine learning and natural language processing algorithms is conducted in compliance with the highest privacy standards.

Dynamic natural language processing

• Natural language processing plays a crucial role in our products, enabling our products to comprehend language, interpret it, and provide actionable insights to the end user.
• We utilize various language interpretability tools to understand whether natural language processing models perform consistently with changes in pronoun gender, verb tense, and topic, and leverage data augmentation techniques to mitigate any existing bias.
• We use several methods to audit the internal function and decision criteria of the models, including Lime and FairML, to ensure machine learning fairness and lack of bias.

Generative artificial intelligence

• Generative AI prompts available to the user are extensively tested and validated via an internal testing framework. This is to ensure maximal user value and minimal risk of hallucinations or personally identifiable data leakage.
• Users can interact with generative AI features through a curated prompt library as well as an open-ended chatbot depending on the product(s) in use.

Transparency in model training and deployment

• We are committed to transparency in our AI and machine learning practices. Users are informed about the types of data used in model training, with a clear emphasis on the exclusion of customer-specific data.
• This commitment extends from the initial stages of model development through to deployment, ensuring that users can trust the integrity of our AI and machine learning solutions.

User-centric ethical standards

• Our ethical standards encompass not only responsible data handling but also the ethical use of machine learning and AI technologies.
• By affirming that customer data is not utilized in our AI and machine learning models, we prioritize the ethical considerations necessary to build and maintain user trust.

In conclusion, at Perceptyx, our perspective on AI revolves around a seamless integration of narrow AI and responsible use of generative AI to deliver unparalleled user experiences. Simultaneously, our unwavering commitment to privacy, ethical standards, and transparency ensures that our advancements in technology are aligned with the values of trust, security, and user empowerment.