Salesforce AI Associate Exam Study Notes

Data values should be consistent across different records, fields, or sources.

Analyzing and understanding human language, both written and spoken.

AI systems should consider diverse perspectives and backgrounds in their design and development.

The ability to explain how an AI system makes its decisions, increasing trust and accountability.

Multiple variations of data values, like state or country names, affecting the consistency of the data.

AI models should be trained on a variety of datasets representing different groups, ensuring fairness and reducing bias.

Lack of transparency in AI systems makes it hard to understand their decision-making process.

The trustworthiness of AI systems is built upon factors like transparency, fairness, and accountability.

The amount of data used to train an AI model. A larger dataset generally leads to better performance and more accurate predictions.

A feature in Salesforce that uses machine learning to analyze data and create models that can be used for various tasks.

A type of AI that uses natural language processing (NLP) and natural language understanding (NLU) to interact with customers in a conversational way.

A feature in Salesforce that uses AI to classify customer cases into specific categories, making it easier to route them to the right experts.

A feature in Salesforce that uses AI to suggest relevant information or actions to users based on their past behavior and current context.

Ensuring that AI systems comply with data privacy laws and regulations, protecting sensitive customer information.

An AI application that helps prioritize leads based on their probability of converting into customers.

An AI application that predicts future sales or revenue using historical data and trends.

AI-powered feature in CRM that analyzes voice calls between a sales representative and customer, generating summaries or transcripts.

Accurate, complete, consistent, relevant, and timely data used by AI for effective decision-making.

The likelihood of a lead converting into a paying customer.

Maximizing the efficiency and effectiveness of sales processes.

Improving customer trust and satisfaction by providing personalized and accurate service.

Analyzing customer interactions to better understand their needs and preferences.

A situation where a company deploys an AI feature that is later found to be biased against specific groups, potentially causing harm. This creates ethical debt because it shows a lack of responsible development and potentially harms users.

The potential harm or risk created by unethical or irresponsible decisions or actions in AI development. This includes deploying biased algorithms, ignoring ethical considerations, and prioritizing profit above ethical conduct.

The potential for AI algorithms to discriminate against certain groups because the algorithms are trained on data that reflects existing biases in society. This leads to unfair and inaccurate results, impacting customer experience and trust.

A method of testing AI models by using datasets that represent diverse groups of people. This helps identify and mitigate biases in the AI system, promoting fairness and inclusivity.

The principle of AI development that emphasizes making AI systems transparent, explainable, and accountable. This means being able to understand how AI decisions are made and holding developers responsible for their impact.

A set of principles developed by Salesforce that guide responsible and ethical AI development. These principles prioritize fairness, transparency, accountability, and inclusivity.

A set of principles that guide responsible and ethical AI development, ensuring that these systems are fair, transparent, and accountable.

A type of bias where AI systems reflect and enforce the values and norms of a particular society or culture.

The accuracy, completeness, consistency, relevancy, and timeliness of data used by AI systems.

Using low-quality data in AI applications can lead to inaccurate predictions, biased outputs, and ultimately, ineffective systems.

AI systems that learn from data to predict future outcomes, like sales forecasts or customer behavior.

AI systems create original content, like images, text, or music, based on existing data.

AI systems that generate new content should be trained on diverse and representative datasets to avoid biases and promote inclusivity.

AI systems should be designed and developed ethically, considering fairness, accuracy, and social impact.

The process of evaluating the quality of data used in AI systems. It helps ensure the accuracy, completeness, consistency, and relevance of data for reliable AI predictions.

Data that describes characteristics of a population, like age, gender, or income. It can lead to bias if used to discriminate or treat people differently.

The challenge of human AI collaboration arises when people become overly reliant on AI, potentially leading to less critical thinking and oversight.

Salesforce excludes demographic data from AI models to mitigate bias and ensure ethical use. This emphasizes using behavioral data instead of personal data.

AI models should be trained on diverse datasets to minimize bias and ensure fairness across different groups.