Generative AI “enriches” customer service, marketing, sales and ecommerce teams. Second generation artificial intelligence introduces the concept of “augmented humanity” which, like augmented reality, improves the characteristics of company resources without thinking of replacing them.
We are faced with a mature interpretation of artificial intelligence, which makes us definitively overcome the fear of being replaced by an algorithm. Generative AI aims to enhance the human experience, thanks to machine learning and the LLM, Large Language Model, with intelligent information processing far superior to any human being. For example, according to this view, the chatbot no longer interacts for humans, but advises them in real time on which answers to provide in the most effective and most informed way possible.
The concept of “increased humanity” was repeatedly mentioned by Marc Benioff, co-founder and CEO of Salesforce, during the AI DAY held in New York in mid-June. Artificial intelligence for Salesforce means Einstein, an application platform created in 2014 with the intention of putting machine learning at the service of the interaction between company and customer at all stages: from the marketing and commercial approach to after-sales assistance.
Einstein GPT, generative AI at the service of CRM
Today Einstein evolves into Einstein GPT, precisely because it is based on generative artificial intelligence, offering “artificial” content, thanks to the integration with OpenAI and with solutions from various companies acquired by Salesforce over the last few years. Business-supporting computing models provide real-time intelligence to support enterprise resources across sales, service, marketing, ecommerce, and IT. Today Einstein GPT is able to provide a trillion – a billion billion – forecasts per week allowing you to convert leads, contact opportunities, 3 times faster. This is because through Einstein GPT not only are millions of (authorised) customer information collected but, above all, it is processed to transform it into suggestions and automated activities to support CRM, i.e. customer relationship management.
Generative AI is above all LLM, a computer model based on neural networks capable of searching, generating, synthesizing and translating content up to predicting it, analyzing huge quantities of public data or data generated by the applications present in the company.
As we know, using these models raises questions related to the reliability of the contents, security, privacy and, finally, ethics which depends on how the models themselves are educated. All these issues have not been overlooked in the construction of Salesforce’s Einstein GPT, indeed they are the starting point in the dialogue on the subject between the supplier and his customers.
Trust is at the forefront of the values underlying Salesforce’s generative AI model, Marc Benioff has repeatedly stressed. This means having deep respect for the data generated by machine learning and deep learning. It means, in particular, to process them without ever violating the privacy of consumers. A condition that is particularly important in sectors such as banking or healthcare, but not only.
In Salesforce’s generative AI model, data must be disconnected from who generated it while, at the same time, guaranteeing knowledge to the client company that uses it. It’s a bit like what is done in the field of marketing analysis: identifying and classifying consumer wishes based on categories that do not consider sensitive information. It doesn’t matter what his name is, what his gender is and what ethnic group he belongs to who has expressed a certain desire in a conversation within a social network, what matters is identifying the desire, measuring it and contextualizing it.
To demonstrate how important ethics are in data processing, in dialogue with client companies, Salesforce immediately involves its ethics team. The principles of Salesforce’s ethical approach were summarized by team leader Kathy Baxter. “Your data is not our product” is the first rule that disavows the ancient axiom born at the dawn of Web 2.0, while ethical platform design and behavior is the last. Only by adhering to these principles in the development of projects for one’s customers can a generative AI be guaranteed that is “friendly” and, at the same time, functional to improve the relationship between company and customer.