The ChatGPT FOMO …get started!

ChatGPT reached 1 million users within 5 days; quite a feat as compared to 3.5 years for Netflix and 10 months for Facebook. The use cases are exploding every day, however, there is growing anxiety in enterprises about missing the opportunity to apply the technology for the right business outcomes. While this technology is advancing fast, there are some initial foundational steps leaders can take to kick-start the journey.

1.????Define AI Policy for Enterprise:

Given the security, regulatory, and privacy issues and low entry barriers, it’s important to define and document a formal AI Policy for the enterprise. ?It should cover Ethical AI guidelines with topics like bias, discrimination, data transparency, and AI use, along with an audit mechanism to check the outcomes.

2.????Narrow down the right use cases

Finding the right-use case to start the journey is critical to show early success. Use Data Analytics from CRM/ITSM systems to narrow down on low-complexity, high-impact use cases. These use cases can demonstrate early success in specific areas like engineer productivity, CX improvement, or business growth. ChatGPT’s Information-summarization capability can synthesize distributed customer feedback as well as NPS & CSAT data to provide meaningful insights.

3.????Selecting the right technology: Build vs Buy

There are options like Azure Open AI, Salesforce Einstein-GPT, or specialized platforms like learn experts (https://learnexperts.ai/) ?which generates a training course within minutes.

Given the diversity of technology and the growing availability of industry-specific/horizontal platform solutions, enterprises have to choose the right strategy either to take a customized approach (Build) or take a readymade platform-based solution (Buy). The decision is about choosing a path that will help faster business differentiation and time to value for its stakeholders.

4.????Strive for accuracy to build trust?

The success of LLMs (Large Language Models) is limited by the quality and accuracy of the data used for training; they cannot understand the context on their own. Most organizations have data integrity problems, and their existing knowledge base needs much more enrichment and curation before it can be used to train the model. Therefore, it is vital to establish a knowledge management governance framework that will focus on knowledge selection, enrichment, and training process with the right approval workflows, keeping the human connection for validation.

Enterprises should leverage the inherent power of the robust “First Draft Response” from GPT and keep finetuning it to attain a better 2nd and 3rd response for improved accuracy.

There’s a dial within AI algorithms that developers can turn up or down, Enterprises need to have strong governance to manage the dial for predictable outcomes.

5.????Security governance

Security is the leading concern for enterprises, hence the adoption plans need to be viewed holistically. This includes overall architecture, data readiness, model training process, data residency, and the authentication techniques being used. It also depends on the solution you choose. For example, Einstein GPT trains using a company’s customer data, based on security principles that maintain secure access to your data, within the Salesforce trust boundary. This protects the privacy of your customers’ personally identifiable information (PII).?

Enterprises are choosing strict model training processes by indexing the data that goes into generative AI tools, and it is limited to only the data that the tool needs to generate a meaningful result. Some enterprises are choosing solutions that connect to and protect the data where it already lives to ensure better control.

Some table stakes include:

·?????End-to-end data encryption

·?????Single sign-on (SSO) and multi-factor authentication

·?????Adherence to ISO and IEC certifications

·??????Compliance with data security regulations such as GDPR and CCPA

·?????Regular security testing and mechanisms to ensure users only see what they are authorized to see.

It’s important to define the security governance process with a human-integrated approach.

6.????Cultural change management

Generative AI use cases are emerging every day, and the speed and magnitude of change are marking it a seismic event that will change the talent mix and organization structures. Hence, enterprises need to sharply focus on cultural and organizational change management as the top priority. The key here will be to communicate the adoption plans in advance, create the bridge training courses, and help employees and partners to align with the overall goal.

Time will tell the extent of the success of this technological phenomenon. However, the relationship between ChatGPT and humans seems to be much warmer and more empathetic than its predecessor. The reason could be the human-like behavior and ability to learn quickly and reproduce the right answers. This relationship sets the foundation for Generative AI to become an integral part of our lives in the near future.