Generative AI for the Management Team

Questions to Kickstart the Discussion

Generative AI, currently the fastest-growing area within machine learning, has gained immense popularity among the general public as the technology can be used to address everything from simple to complex everyday problems. By now, you've probably heard of, and perhaps even tried, some of the popular tools like Midjourney, DALL-E, ChatGPT, or Bard, all of which are powered by generative AI. A few years ago, none of these tools had been launched - today, they're a hot topic in politics, classrooms, and the business world.

McKinsey's annual global AI survey indicates a dramatic surge in the number of generative AI tools available on the market. As more tools are launched, the amount of companies using generative AI increases. Today, about a third of the participants in McKinsey's survey report that their business regularly use generative AI in at least one corporate function.

After recognizing the potential of the technology, decision-makers across various industries are now considering how generative models can drive business value and competitive advantages - a challenge made even more daunting given the rapid pace of advancement.

In this guide, we provide an introduction to generative AI, including use cases and questions that you, as a CEO, can pose to the responsible CXO to kickstart a discussion on generative AI.

‍

What is Generative AI?

Generative AI is a type of AI used to create new content such as sound, images, text, code, and videos. Generative AI learns from the patterns and structures in its training data and can then produce new content based on that. This means that a generative model, for instance, can create realistic images of cats, even though it has never seen a real cat, simply by learning what a cat looks like from thousands of existing images.

‍

‍

The applications for generative AI are vast. For instance, this type of AI can draft a business plan or management report, summarize a research paper, create images for social media posts, or generate synthetic data to train AI models.

The strength of generative AI lies in its flexibility. The foundational models that are developed in the process can be applied across various domains without the need for further modifications or training. This means that a company can leverage a single model across multiple application areas, only needing to adjust the application layer. In contrast, traditional AI is often tailor-made for a specific task and requires vast amounts of underlying data to be applicable. With generative AI, it has never been easier to deploy and scale powerful AI applications.

‍

Generative AI & Natural Language Processing

Generative AI often involves language, either through the user providing the model with text-based instructions or the model producing text as output. For the large generative language models, known as Large Language Models (LLM), to utilize language, they need to "understand" it. This is where Natural Language Processing (NLP) comes into play.

NLP encompasses techniques that enable computers to understand and generate natural language in the same way humans do. Natural language processing is a subfield within linguistics, computer science, and artificial intelligence. There are various applications within this domain, such as classifying text, comparing the content of texts, extracting keywords from text, summarizing large volumes of text, translating text, or generating text.

A well-known example of a generative model (LLM) is ChatGPT, which produces human-like text responses based on a user's text input. The core technology behind most large language models is the "transformer model," introduced 2017 by researchers at Google Brain. The majority of today's language models are built on this particular model architecture. GPT stands for "Generative Pre-trained Transformer".

Other examples of AI tools built on generative models include the image generators DALL-E 2 and Midjourney, both launched in 2022. Google's chatbot Bard, a competitor to ChatGPT, the productivity tool Microsoft 365 Copilot, Photoshop's generative tool Firefly, and the development tool GitHub Copilot were all introduced in 2023. These tools have gained significant media attention over the past year, and due to their accessibility, generative AI is now used by a wide audience in ways that were unimaginable just a few years ago.

‍

Generative AI for various business functions

In this section, we provide examples of application areas and questions you can ask your CXO to start the discussion about generative AI.

‍

1. Finance & Accounting

Generative AI can be used in finance and accounting for tasks such as generating reports or automating administrative workflows. The ability to extract and understand text from invoices, orders, and contracts will lead to significant efficiency gains in the finance function. Additionally, generative AI can be used to create simulations and scenarios, which will likely revolutionize strategy and budgeting work within a few years.

Questions to discuss and delegate:

• Are you currently exploring how generative models can automate parts of the reporting process, such as drafting quarterly reports?
• Most companies now have incorporated automated invoice interpretation into their processes. Can these be further streamlined with smarter interpretation and processing using the latest language models?

‍

2. Procurement & Logistics

Within procurement and logistics, traditional AI can optimize delivery routes and improve demand forecasts - read more about it here.

As procurement contains lots of communication, generative AI has the potential to accelerate and streamline communication processes with suppliers and partners. Language- and document-intensive quality processes also have great potential for improvement. When language models are combined with the ability to perform external web searches, such as in Google Bard, a powerful tool is created for faster and better research, such as finding new suppliers or understanding market pricing.

Questions to discuss and delegate:

• Which part of your supply chain has the greatest potential to be improved with generative AI?
• Can automated communication improve the relationship and support towards suppliers and partners?
• How automated are your procurement processes? Can purchase orders, for example, be interpreted and processed better with generative AI?

‍

‍

3. Human Resources

AI has already made its way into the HR field. For example, traditional AI is used for matching CVs to job postings, which can streamline the recruitment process. The subfield, generative AI, also has great potential in HR operations. For example, it can be used to create or customize job postings or to personalize and streamline the recruitment and onboarding process.

Questions to discuss and delegate:

• Have you considered how generative AI can streamline the recruitment process, such as by creating drafts for job postings or summarizing incoming applications and CVs?
• Do you see opportunities to use generative AI to create individualized onboarding and offboarding processes, or training programs?
• Can generative AI help you predict future skill needs within the company?

‍

4. Legal & Compliance

Legal and compliance is often a document-intensive business function. Generative AI can be used to find information in a large document structure, summarize large amounts of text, and to create drafts or review legal documents.

Questions to delegate and discuss:

• Which legal documents could you create drafts for with the help of generative models?
• Could you use an AI agent to help you find the right information in a large document structure?
• Can language models help you see when contracts expire, when key employees leave, or when new legislation affects you?

‍

5. IT

Generative AI has also entered the IT department, especially in software development. Developers can now get help from AI with generating, debugging, and testing code. The productivity gains here are significant, especially in a labor market with a long-standing shortage of developers.

Questions to discuss and delegate:

• Could generative models help you write, test, and debug code?
• Do you use GitHub Copilot today?
• Do you use ChatGPT? If so, have you evaluated the risks and ensured that your data is handled securely?
• How can generative AI help you speed up development, for example through AI-driven rapid prototyping?
• Can the IT department create more effective and accessible support with the help of language models?

It is important to note that there are security risks that IT needs to address. For example, Samsung recently leaked their own corporate secrets as a result of employees using ChatGPT.

‍

6. Product development

With its potential to boost creativity and streamline idea and concept development, generative AI will become an increasingly essential part of the entire product development process.

Questions to discuss and delegate:

• Are there parts of your product development process that are particularly resource-intensive or time-consuming?
• Are personalized and customized products a competitive advantage in your industry?
• Can you use generative AI to generate ideas and concepts, develop rapid prototypes, or something else?
• Can generative AI be used to simulate the potential benefits and drawbacks of different product versions?

‍

‍

7. Manufacturing

Modern manufacturing is often one of a company's most data-intensive environments, creating good conditions for AI. Traditional AI can be used to predict maintenance needs, optimize workflows, and improve quality control. Generative models have proven to be very powerful for creating scenarios and simulating data and events, which have the potential to be a powerful tool in manufacturing.

Questions to discuss and delegate:

• Have you considered how generative AI can be used to improve efficiency by simulating and optimizing production flows?
• Can quality work be improved by simulating tests, troubleshooting, and inspections?
• Predictive maintenance (PDM) is a very interesting area where generative AI combined with AI-driven forecasting can simulate and predict maintenance and repair needs - could this be relevant for you?

‍

8. Marketing

Marketing is one of the most obvious areas for generative AI right now. Using AI tools in marketing not only frees up time and resources for more value-creating activities, but can also promote creativity, idea generation, and personalization. As a marketer, using the existing generative AI tools in your daily work is mandatory.

Questions to discuss and delegate:

• How can generative AI and tools like ChatGPT, Midjourney, BARD, and DALL-E 2 help you reduce the costs of creating content for campaigns, websites, and social media?
• How can you use generative AI to create personalized campaigns?
• How can the marketing team use ChatGPT and BARD to do better research and thus develop and reach out with more accurate messages?
• How does access to generative AI affect your need for external partners? Do you need the same agencies and freelancers as you did a year ago? Are your existing partners better than you at generative AI? (They should be!) There is likely a lot of money to be saved here.

9. Sales

Generative AI can help a sales department in many ways, such as creating personalized sales strategies or product recommendations based on existing customer data.

Questions to discuss and delegate:

• How can generative AI improve your understanding of what your customers want, and what arguments/pitches you should use?
• Does your pre-sales and sales team use tools like ChatGPT and Bard to research customers and your competitors?
• Do you use generative AI to quickly draft and analyze contracts?
• Could generative AI assist salespeople in gathering real-time information during customer meetings or in drafting sales materials?
• How can language models be used to send better and more relevant emails and email sequences?
• Can generative AI help you onboard new sales representatives faster?

‍

‍

10. Customer Service

It's becoming increasingly popular for companies to train GPT models on their own internal data to develop AI-powered support agents. These agents can either directly interact with customers or serve as a support tool for human customer service representatives. We strongly recommend that all executive teams closely examine this area, as advancements in generative AI will transform the support function in the near future.

Questions to discuss and delegate:

• Can AI-powered support agents trained on your data and documents directly address customers, thus reducing response times?
• Can generative models assist human customer service representatives, for instance, by summarizing support cases?
• Have you structured and categorized the data in your case management system so that it's prepared for AI applications?
• What does your vendor's AI roadmap look like? Can they offer advanced tools based on generative AI, or will you need to develop your own solutions?
• How can the onboarding process for new customer service representatives be improved?
• Can access to internal wikis and Q&A be streamlined and accelerated with language tools?

An inspiring example is the American fast-food company Wendy's, which has integrated generative AI into its drive-through. Learn more about it here.

‍

A strategic opportunity for the management team

This was part three of our AI series for CEOs. Part 1 - "Challenge Your Management Team with the Right Questions about AI" can be found here. Part 2 - "AI is Crucial for Management - Here's How to Get Started" is available here.

‍

‍

‍

Is your company ready to embrace the AI revolution? Reach out to me to start your AI journey today: mattias.guilotte@violet.ai

‍