Welcome to “Generative AI for Business: Insights from Eugina Jordan.” In this enlightening episode, we dive deep into the transformative potential of generative AI across various industries. Hosted by Hema Kadia, this discussion with Eugina Jordan explores the foundational aspects of AI and highlights key applications transforming sectors like marketing, customer service, and finance. 

From creating compelling content to enhancing customer interactions and automating financial reports, generative AI is a game-changer. Eugina also addresses the critical challenges businesses face in integrating AI and offers insights into the future trends shaping the AI-driven business landscape. Join us as we uncover how generative AI is an innovation and a strategic tool for driving efficiency and creativity in today’s fast-paced world.

Key Generative AI Applications in Various Industries

Hema Kadia: Thank you for joining me, Eugina, to discuss generative AI for business. Let’s dive into the first question. What are the top use cases for Generative AI across different industries?

Foundations of AI and Generative AI

Eugina Jordan: We can first identify what AI is and types of AI, and then we’ll dive into generative AI. For people who don’t know, AI has been around since the 1950s. It’s the intelligence that is created and performed by machines, by computers. And computers were created in the 50s. There are humans like you and I. That’s our intelligence. Machine intelligence is by computers. Once people know it, explaining what machines can do is simple.

Machines, with their intelligence or artificial intelligence, can predict things. That’s predictive AI, and it’s been around in many industries for a while. In addition, there is also communications AI, and that’s your chatbot. It’s developed in many industries. They’re moving from simple chatbots to generative AI and predictive AI chatbots. There’s also reinforcement learning AI. It’s been around in gaming industries and robotics forever. In robotics, it’s being combined with predictive and generative AI. Natural language processing (NLP) also feeds into multi-modes with GenAI.  All those different types of AI are feeding GenAI to produce something. That’s why it’s called generative AI. It produces images, maps, conversations, and decisions via machine.

Top Business Use Cases of Generative AI

To understand the types of AI in GenAI, let’s return to your question about the use cases. We will concentrate on business use cases and separate them by function and industry. The most prominent use cases are marketing, customer experience, and finance.

Generative AI Transforming Marketing: The most prominent use cases by function are number one, for marketing, with content creation and image creation, which maximizes time for cash-strapped or time-strapped marketers. It can create really good content. Humans fit in as editors. They provide really good prompts for AI to create blogs, white papers, and social media posts, and then they edit for tone, clarity, and so on. This is a very prominent use case right now.

Enhancing Customer Service with Generative AI: For customer service, as we discussed, there are AI chatbots that can predict and generate responses. It needs to be monitored by humans because AI hallucinates. We want to avoid issues where AI-generated conversations with chatbots hallucinate, causing companies, airlines, and financial institutions to end up paying millions of dollars for those hallucinations. The second one is chatbots, which improve the customer experience.

Financial Innovations with Generative AI: The last one by function is finance. The finance function can benefit from automating reports, forecasts, and risk management. 

Overcoming Engineering Challenges in Generative AI: I’m not covering engineering because it is challenged with code accuracy and requires a lot of human oversight. Hugging Face is trying to fix those issues. In a year from now, engineers will be able to use artificial intelligence to create their code.

Expanding Business Applications of Generative AI: If we look at other use cases besides functional ones, there are overall business use cases. Companies like Gemini, ChatGPT, and pilots can optimize different aspects of business productivity in our work. Unfortunately, it’s not properly implemented, creating challenges because it must be trained on its data and regulations. You cannot just give the tool to your team and say, “Run with it.” You need to train the tool properly.

These are the most popular use cases that can be applied across different industries, from healthcare to finance to telecom, to improve their marketing, customer experience, and business.

Generative AI Synergies in Telecom

Hema Kadia: If you look at the synergies in the telco industry, marketing, and customer experience are common areas. As you explained, the network operations team has been using predictive analytics, which continues in network operations. We have yet to see prominent use cases for generative AI in telco on the operations side.

Sector-Specific Generative AI Development 

Hema Kadia: What is your view on having industry-specific generative AI, and do you see some sectors excelling better?

Eugina Jordan: Great question. I’ll answer it because, between us, we have over 50 years of experience in the telecom industry. In healthcare and telecom, the word “cell” has different meanings. In healthcare, pharma, or anything related to medicine, the word “cell” means something different than what we classify as a cell in telecom, which is a cell site. If a generic LLM (large language model), and there are plenty of open and closed LLMs and also SLMs (small language models) out there, is untrained, it will generate generic responses. That is a challenge. The strategy would be taking an open LLM—there are plenty on the market—training it with your data, upskilling your people on how to input the data into that generic LLM, and figuring out the process to train and implement. As we discussed, you cannot use a generic LLM even for marketing. Generic LLMs perform well in marketing because they’ve been trained, and there aren’t many different types of marketing. So my answer is yes. If there is specific data, if you use specific language or have lingo in your industry, you need a specific LLM. 

Collaborative Efforts in Generative AI Across Industries

Hema Kadia: We discussed the global initiative where telecom operators formed the Global Telco AI Alliance. Do you see similar initiatives in other industries?

Eugina Jordan: We discussed the Global Telco AI Alliance announcing they’re building LLMs in multiple languages. We laughed because, while I love my telcos, we don’t like to share. This alliance addresses that issue by combining data from US, European, and Asian operators to create a robust LLM. Surprisingly, there aren’t many specific industry alliances like this. Telco is unique. In other industries like healthcare or automotive, they have AI initiatives within their organizations. For example, I work for TIP, and we have a Telco AI group because we have access to the ecosystem. To answer your question, I’m seeing two main trends in the industry regarding GenAI alliances and organizations:

• Formation of Alliances: Companies are forming alliances, such as OpenAI, partnering with various small and large companies, including hardware and software firms.
• Regulations: Governments are working on regulations. Different federal and state departments in the US are addressing it with various bills. The European Union is also tackling concerns around ethical and responsible AI.

Ethical AI and Global Challenges: A key regulation is that companies must specify if AI generates content. This transparency is crucial because these models were often trained on unauthorized content. Recently, a summit in Seoul brought together multiple governments to discuss a unified approach to AI. However, geopolitics is a challenge, as China is excluded from these discussions, leading to isolation. In China, LLMs are driving down costs, while elsewhere, the competition focuses on innovation, not cost.

Future of AI Alliances: Time will reveal which alliances will thrive. The more the private and public sectors collaborate on developing regulations, the better the future of AI will be.

Hema Kadia: To your point, I agree about the Telco Alliance, but there are also many partnerships beyond the alliance. Recently, we saw South Korea Telecom (SKT) investing heavily in a startup called Perplexity AI. There are numerous partnerships where operators invest in generative AI, both on the device and the pure telco sides.

Challenges in Integrating Generative AI

Hema: What challenges do businesses face when integrating generative AI into their operations?

Eugina Jordan: There are four challenges. These challenges have been discussed a lot, and we need to start moving from talking about them to solving them. This may be where the Telco Alliance will come in.

Data Management Challenges: The first challenge is data. In many organizations, the data is not easily accessible. It’s fragmented, siloed, and not quality data. It still needs to be cleaned up, and there are no policies around security or privacy on how that data will be managed. For example, there are many laws that prohibit data from being moved from one geographical location to another. This is understood with any other type of service, and that’s why many hyperscalers have local data centers; the data cannot move. It’s about where you’ll get the data and how to secure and implement it. How are you going to host it? If a failover is necessary when one of your cloud locations is down, where will it failover? If it’s in a different country, what are the implications? Ensuring data movement does not violate any regulations is crucial.

Team and Skills Development Challenges: The second challenge is the team. We all hear clickbaits. “AI is going to take jobs.” AI is not going to take jobs. AI will make people like you and me more efficient and productive and give us visibility into how our business is doing. It’s going to make us better professionals. But it will take time, and people need a growth mindset. Organizations need to empower individuals to get up-skilled on GenAI. It’s about having the right people, learning platforms, and processes to train those people. I always encourage people to go out there and learn. There are so many free resources. If you are worried about your job, spend 30 minutes watching a YouTube or Google training video instead of doing something else. That’s how I learn. I’m a 53-year-old woman who wants to stay relevant. Be like me.

Identifying Objectives and ROI Challenges: The third challenge I put my executive hat on is identifying objectives and ROI for your GenAI implementation. Before you ask your team, once you’ve up-skilled them and know where your data is from, or in parallel, identify the objectives. Identify the return on your investment. Don’t just run and say, “Oh my God, we need to implement GenAI.” Sit down and stop clicking on the clickbait about being behind. Only 20% of businesses worldwide are in the first phase of implementing GenAI. Identify objectives and return on your investment. What areas will it impact? Will it reduce churn in customer experience by a certain percentage? Will it improve the productivity and quality of code in engineering? Over 50% of businesses now test and build models to jump on the bandwagon, and they’ll have to redo that work. Organizations that step back and identify objectives and ROI will win in this long race.

Responsible and Ethical AI Challenges: The fourth challenge is responsible and ethical AI. We saw the news that one of the co-founders of OpenAI, Ilya, started a new startup today, the SSI (Safe Super Intelligence). Responsible AI is becoming a significant topic, not just a buzzword. As business leaders, we need to continue thinking about how we will protect the data. Are we using LLMs developed ethically? How will we implement regulations, if there are any? How will we protect our users? This should be a consideration across all areas. Is the data ethically sourced and responsibly used? Are we teaching our team to use AI responsibly and ethically? Are we developing tools that are not biased? As executives, do we include ethical implementation as part of our objectives?

Effective Strategies for Implementing Generative AI 

Hema Kadia Eugina, you mentioned these four challenges. Have you seen some best practices or effective strategies for how different businesses address these challenges?

Eugina Jordan: Unfortunately, not yet. There’s probably one business that comes to mind—McDonald’s. They recently decided to stop using AI-generated chatbots for their kiosks due to errors and mistakes. Companies decide not to proceed because they’ve learned from their pilots and are now looking to proceed with something else. So, we’re still early on.

When I’m asked that question, I give a timeline. OpenAI was formed in 2016. This was the first significant milestone when Elon Musk bought a server from NVIDIA, gave it to the OpenAI founders, and said, “Run with it.” They launched in November of 2022. Just two years ago; there are not many examples right now. I see many launches and different tools being introduced, and we just need to continue observing and learning from these developments.

Future Trends of Generative AI in Business

Hema Kadia Lastly, Eugina, how do you see the future of generative AI evolving in the business landscape?

Eugina Jordan: That is a great question because I always think about where I want to be in five to six years. Unfortunately, as a society and business leaders, we don’t yet have a clear vision for AI’s future because we are still learning. While I can’t predict the future, I can share what I hope will happen in the next two to four years.

Improved Customer Experience: We’ll get a better customer experience. The chatbots designed with generative AI for many industries—healthcare, finance, travel, retail, you name it—will provide a great experience. If you go on any website and there’s a chatbot, and it’s a generative AI chatbot, I hope it’s going to provide a great experience. This is where RAG (Retrieval-Augmented Generation) will come in, ensuring the chatbot doesn’t hallucinate and maybe even show a bit of empathy. I talk to the GPTs I create as if they were human, and they often respond kindly. In one Asian country, I believe it’s Korea, a company developed technology using gen AI to reduce yelling at customer representatives, helping them enjoy their jobs more. I hope this is the direction gen AI takes for customer chatbots, where we treat each other more humanely.

Increased Productivity: Number two, we will be more productive. This productivity will give us time back to spend with our families. By being productive—creating content or plans, for example—I’ve created many different plans in chat GPT with the right prompting, which is fantastic. It will help me launch what I want to launch much faster. Ideally, this productivity won’t keep me at my desk for 14 hours but will allow me to enjoy life outside of work.

Advanced Analysis and Strategic Thinking: Third, we thrive on analysis as analytical people. In the future, instead of using our brains for low-level analysis, gen AI will help us with initial analysis, allowing us to focus on high-level analysis for different functions. I hope we can optimize product design and development, design different prototypes with gen AI, and fast-track innovation.

Sector-Specific Improvements: For sector-specific improvements, gen AI can significantly impact healthcare. Currently, healthcare is disjointed, but gen AI will help people have personalized health and prescription plans that they can afford. This is a future hope. I hope gen AI will accelerate drug development in pharma, creating medications that alleviate suffering. Some research is already underway in this area. If AI can help with that, maybe not in five years but perhaps in ten, it would be fantastic.

We already discussed finance, but another industry is retail. We all buy things online and in stores. We can make that experience more enjoyable and fun.

Hema Kadia: I fully agree with you, Eugina. Generative AI is not only an automation tool that makes processes and systems more efficient, but it also improves our daily lives. I see a bright future, from customer experiences to marketing and daily interactions for enterprises and consumers. Generative AI offers daily benefits for both businesses and individuals.

Embracing the Transformative Power and Challenges of Generative AI

In conclusion, “Generative AI for Business: Insights from Eugina Jordan” provides a comprehensive look at how generative AI transforms marketing, customer service, and finance industries. Eugina Jordan highlights the foundational elements of AI, its evolution, and the specific use cases where generative AI is making a significant impact. From automating content creation to enhancing customer interactions with advanced chatbots, the potential of AI to drive productivity and innovation is clear. However, integrating AI comes with challenges, including data management, skill development, and ethical considerations. Eugina emphasizes the importance of strategic implementation and responsible AI practices. The future of generative AI promises improved customer experiences, increased productivity, and advanced strategic thinking, positioning it as a pivotal tool in the business landscape. Thank you for joining us in this insightful discussion, and we hope to continue exploring the dynamic world of AI in future episodes.

To stay updated on the latest insights into “Generative AI for Business,” subscribe to Eugina’s weekly newsletter here. For a more focused look at Generative AI in the Telco-Tech sector, subscribe to our TeckNexus newsletter here.