Skip to content


Frequently Asked Questions About Generative AI

Generative AI ignites conversations among business leaders, politicians, advocates and critics alike. It’s created new opportunities, introduced more technical considerations and raised a ton of questions. Every business takes its own journey, but we are exploring genAI, finding use cases and developing solutions together.

Organizations recognize the need to demystify genAI and to comprehensively understand its capabilities and limitations. They’re investing time and resources to educate themselves about the intricacies of the technology, ensuring they can harness its power while mitigating potential risks. But as businesses dig deeper into generative AI, they uncover new challenges, and many are left uncertain of the best path forward.

That’s why we recently presented an interactive webinar on the topic, hosted by Jonathan LaCour, our CTO, and Dr. Ryan Ries, our Data, Analytics and Machine Learning Practice Lead. During this "ask us anything" style session, we invited our audience to submit their burning questions about genAI, and the response was both exciting and encouraging. Let’s look at some of the common themes and questions we heard.

What’s Exciting About Generative AI?

It’s exciting to see how quickly genAI has evolved from an interesting tool to a vast assortment of use cases, platforms and possibilities. Technology is constantly shifting and changing, and hype cycles often revolve around cool concepts that may not be useful. The difference now is that genAI has proven to be immediately applicable and beneficial to a wide range of individuals and businesses. Its potential for enhancing efficiency and productivity can resonate with anyone who has a variety of tasks to accomplish each day.

It’s not just the technology that’s exciting, but also the incredible opportunities it presents. GenAI can revolutionize the way we work, enabling us to achieve more in less time. People can automate repetitive tasks, generate creative content and assist in decision-making processes, which can significantly enhance efficiency and free up valuable resources for more strategic and innovative endeavors. GenAI offers endless possibilities, and it’s thrilling to witness the incredible solutions and innovations that individuals and businesses can create. 

Whether developing personalized customer experiences, automating complex processes or unlocking new realms of creativity, genAI can transform industries and drive us toward a more efficient and innovative future. However, it's crucial to approach genAI with responsibility and ethical considerations. As with any powerful technology, it's essential to ensure that it is used in a manner that aligns with ethical guidelines and respects privacy and security. 

It’s exciting to use generative AI on AWS, as the innovative company has focused on data security from the beginning. AWS aims to provide a marketplace of foundation models and a secure platform for businesses to leverage and experiment with different LLMs without a one-size-fits-all approach. The company prioritizes protecting data and ensuring customers remain safe while providing opportunities to leverage cutting-edge, high-performing technologies and services tailored to their use cases.

What Are Some Emerging Use Cases for GenAI in Industries Like Financial Services, Cybersecurity and Real Estate?

GenAI has opened up a world of possibilities across industries. While some use cases are designed to be a niche solution, many others, like intelligent document processing, can be applied broadly, regardless of the type or size of the business.  

For example, in the financial industry, one exciting use case is the automation of newsletters. With genAI, financial advisors and organizations can easily create engaging and personalized newsletters by leveraging existing articles and content. Using a retrieval augmented generation (RAG) implementation, genAI can generate new content based on recent articles, which can help save a lot of time and effort. This allows them to provide valuable information to their customers without writing from scratch.

Another use case in the financial sector is the creation of chatbots for analysts. Venture capitalists and private equity firms often track numerous companies before making investment decisions. With genAI-powered chatbots, analysts can ask questions about specific companies and gain deeper insights into their performance and market trends. This enables them to make more informed investment decisions and stay ahead in a competitive market.

Similarly, chatbots can be incredibly helpful in cybersecurity by providing information on cybersecurity risks and helping companies better understand compliance issues. Users can ask questions and receive quick and accurate responses, helping them find potential threats and take appropriate measures to protect their systems and data. GenAI can also help cybersecurity teams by analyzing vast amounts of data and providing quality support with faster and more effective solutions.

Real estate companies are using genAI to create personalized customer updates by summarizing and combining data from various sources, such as property listings and customer preferences. Real estate agents can easily generate customized messages daily or weekly. This allows real estate companies to provide concise and tailored information to their customers, enhancing the customer experience and streamlining the sales process.

These are just a few examples of‌ emerging use cases for genAI. As ‌the technology continues to evolve, we expect to see even more innovative applications that enhance efficiency, improve decision-making processes and drive growth in these sectors. 

How Do You Keep Intellectual Property Secure?

Many approaches and serious considerations exist for keeping intellectual property (IP) secure with genAI. It’s essential to approach third-party platforms cautiously. It may be tempting to invite AI bots into meetings or copy and paste data into AI tools, but it's crucial to understand the potential risks involved.

One of the safest and most effective ways to protect your data is to use a trusted platform like AWS. The company provides a secure environment for customer data, and as the biggest cloud provider in the world, it’s already home for the data of many businesses. Leveraging AWS genAI services ensures that your data remains within the walls of your cloud environment, minimizing the risk of unauthorized access. 

AWS offers tools like Bedrock and JumpStart models on SageMaker that prioritize data protection. If you’re using Bedrock and API models within it, your data never leaves the AWS ecosystem, providing an extra layer of security. Similarly, with JumpStart, you control your container, ensuring that your data stays within your environment. 

In addition to technological measures, setting good corporate policies and educating your team on the importance of IP security is crucial. Protecting IP lies more in the process rather than the tools themselves. Be mindful of what you claim as IP and ensure that your team understands the implications of interacting with external platforms. 

Should You Use a Third-Party Model, or Build, Train and Operate Your Own?

When deciding whether to use a third-party model or build your own, one of the first aspects to consider is the size and cost of building your own large language model (LLM). Creating your own LLM can be an enormous undertaking, requiring a massive corpus of training data and potentially costing millions of dollars. In most cases, it isn't the most practical or cost-effective solution.

However, the beauty of foundation models lies in their ability to be fine-tuned and trained for specific use cases. If you have a business need requiring a customized LLM, you can take advantage of the numerous existing models and fine-tune them to suit your requirements. Platforms like Amazon Bedrock and Hugging Face offer a wide range of models that you can choose from and customize for your use case.

Whether to use an existing model or build one depends on your specific use case. If you aim to deliver an LLM API and want to differentiate your model from others, making your own model might be the best option. This allows you to optimize for speed, accuracy or other specific requirements that align with your business goals. However, building your own model can be expensive, with training jobs potentially costing hundreds of thousands of dollars.

On the other hand, if your use case can be satisfied by existing models, it's typically best to try out the available models before considering training your own LLM. Fine-tuning a pre-existing model often yields similar performance to building your own, but at a lower cost. It's a good idea to evaluate the performance of different models and choose the one that best fits your needs.

Training your own LLM requires significant infrastructure and resources. Acquiring the necessary hardware, such as GPUs, can be challenging, especially when competing against hyperscalers who offer APIs and have dedicated resources for training large models. AWS, for example, provides specialized chips like Trainium and Inferentia that can be used for training models. Leveraging these resources can save you the effort and cost of building your own infrastructure.

It's worth noting that models have evolved rapidly in recent months, with the ability to accept larger prompts and use more data for generating responses. This opens up greater flexibility and the ability to leverage models with more bespoke data for specific business needs, rather than generic, over-the-top concepts.

Where Can I Learn More About Generative AI and Machine Learning?

If you want to learn more about genAI and machine learning, several resources are available to help you dive deeper into these topics. One highly recommended source is Andrew Ng's DeepLearning.AI, which offers a range of courses, many of which are free. Andrew Ng is a former Stanford professor and provides excellent educational materials to help you understand the fundamentals of AI and machine learning. You can also check out our guide on How to Use GenAI in Your Business, along with early-release chapters of the O’Reilly book, Designing Large Language Model Applications, to find more insights, use cases and examples of genAI success stories. 

For those specifically interested in exploring genAI, numerous tools are available for you to explore and play around with different models. One such tool is LLM for Python, which can be easily installed using pip. With LLM, you can download models from various sources like Hugging Face and GPT4All, even on a regular laptop.

To get a feel for what genAI can do, you can try out ChatGPT or others, which are freely available. These can give you a glimpse into the capabilities of these models. Once you're ready to move beyond toy systems and start building production-level solutions, you can explore real-world use cases and leverage machine learning more practically and effectively.

Are you ready to discuss the next steps toward leveraging genAI or another machine learning use case? Connect with us to schedule a complimentary 60-minute session with an AI/ML expert to discuss your goals.



How does generative AI impact job roles and employment in sectors heavily reliant on creative and analytical tasks, and what strategies can organizations employ to mitigate potential job displacement?

Generative AI's impact on employment is profound, particularly in sectors reliant on creative and analytical tasks. It offers tools that augment human capabilities, allowing for more innovation and efficiency. Organizations can mitigate potential job displacement by focusing on retraining and upskilling employees, enabling them to work alongside AI technologies effectively. Emphasizing the collaborative potential of human-AI interaction helps businesses harness generative AI's benefits while maintaining a skilled workforce.

How can generative AI be integrated into traditional industries such as manufacturing or agriculture to improve efficiency and innovation, and what are the potential challenges in adopting such technologies in these sectors?

Integrating generative AI into traditional sectors like manufacturing and agriculture can significantly enhance operational efficiency and drive innovation. AI can optimize production lines, predict maintenance needs, and customize product designs in manufacturing. In agriculture, it can improve crop yield predictions, pest management, and resource allocation. The challenges include the high initial investment, the need for technical expertise, data privacy concerns, and the adaptation of existing workforce skills to leverage AI technologies effectively. Overcoming these hurdles requires strategic planning, investment in training, and a focus on long-term value creation.

How can small to medium businesses (SMBs) leverage generative AI effectively without significant investment in data science resources, and what steps should they take to integrate this technology into their operations?

Implementing generative AI can seem daunting due to perceived high costs and technical complexity. However, SMBs can start by identifying business processes that benefit from automation and enhanced creativity. Utilizing AI-as-a-Service platforms can be a cost-effective way to access generative AI capabilities without heavy upfront investment, allowing SMBs to experiment with AI features and integrate them gradually into their operations, scaling as they grow more comfortable and proficient with the technology.

Author Spotlight:

Ryan Ries

Keep Up To Date With AWS News

Stay up to date with the latest AWS services, latest architecture, cloud-native solutions and more.