Generative AI

Fireside chat recap: The future of generative AI

8 min read
Tags:

In a recent fireside chat, Nylas Co-founder and CTO Christine Spang and OpenAI Applied Engineering Manager Evan Morikawa sat down to discuss the exciting prospects of Generative AI technology, exemplified by models like ChatGPT. 

A few key themes emerged from the conversation: 

  • What is generative AI, and how does generative AI work? 
  • The importance of generative AI technology and the benefits of generative AI across industries 
  • Practical applications of Generative AI – how technical leaders and software engineers are using the technology today
  • The impact of generative AI on the future of work 

Watch the video recording here, or read on as we dive deeper into these themes, further discussing generative AI’s impacts, benefits, and applications and exploring its potential to shape the future of work. 

What is generative AI, and how does it work? 

Generative AI is a category of AI that helps users generate content, including text, images, music, or even videos. Models like ChatGPT operate on deep learning techniques and extensive training data. Learning patterns and structures from the data generate coherent and contextually relevant responses. Continuous training enables these models to improve performance over time.

Prompt engineering also plays a crucial role in the effectiveness of generative AI models like ChatGPT. By providing specific prompts or instructions to the model, developers can guide its behavior and generate desired outputs. Thoughtfully crafted prompts help elicit more accurate and contextually appropriate responses, enhancing the overall performance of the generative AI model. Through iterative experimentation and refinement of prompts, developers can fine-tune the model’s outputs and improve its ability to understand and generate desired content.

In the fireside chat, Evan mentions prompt engineering is a rapidly evolving field, and its nature is constantly changing across different domains. He compares prompt engineering to the skills managers and teachers have developed over time, indicating its importance in effectively working with generative AI models. The conversation highlights the need to tailor prompts to the appropriate level of abstraction and specificity to achieve desired outputs. 

Christine also acknowledges the importance of breaking down problems into manageable tasks and providing clear instructions to the AI model. Evan and Christine emphasize the importance of improving prompt engineering skills to collaborate effectively with AI technology.

Want to see a demo of Nylas’ developer advocates experimenting with ChatGPT prompt engineering? Check out the recording here

Understanding the importance of generative AI

Generative AI technology, powered by models like ChatGPT, has rapidly gained prominence due to its ability to generate human-like text based on prompts. This scalable and powerful technology holds immense potential for various industries and applications. Using generative AI, employees can tap into its creative potential and explore new possibilities in their work. Technical leaders should encourage its use for teams as it can enhance productivity, foster innovation, and provide a competitive edge in industries where creative problem-solving and unique outputs are valued.

“Everybody needs to deeply consider and plan for the impact on their organization, both now and in the future. Recognizing that this is already significant enough to warrant attention is crucial. Everyone must acquire these tools today, as they are skills that require practice, familiarity, and mastery. If you, as an individual, do not learn these skills, you will be left behind, as the way people work in the coming years will undergo substantial changes. This could ultimately lead to being outperformed by others who have invested their time in learning these skills.” 

– Christine Spang, Co-founder, and CTO, Nylas

In the session, Evan also expresses his perspective on generative AI. He sees it as a transformative force that will impact various professions, although not uniformly or to the same extent. He highlights its potential value and active exploration in finding its most useful applications. As a software engineer, Evan shares his firsthand experience integrating models like GitHub Co-pilot, which shifted how he builds things and alters his coding workflows. 

“There is definitely a danger of these things becoming too much of a crutch. The best analogy I like is imagining it as a senior engineer, mentor, or tutor constantly by your side. However, when learning to code, relying solely on another engineer to do everything for you inhibits the necessary internalization and growth. This applies to heuristics used in schools as well. While it can be a helpful tutor, good tutors understand the importance of not writing your essays for you until you have a solid grasp of the fundamentals.”

– Evan Morikawa, Applied Engineering Manager at OpenAI 

Unveiling the benefits of generative AI

Generative AI brings a host of benefits to the table. A few examples include: 

  • Automating tasks
  • Enhancing productivity
  • Facilitating complex problem-solving 
  • Personalizing recommendations 
  • Aiding in the creation of creative content 
  • Translating various languages 
  • Generating data samples 

From a business perspective, Generative AI can help companies to streamline operations, improve customer experiences, and foster innovation. 

During the fireside chat, Christine explains that Nylas has encouraged using ChatGPT at the company by providing licenses to the engineering team and offering ChatGPT Plus subscriptions to all employees. She describes how Nylas views generative AI as a valuable learning tool that boosts efficiency and helps individuals get tasks done professionally and personally. The Nylas team has discovered its usefulness in various ways, from obtaining accurate explanations to generating Google Sheets formulas. Non-technical staff members have found it liberating, allowing them to handle tasks and take ownership of their output independently.

Evan explains, “You gotta treat it as your friend, someone relatively knowledgeable that you generally trust. Depending on the context, there’s a huge difference in application and safety. The industry actively focuses on closing the gap and enhancing the model’s ability to introspect and discern the truth. We’re just scratching the surface with these techniques, and while we need to adapt our interactions with it today, it’s likely to change frequently in the coming months and years.”

The impact of generative AI on the future of work

It’s no surprise that the future of work is being transformed by advancements in AI, particularly in the field of generative models. By incorporating AI models like ChatGPT into everyday tasks, employees can harness their capabilities for drafting emails, generating code snippets, offering customer support, and more. However, addressing concerns regarding data privacy and intellectual property protection is crucial.

Christine and Evan continue to talk about balancing innovation and security, the importance of building reliable software, and ethical considerations in AI and machine learning related to generative AI and the future of work. 

Evan emphasizes OpenAIs significant focus on security regarding AI accounts and preventing unauthorized access. He compares the security measures taken for AI models to those employed by companies like Nylas, which handle sensitive data from email, files, and customer records. 

“We have built a robust security posture, considering application security and data privacy. [OpenAI’s security team] is aware of the potential risks posed by state actors attempting to access AI model weights and is actively preparing for more intense security measures. Data privacy and ownership are paramount, and we are extremely careful about the data that goes into training models, ensuring opt-out options and addressing ownership questions. The complexity of legal and ethical considerations around data is an important part of the conversation, and our teams are actively working on it.”

– Evan Morikawa, Applied Engineering Manager at OpenAI 

The conversation concludes with Christine raising the question about the voices calling for a slowdown in the development of AI due to its potential risks. Evan acknowledges the need for caution and highlights the importance of considering incentives and motivations in AI development. He mentions OpenAI’s efforts to establish a nonprofit structure to avoid being driven solely by corporate interests. 

Christine and Evan express the significance of finding a good framework for regulation and governance to manage the rapid pace of technological advancements without hindering innovation. They emphasize the need to address AI’s potential benefits and downsides and ensure that adverse outcomes are not the predominant result.

Conclusion

As exemplified by models like ChatGPT, generative AI technology is poised to shape the future across industries. Its automation, problem-solving, and creativity benefits hold immense potential for improving efficiency and driving innovation. As we embrace the potential of generative AI, addressing concerns around privacy, intellectual property, and responsible use is crucial to ensure this technology’s positive and ethical integration into our lives.

Watch the full recording to learn more about embracing this groundbreaking technology to unlock innovation and stay ahead in the ever-evolving digital landscape. To learn more about how developers can build using ChatGPT with Nylas today, check out our recent blog post.

Related resources

How to build with Nylas using ChatGPT today

Learn how developers can build using ChatGPT with Nylas to maximize efficiency and productivity.

Voice-activated emailing: A guide to using ChatGPT and speech recognition to send emails

Have you ever thought about using your voice and ChatGPT to send your emails? We will create a DearPyGUI desktop app to do exactly that.

How to create a sentiment analysis dashboard using Python, Streamlit, and ChatGPT

If you want to know how your customers feel about your product, create an Streamlit dashboard using ChatGPT and Sentiment Analysis.