Generative AI seems to have taken the world by storm, introducing new productivity tools and even passing medical and bar exams. While the tech industry seems to still work out what this will mean for the future, we are glimpsing at the people at the heart of the revolution, developers and Developer Relations practitioners.
The data from the recent State of the Developer Nation report suggests that 63% of developers are engaged in some aspect of AI-assisted development, making it evident that this technology is rapidly maturing and transforming from a mere trend to a valuable tool.
Which developers are working on Generative AI projects?
We know developers are highly interested in AI technologies, with 57% of developers globally already working on or learning about Generative AI. But who are these developers?
Let’s take a closer look! Recently we hosted a session where we looked at what developers have to say about working on Generative AI projects. You can re-watch this session on demand and discover answers to these questions
How are developers involved in generative AI projects?
How are they engaging with generative AI projects?
Where are these developers located?
What is their level of experience?
Where did they learn how to code?
What are their roles?
How do developers involved in generative AI projects, get information about software development, and what types of content do they prefer?
How are Developer Relations adopting AI?
While developers are at the frontiers of emerging technologies, what about those of us engaging with developers daily and catering to their needs?
Recently we asked our DevRelX Community members about the topics they are most interested in, and they, just like the developers themselves, are most captivated by the current state of Generative AI.
We also asked about our members’ views on how AI technologies are affecting the Developer Relations field and developer communities overall. You can also join the discussion with them and share your thoughts and insights on this Slack thread.
Here are some of the key highlights and resources shared by our members.
⭐ Ash Ryan Arnwine, Director of Developer Relations at Nylas
At Nylas, our DevRel team is using a framework to help us explore where AI can be useful in key areas of our work. My hope is that our team can grapple with emerging AI capabilities while keeping a couple of guiding principles in mind.
First, the current wave of generative AI is the newest "bicycle for the mind"; it doesn't need to solve every problem perfectly to be a valuable tool. It's on us to learn this tool and wield it constructively in the service of our developers.
The second principle is related: we experiment aggressively in private; we ship to developers with care and consideration. We're using AI to aid in completing the meta task of inspiring developers on ways to enhance their Nylas integrations with AI. On the developer experience side, it feels like the possibilities are almost infinite! And the more we play, the better informed our product ideas and AI-enhanced DevRel work can become.
Learn more about Ash’s team’s experiments with AI.
⭐ Karl Hughes, CEO at Draft.dev
We've been starting to explore/use AI a bit mostly just for outlines and briefs for articles. The content it writes still needs a lot of hand-holding from knowledgeable engineers to ensure it's accurate. The inability to get very deep, especially in esoteric or nuanced topics still holds it back though.
Learn more about Karl’s thoughts about the future of AI writing tools.
⭐ Kamran Ayub, Principal Consultant at Lovely DevEd
AI isn't going to go away and will keep getting more sophisticated unless you believe competition will stop and capital will stop flowing to these startups and services.
Hallucinations are a problem but that will likely be handled soon, we are already beginning to see services like Azure OpenAI offer ways to safely and securely deploy AI solutions that avoid hallucination, sandbox confidential data, and integrate across enterprise tools.
Generative AI is "creative" but not as creative as humans -- it cannot generate original thought, it cannot think for you, which means it will be even more important for companies to create quality creative content since quantity will be easy. In other words, start thinking of gourmet content instead of fast food content creation. Someone I know uses the phrase, "AI-Generated, Human Curated" -- Human curation, human creativity, will be a differentiator in the sea of average-to-below-average bot content.
AI will help assist folks with disabilities. If they cannot speak, you can now use AI voices to make technical videos that aren't terrible using Text-to-Speech. If you have trouble learning, you can now ask questions about code or what you're seeing with GitHub Copilot. If you have trouble articulating, AI can help you be clearer. If you don't speak the language, AI can help you translate accurately. AI-assisted accessibility will provide more ways for people to contribute and more ways to make your content more accessible.
AI can help with dev tools in tons of ways. It can speed up rich content creation: visual essays, videos, music, etc. all of which could be hard to do before and took a lot of time will get easier and easier to create -- it'll feel more like "directing" than creating. We're already seeing some work being done with support bots or "Ask Docs" functionality. GitHub Copilot is evolving rapidly and it's already saving me time and solving real issues (see my blog post). This will change the way developers interact with docs and education.
AI can help assist with workflows. Imagine being able to summarize community posts, Discord channels, and Slack channels -- I can almost guarantee we'll begin seeing Community Copilots that help coordinate between different channels and outlets.
Microsoft is already doing a land grab for enterprise AI with their new services and offerings. It's their fastest-growing service. They have all the MLOps and DataOps infrastructure and they're creating a Copilot Platform across Microsoft 365, GitHub, Windows, ChatGPT, Bing, etc. -- that is going to be the new app frontier, accessible to billions of people. How will your dev tool help people in an everyday context outside your product UI? That'll be an interesting question to wrestle with.
AI impacts data privacy and regulations. We'll start seeing crazy data breaches, jailbreaking, and other security-related issues.
Learn more about Kamran’s experiments with AI tools.
⭐ Christie H. Kristensen, VP, Global DevRel at Mastercard
We recently published a blog about how the Open Banking API team uses AI.
Mastercard has a suite of Open Banking (US) APIs that are powered by AI and Machine Learning. Although AI is beginning to make headlines, we have been using AI for the better part of a decade and AI is embedded into a whole range of Mastercard’s products. Mastercard is committed to the responsible use of AI in our products and will remain dedicated to customer data privacy.
Stay tuned for the next post which will cover the challenges with AI!
⭐ Eric Ciarla, Co-FounderCo-Founder Mendable by SideGuide
Here are a few thoughts I have to share:
Improved Efficiency and Accessibility: Generative AI products like Mendable.ai drastically streamline developer relations by making it easy for developers to access and navigate the knowledge base. They can find answers quickly, without needing human intervention, which speeds up their workflows and improves overall satisfaction.
Greater Engagement and User Retention: With an AI-powered search and chat interface, developers can interact with the knowledge base in a more intuitive and engaging manner. This engagement can increase retention, as developers appreciate the readily available assistance and feel more connected to our product and brand.
Scalable Support System: Our AI-powered solution can handle a vast number of queries simultaneously, without the need for scaling human resources. This allows us to provide consistent, reliable support to an ever-growing user base and ensures that every developer, regardless of their timezone or location, can get the help they need when they need it.
⭐ Brett Bush, Senior Software Engineer at LogicGate
I’ve got a few notes too!
APIs are centre stage: AI services and their APIs are available off-the-shelf, allowing for straightforward integrations to be built with existing platforms, as opposed to building AI functionality internally. However, in order to integrate, the value of APIs for non-AI platforms has also risen, as these APIs will benefit from being built out and streamlined in order to support enhanced functionality by integrating with AI APIs.
Token-based API pricing: OpenAI’s API model is placing the token-based API pricing model in the spotlight, perhaps paving the way for more adoption of pay-as-you-go API pricing models as opposed to subscription or tiered models.
OpenAPI x OpenAI opportunities: Given the text-based and semantic nature of OpenAPI Specifications, they provide a compatible format for describing an API to LLMs, allowing for possibilities ranging from code snippet creation, debugging, and helping with building complex solutions with a given API.
Developer Portals: With the wiki and webpage formats of many Developer Portals, there is an opportunity to supplement Developer Portal documentation with AI-backed chatbots. This would allow developers to query specific questions from a single prompt, saving them time from parsing through pages of documentation or contacting a developer relations representative.
And what are your views on how AI technologies are affecting the Developer Relations field and developer communities? Share your thoughts with us on Slack!
コメント