Findings from the Go developer survey

Data from the Go Developer Survey, conducted by Google’s Go team, showed a strong preference among developers using Go for AI application production. Gathering responses from 6,224 participants, the survey’s findings, published on April 9, 2024, provides insightful perspectives into the developers’ preferences and sentiments. 

93% of developers expressed satisfaction with Go over the past year, highlighting Go’s alignment with developers’ needs and expectations. To add to this, 80% of respondents stated their trust in the Go team’s decisions, signaling confidence in its future direction and development.

Language preferences and migration

Developers stated a preference for Go for developing AI-powered applications and services. However, Python retains its status as the initial go-to language for initiating AI projects, due to its entrenched position in AI development. 

Despite this, a large number of developers either currently use Go for their AI projects or show interest in transitioning to Go, thanks to the language’s growing relevance in AI workload management.

AI-Powered service development

Developers are actively engaged in creating a variety of AI-powered services, with summarization tools, text generation tools, and chatbots being the most commonly developed applications. Go now shows more diversity in facilitating AI-driven solutions, catering to a growing range of functionalities and user needs.

AI model usage and integration

The survey highlights the AI models that have gained popularity among developers’, with 81% of respondents using OpenAI’s ChatGPT or DALL-E. Meta Llama follows with 28%, while Mistral AI/Mixtral and Google’s AI models like Gemini, Imagen, or PaLM have 18% and 13% usage rates, respectively.

AI Integration libraries and services

When it comes to integrating AI models, 69% of developers rely on OpenAI. Hugging Face TGI or Candle is used by 22%, and LangChain by 20%, indicating that developers leverage diverse tools and services to properly exploit AI capabilities. This diversity highlights the dynamic nature of AI development, wherein multiple solutions coexist to meet different requirements and preferences.

Go learning and community engagement

Time constraints and limited opportunities are primary obstacles that keep Go developers from achieving their learning objectives, according to the survey. Despite this, a significant portion of the community remains engaged. Approximately 33% of the survey respondents reported active participation in the Go developer community, either online or in person, over the past year – marking the community as a key tool for knowledge sharing, networking, and fostering innovation.

Development environments usage

A majority, 61%, prefer Linux as their operating system for writing Go code, with macOS closely following at 58%. Windows is used by just 23% of the respondents, and the Windows Subsystem for Linux is used by only 17%. 

Looking into editor usage, Visual Studio Code was the most popular editor, used by 43% of developers, followed by GoLand/IntelliJ at 33% and Vim/Neovim at 17%. Businesses can use these preference insights to optimize their tech stacks and support for Go development teams.

Security and compliance in Go development

For Go developers, insecure coding practices is the primary security issue, with 42% of respondents highlighting this concern. Organizations need to invest in secure coding training and implement robust code review processes to mitigate security risks, aiming to improve the security posture of Go-based applications and align them with the broader organizational data integrity and trust objectives.

Tim Boesen

April 17, 2024

3 Min