Online knowledge communities combine enormous contribution volumes with uneven participation, concentrated project activity, and persistent trust friction. GitHub’s 2024 and 2025 platform measurements show the scale of collaborative production, while Stack Overflow surveys show how developers search, answer, learn, and judge machine-generated help.
Contents
- The scale of online knowledge contributions
- New contributors and the global participation pipeline
- Public, private, and open-source contribution patterns
- How contributors concentrate across leading projects
- Trust, usefulness, and friction
- Time spent sharing, finding, and learning
The scale of online knowledge contributions
GitHub’s activity measure is broader than code commits alone. Its definition includes comments, created gists, issues, pull requests, discussions, pushed commits, and reviewed pull requests. That makes the figures useful for understanding contribution as a broad form of online knowledge exchange.
During GitHub’s measurement period from October 1, 2023, through September 30, 2024, the platform reported 518 million total projects, with project volume up 25% year over year. Developers made 5.6 billion contributions across all GitHub projects in that period. Nearly 1 billion contributions went to public and open-source projects. These figures come from GitHub’s Octoverse 2024 report.
The same report counted 137,000 public generative-AI projects, a 98% year-over-year increase. Contributions to generative-AI projects rose 59%, and total contributions to all generative-AI projects increased by almost 60%. Developers also created more than 70,000 new public and open-source generative-AI projects during the period. The growth indicates a rapidly expanding area of shared technical knowledge, although the measures describe activity and do not establish why it increased.
GitHub’s 2025 Octoverse overview reported more than 180 million developers building on the platform and 630 million total projects. It also reported 1.12 billion total contributions across the platform for the 2025 measurement year. The large difference between the 2024 and 2025 contribution figures reflects the supplied annual overviews’ reported measures and definitions; they should be read as period-specific platform statistics rather than as a single continuous series. Source: Octoverse 2025.
New contributors and the global participation pipeline
New participation is a central feature of online knowledge communities. GitHub reported that more than 36 million developers joined in its 2025 measurement year, a 23% year-over-year increase. The growth averaged more than one new developer joining every second. Developers created more than 230 new repositories every minute, producing 121 million more repositories than in the prior year.
The 2025 overview also reported 395 million public and open-source repositories, 72 million more than the prior year. This gives public collaboration a substantial visible footprint even though the largest share of contributions was associated with private repositories.
The reported regional flow of new developers was uneven:
| Region | Approximate new developers per minute, 2025 |
|---|---|
| Asia-Pacific | 25 |
| Europe | 12 |
| Africa and the Middle East | 6.5 |
| Latin America and the Caribbean | 6 |
These are approximate rates reported by GitHub for the 2025 measurement year, not population-adjusted participation rates. India alone added more than 5 million developers to GitHub during that year. Source: Octoverse 2025.
The 2024 measurement also recorded 1.4 million developers making their first open-source contribution. That number matters because joining a community and contributing publicly are different steps. A platform can grow quickly while newcomers remain concentrated in private work, learning exercises, or small projects.
Public, private, and open-source contribution patterns
Public visibility does not equal contribution volume. In GitHub’s 2025 overview, private repositories accounted for 81.5% of contributions. Public repositories represented 63% of all repositories, while private repositories grew 33% year over year and public repositories grew 19%.
GitHub’s 2024 period showed a similar pattern: more than 82% of contributions went to private repositories. Developers made 4.3 billion contributions across more than 181 million private repositories. At the same time, nearly 1 billion contributions went to public and open-source projects. Source: Octoverse 2024.
The contrast can be summarized as follows:
| Measure | Reported figure | Period |
|---|---|---|
| Private share of contributions | 81.5% | 2025 measurement year |
| Public share of repositories | 63% | 2025 measurement year |
| Private repository growth | 33% year over year | 2025 measurement year |
| Public repository growth | 19% year over year | 2025 measurement year |
| Contributions to private repositories | 4.3 billion | Oct. 1, 2023–Sept. 30, 2024 |
Public and private work also connect through shared tools and packages. JavaScript package consumption through the npm registry rose 15% year over year during GitHub’s 2024 measurement period. That is a measure of use rather than direct authorship, but it shows how community knowledge can circulate through dependencies even when the consuming work is private.
The 2024 report also counted more than 1 million maintainers, teachers, and students using free GitHub Copilot access. AI-related activity therefore appeared in both contribution and learning contexts, while the supplied figures do not measure the quality or durability of the resulting knowledge.
How contributors concentrate across leading projects
Contribution is distributed across many repositories, but leading public projects can attract thousands of participants. In GitHub’s 2024 public-project ranking, home-assistant/core had more than 21,000 contributors. microsoft/vscode and ProvableHQ/leo each had more than 20,000.
Other projects in the ranking also showed substantial participation:
- firstcontributions/first-contributions had more than 13,000 contributors.
- flutter/flutter had more than 10,000 contributors.
- NixOS/nixpkgs and vercel/next.js each had more than 9,000 contributors.
- langchain-ai/langchain had more than 8,000 contributors.
- godotengine/godot had more than 7,000 contributors.
These are contributor counts for named public projects in the 2024 ranking, not estimates of unique people across GitHub. A person may contribute to multiple repositories, and the figures do not show how frequently each contributor participated or how much work each contribution represented. Source for all project counts: Octoverse 2024.
Entry points can influence who participates. Among contributors to GitHub’s top 10 For Good First Issue projects, 34% made their first contribution after signing up for Copilot in the 2024 measurement year. This association describes timing, not a causal effect. It does, however, quantify how a tool-enabled onboarding path appeared alongside first contributions in that group.
Trust infrastructure is another part of participation at scale. In 2025, 47 of GitHub’s top 50 open-source projects, or 94%, used or were independently scanned by OpenSSF Scorecard. The statistic concerns security-practice coverage, not whether every project was equally secure or trustworthy.
Trust, usefulness, and friction
Stack Overflow’s 2024 Developer Survey shows that online knowledge exchange increasingly sits beside AI-assisted answers. Seventy-six percent of respondents were using or planning to use AI coding tools, up from 70% in 2023. Seventy-two percent viewed those tools favorably or very favorably.
Favorable views did not eliminate verification concerns. Forty-two percent trusted AI output accuracy in their workflows, while 31% did not trust it and 27% were neutral. Sixty-six percent identified distrust of AI output as a top organizational concern, and 63% cited missing codebase or institutional context. Among professional developers, 45% said AI tools were bad or very bad at handling complex tasks.
Respondents also identified practical benefits. Eighty-one percent cited increased productivity as AI’s biggest benefit, followed by 62% citing faster learning. Fifty-eight percent cited improved efficiency, and 30% cited improved coding accuracy. These are reported survey responses, not controlled measures of causal impact. Source: Stack Overflow’s 2024 analysis.
Human response time remains important even when automated tools are available. Fifty-three percent of surveyed developers agreed or strongly agreed that waiting for answers disrupts their workflow in 2024. That friction helps explain why searchable discussions, reusable explanations, and maintained project documentation remain valuable forms of community infrastructure.
Time spent sharing, finding, and learning
The time cost of knowledge exchange is substantial. More than 60% of surveyed developers spent at least 30 minutes per day searching for solutions in 2024, and one in four spent at least 60 minutes per day looking for answers. These survey percentages describe respondents’ reported behavior, not a platform-wide time total.
Contribution also includes repeated explanation. Three out of four surveyed developers answered questions they had answered before. Forty-seven percent spent at least 30 minutes per day answering questions. Thirty percent said knowledge silos affected their productivity at least 10 times per week. Together, the measures point to a cycle in which people both retrieve existing knowledge and recreate or clarify it for others. Source: Stack Overflow’s 2024 developer-workflow analysis.
Online media has become a larger part of how people learn to code. The share of developers learning through online media such as blogs and videos reached 82% in 2024, up from 60% in 2021. In the 2024 Stack Overflow learning survey, 59% said routine, hands-on experience was how they learned things confidently, while 23% selected writing about a topic. Teaching was preferred as a learning route by 14% of individual contributors and 9% of people managers. Source: Stack Overflow’s 2024 learning analysis.
The overall pattern is clear in the reported measurements: online communities support billions of contributions, millions of new participants, and extensive public reuse, while much activity remains private and much knowledge work still depends on human search, explanation, context, and trust.