A look back before we look forward: A Developer Survey retrospective

[kasenur]

We are on the precipice of brand new results from the 2026 Developer Survey. We have 15 years of data and insights from this survey at our fingertips that shows more than just what one year’s collection of insights will explore. A year-over-year look at the Developer Survey offers the long view of what is changing in the broader technology ecosystem. This analysis compares the 2024 and 2025 Developer Survey data across three connected stories: the evolution of AI, humans at work, and demographics and community. The survey results reveal what the headline news cannot: how people actually fold these tools into daily work, where they still want human judgment, and how changing workflows affect the communities developers rely on.The AI evolution from copilots to autonomous systemsAI tool usage has grown in our survey since we initially started asking about AI tools in the development workflow in 2023. 44% of developers indicated they were using AI tools in 2023, growing to 62% in 2024 and 79% in 2025, with almost half of all respondents indicating they used AI tools daily. What can the survey show us about this growth in AI usage that we don’t already know from news stories about frontier lab spending, AI stocks and earnings reports increasingly citing organizational restructuring enhanced with AI efficiencies?We started asking about AI agents in the 2025 Developer Survey, with 31% of respondents indicating they were using them and most using ChatGPT (82%) and GitHub Copilot (68%). We held a smaller pulse survey in April 2026 asking about agents again, and this time we see agent use as almost doubled (59%) and Claude Code increasing in usage (from 41% to 55%). Something else the pulse survey on AI agents pointed out was that usage was driven by those that were daily users and those that were in more executive roles.AI tools and their usesWhen developers step into the world of AI tooling, ChatGPT and GitHub Copilot serve as the primary gateways. In our 2025 survey, 82% of those utilizing out-of-the-box assistants selected ChatGPT, while 68% turned to GitHub Copilot. Google Gemini trailed at 47%, followed by Claude Code at 41% and Microsoft Copilot at 31%. Farther down the long tail, niche offerings saw smaller footprints: Perplexity reached 16%, v0.dev hit 9%, Bolt.new gathered 7%, and Lovable.dev captured 6%. This breakdown suggests developers gravitate toward established platforms prior to experimenting with tailored autonomous setups.Year-over-year trends reinforce this dynamic, even though survey design evolved. Back in 2024, when respondents marked single active usage options, 82% applied AI toward writing code, 68% relied on it for research, 57% tapped it for debugging, 40% drafted documentation, and 35% created general content. Because 2025 split these tasks between primary and secondary reliance, direct line-item comparisons are tricky. Nevertheless, the underlying trajectory stays clear: discovery and drafting lead the pack in automation, while actual software engineering remains predominantly assisted rather than autonomous.Learning to use AIIndividuals who are learning to program are adopting automated tooling at an accelerated pace even as caution creeps in. In 2024, 63.0% of student respondents reported utilizing AI resources; by 2025, that metric climbed to 73.3%. This includes 39.5% engaging with these platforms daily, 18.7% weekly, and 15.1% monthly. This technology shift is part of everyone’s educational journey, with newcomers jumping in just as frequently as experienced professionals.However, overall enthusiasm has noticeably tempered. Positive sentiment among learners dropped from 72.2% in 2024 (comprising 22.4% very favorable and 49.8% favorable) down to 52.8% in 2025 (19.3% very favorable, 33.5% favorable). Meanwhile, skepticism quadrupled, rising from 6.4% up to 26.3%. The overall survey population showed a similar cooling trend, moving from 72.0% positive and 6.4% negative in 2024 to 59.7% positive and 20.4% negative in 2025. Initial novelty is giving way to pragmatic evaluation as students confront inaccurate, incomplete, or confusing outputs.Confidence in output accuracy similarly waned among students, with moderate or high trust sliding from 49.1% in 2024 (44.9% somewhat, 4.2% high) to 37.4% in 2025 (31.3% somewhat, 6.1% high). While 2025 introduced a neutral option that redistributed responses, the broader shift highlights a more discerning audience. Sentiments around handling intricate assignments echo this sentiment: favorable ratings for complex work fell from 41.5% in 2024 to 33.7% in 2025, while 18.6% admitted they avoid using AI for advanced tasks or remain unsure of its reliability.Autonomous agents haven't quite won over the learning cohort, either. In 2025, 28.4% of students utilized agents, whereas 44.1% expressed no intention to implement them, compared to 37.9% across all survey takers. Concern over AI replacing jobs stayed flat at 15.0% across both years, though uncertainty dropped from 27.1% to 21.3% as unconcerned responses grew from 57.9% to 63.6%. Usage might be accelerating, but healthy skepticism remains, underscoring that using tools for study doesn't mean blindly embracing every feature on the horizon.Work is about more than just a paycheckWork-life balance continues to be an evolving story with the shifting office landscape. We have been tracking work setting in the Developer Survey since 2019 when only 12% of survey respondents indicated they worked fully remote. Since then, we saw a peak in remote work being reported in 2022 with 43% being fully remote. Hybrid work settings became the new majority for respondents in 2023 and 2024 at 42%. Last year we started capturing whether the hybrid situation was left up to choice, 13% indicated that their work situation was flexible enough for them to choose when to come into the office. We continue to ask about work setting this year and are curious about the implications it has with job satisfaction.J

 

[PinarToprak_61]

Çok faydalı, hemen deneyeceğim.