When was the last time you recall learning something new? It might have been after passing a test in class or publishing a paper for an academic program. It might have been while you were troubleshooting code or researching a new tool to solve a problem at work. Maybe the last thing you learned was so simple or subtle that it never registered as a new learning until you were able to apply it later. Developers and technologists have been learning to code using online media such as blogs and videos increasingly in the last four years according to the Stack Overflow Developer Survey–60% in 2021 elevated to 82% in 2024. The most recent useful resource that builders may make the most of for studying is generative AI which is rising as a key device that provides real-time problem-solving help, customized coding ideas, and revolutionary methods to boost talent improvement seamlessly built-in inside each day workflows.
There was loads of pleasure on this planet of software program improvement about AI’s potential to extend the pace of studying and entry to extra data. Hypothesis abounds as as to if studying will likely be helped or hindered by AI development. Our latest survey of over 700 builders and technologists reveals the method of realizing issues is simply that—a course of. New insights about how the Stack Overflow group learns exhibit that software program professionals choose to realize and share data by way of hands-on interactions. Their preferences for sourcing and contributing to teams or people (or AI) supplies colour on the evolving panorama of data work.
We requested builders what helps them really feel assured about realizing what they know, and a majority of working skilled builders agree doing work is far more significant to the method (59%) than studying about it (45%), educating it to somebody (39%), and even the chance to interface with the first supply (38%).
On the subject of feeling assured about your data base, builders throughout expertise ranges are remarkably aligned. Nearly all of each early-career builders (19% of these with 5 or much less years of expertise) and seasoned professionals (18% of these with 25 years or extra expertise) rank hands-on, in-depth routines as their path to data with confidence. Even folks managers, who could be additional faraway from actions equivalent to day-to-day coding, worth sensible expertise virtually equal to their particular person contributor counterparts (17% and 19% respectively).
This choice isn’t linear, as different studying sources wax and wane in recognition for these with extra and fewer work expertise: studying peaks for much less skilled (ten or much less years) professionals, and search engine use or sourcing from boards and on-line communities turns into widespread amongst late-career professionals (10 – 15 years expertise) however falls out of favor among the many most senior professionals (ranked fourth out of the highest six studying sources for these with 25+ years of expertise). Provided that particular person contributors are huge proponents of studying by way of educating (most well-liked by 14%) and folks managers choose utilizing conventional search engines like google and yahoo to be taught new issues (additionally 14%), we will surmise that the function dictates how knowledgeable learns and never the work expertise.
The motivations behind studying actually are related to extrinsic targets (e.g. getting a certification that may result in a pay improve) however studying can be associated to a need to be autonomous (e.g. proudly owning a GitHub repo for a brand new device as an alternative of submitting a difficulty to an present device) and builders in a person contributor function have extra runway to extend their autonomous worth than builders who’re liable for a workforce or division of others which in flip adjustments their motivations to maybe be extra synergistic (e.g. delegating particular studying actions to workforce members slightly than tackle a brand new duty). Position and private targets can have an effect on how knowledgeable developer chooses to interact in studying in several methods all through their profession.
The trail to data is not solitary. Our survey reveals a powerful choice for collective studying that transcends expertise ranges. The youthful the respondent, the extra seemingly they’re to choose collective sources (a variety of 27-46%), and whereas this choice weakens with older generations it doesn’t strengthen their choice for studying through particular person sources however slightly for having no choice in any respect (a variety of 18-36%).
Respondents cite the rubber duck method, a course of the place a programmer verbalizes their code and logic, as a part of the explanation they like sourcing data from a collective supply on this survey. This sentiment is validated within the 14% of particular person contributors preferring studying by way of educating, in addition to the element offered in an accepted reply or a solution with many feedback on Stack Overflow posts. Respondents that had been vital of collective sources of data point out lack of belief and quotation, and typically lack of consensus. Related issues with belief in AI instruments amongst builders are validated within the latest developer survey responses.
Whereas builders strongly choose collective sources of data, their working preferences inform a extra nuanced story. These with 5 or extra years of expertise are virtually evenly cut up between preferring collaboration and solo work (24% and 26% for 5-25 years of expertise in comparison with 31% and 32% for 25+ years of expertise) whereas these with lower than 5 years expertise choose collaborative work (36%) and people with 5-25 years expertise choose neither (53%). Skilled builders lean into utilizing AI as a companion at work about equally in all ranges of labor expertise, ranging 10-12% amongst the teams. Might this recommend that AI instruments are discovering a “candy spot” the place AI enhances the prevailing workflow of builders throughout numerous ranges of expertise? Respondents that don’t choose collaboration specific that their introverted personalities are extra conducive to working alone and that collaboration provides extra time to their course of.
When evaluating preferences for studying sources and work type, some anticipated and sudden outcomes emerge. Those that choose collaboration present a stronger tendency towards collective studying sources (52% overlap) and are much less prone to choose particular person or AI-based studying strategies (9% and eight% respectively). Respondents preferring to work alone present increased percentages for particular person studying preferences (28% don’t have any choice and 22% choose particular person sources of data to be taught from), but nonetheless fee collective sources highest at 34%. Even those who choose working with AI slightly than alone or collaboratively fee collective studying as a choice at 30% behind 37% with AI search as their studying choice.
We all know Stack Overflow customers are utilizing AI at work (virtually two thirds of professional devs are using AI), however AI instruments don’t match of their studying course of in accordance with this survey. After we requested skilled builders why they really feel assured about realizing one thing, AI instruments had been a lesser consideration together with conventional faculty (each chosen by 26%). Some builders reward AI’s chat interface and broad data base, but others stay skeptical, citing workflow integration challenges and an total lack of belief. Sound acquainted? These are the identical issues we have heard about collective data sources—however communities have earned their trust over time. The sturdy choice for collective data sources means that group platforms and collaborative studying environments are nonetheless vital, whilst AI instruments turn into extra subtle. For these constructing developer instruments and platforms, the message is evident: deal with enhancing, not changing, group interplay. Essentially the most profitable AI implementations could be those who facilitate higher group studying slightly than attempting to substitute for it.
