Introduction
In May 2024, developers let us know which GenAI tools they were using and how it was improving their quality of time spent but not necessarily saving them time. The difference in perceptions of productivity a year ago in the Developer Survey's AI insights show that willingness to adopt new AI tools did not deliver the promise of productivity most had forecast. With our 2024 Developer Survey, we ask about usage and perception to compare to last year's responses, and added a few new questions about top-level concerns with AI at work.
Highlights
Current Use of AI Tools
Question: Do you currently use AI tools in your development process?
Current Use: Professionals vs. Learning to Code
More developers learning to code are using or are interested in using AI tools than professional developers, and this makes sense because developers learning are less likely to have preset loyalties to existing technologies.
| Category | Yes | No, but I plan to soon | No, and I don’t plan to | Sum: Use or Plan to Use |
|---|---|---|---|---|
| Learning to Code | 66.2% | 17.28% | 16.52% | 83.48% |
| Professional Developers | 63.26% | 13.35% | 23.39% | 76.61% |
Current Use: Top 10 Developer Roles
Of the top ten roles developers selected in the 2024 Developer Survey, mobile developers, full-stack developers and engineering managers are all most likely to be using or want to use AI tools. Different roles have different contexts for code; people managers are likely to have more influence over technology purchases, it makes sense that they are represented in top 3 roles using AI tools more than average.
| Category | Yes | No, but I plan to soon | No, and I don’t plan to | Sum: Use or Plan to Use |
|---|---|---|---|---|
| Developer, mobile | 65.19% | 17.28% | 17.53% | 82.47% |
| Developer, front-end | 69.06% | 12.93% | 18.01% | 81.99% |
| Engineering manager | 62.49% | 18.83% | 18.68% | 81.32% |
| Developer, full-stack | 65.3% | 12.92% | 21.78% | 78.22% |
| Data engineer | 60.79% | 16.33% | 22.88% | 77.12% |
| Developer, back-end | 61.97% | 13.92% | 24.11% | 75.89% |
| Student | 61.51% | 10.36% | 28.13% | 71.87% |
| Academic researcher | 57.8% | 10.76% | 31.44% | 68.56% |
| Developer, desktop or enterprise applications | 44.99% | 17.22% | 37.79% | 62.21% |
| Developer, embedded applications or devices | 43.46% | 14.09% | 42.45% | 57.55% |
Current Use: Years of Experience
Developers with less than 10 years experience are also more likely to have used or plan to use AI tools (> 77%). Those with less experience are closer to learning communities, these developers are influenced by peers that are learning and are more likely to be using AI tools.
| Category | Yes | No, but I plan to soon | No, and I don’t plan to | Sum: Use or Plan to Use |
|---|---|---|---|---|
| 1 to 4 years | 70.72% | 10.13% | 19.15% | 80.85% |
| Less than 1 year | 70.11% | 9.95% | 19.94% | 80.06% |
| 5 to 9 years | 63.88% | 13.43% | 22.68% | 77.31% |
| 10 to 14 years | 58.29% | 15.17% | 26.54% | 73.46% |
| 15 to 19 years | 55.24% | 16.71% | 28.05% | 71.95% |
| 20 + years | 49.16% | 18.27% | 32.57% | 67.43% |
Current Use: IC’s vs. People Manager
People managers are more likely than individual contributors to be using or interested in using AI tools. This makes sense, people managers are more likely to respond positively to "I have a great deal of influence" on technology purchases, such as senior executives (99%), engineering managers (87%), and product managers (77%).
| Category | Yes | No, but I plan to soon | No, and I don’t plan to | Sum: Use or Plan to Use |
|---|---|---|---|---|
| People Managers | 65.35% | 16.82% | 17.84% | 82.17% |
| Individual Contributors | 62.07% | 13.7% | 24.23% | 75.77% |
Current Use: Top 10 Countries
India, Ukraine, and Brazil are the most likely to be using or planning to use AI tools in their work (> 81%). In a report Coursera published to spotlight learning trends in different countries, India increased GenAI course enrollment in 2024 by 1,648%, Brazil by 1,079%.
| Category | Yes | No, but I plan to soon | No, and I don’t plan to | Sum: Use or Plan to Use |
|---|---|---|---|---|
| India | 67.94% | 21.53% | 10.53% | 89.47% |
| Ukraine | 72.34% | 14.58% | 13.08% | 86.92% |
| Brazil | 65.98% | 15.14% | 18.87% | 81.12% |
| Poland | 62.97% | 11.29% | 25.74% | 74.26% |
| Canada | 58.21% | 13.34% | 28.45% | 71.55% |
| Netherlands | 60.24% | 10.41% | 29.35% | 70.65% |
| Germany | 57.1% | 12.65% | 30.26% | 69.75% |
| United States of America | 54.25% | 13.38% | 32.37% | 67.63% |
| France | 54.2% | 11.46% | 34.34% | 65.66% |
| United Kingdom of Great Britain and Northern Ireland | 50.7% | 13.03% | 36.27% | 63.73% |
Perception of AI tools
Question: How favorable is your stance on using AI tools as part of your development workflow?
Favorable = Very & Favorable Indifferent = Indifferent & Unsure Unfavorable = Very & Unfavorable
Favorability: Professionals vs. Learning to Code
Professionals and those learning to code are fairly split on favorability, those learning to code are slightly more favorable and less indifferent.
| Category | Favorable | Indifferent | Unfavorable |
|---|---|---|---|
| Learning to Code | 74.05% | 15.8% | 5.2% |
| Professional Developers | 71.96% | 18.87% | 6.47% |
Favorability: Top 10 Developer Roles
Front-end developers, data engineers and academic researchers are the most favorable of AI tools out of the top 10 developer roles from the Developer Survey. These roles have had opportunity to experience benefits from GenAI.
| Category | Favorable | Indifferent | Unfavorable |
|---|---|---|---|
| Developer, front-end | 75.63% | 16.87% | 4.94% |
| Data engineer | 75.21% | 16.77% | 4.67% |
| Academic researcher | 74.14% | 13.9% | 9.03% |
| Engineering manager | 73.16% | 18.09% | 6.29% |
| Developer, full-stack | 72.24% | 18.69% | 6.44% |
| Developer, mobile | 71.58% | 19.13% | 5.39% |
| Student | 70.73% | 18.08% | 7.41% |
| Developer, back-end | 70.08% | 20.97% | 6.35% |
| Developer, desktop or enterprise applications | 67.31% | 21.09% | 7.38% |
| Developer, embedded applications or devices | 62.84% | 23.83% | 9.4% |
Favorability: Years of Experience
Similar to trends seen with the current usage among groups of experienced developers, those with less experience have higher rates of favorability and those with more experience are more likely to be indifferent.
| Category | Favorable | Indifferent | Unfavorable |
|---|---|---|---|
| Less than 1 year | 73.24% | 17.63% | 6.26% |
| 1 to 4 years | 74.56% | 17.49% | 6.1% |
| 5 to 9 years | 71.46% | 18.89% | 6.51% |
| 10 to 14 years | 70.25% | 20.32% | 6.56% |
| 15 to 19 years | 69.24% | 19.93% | 7.71% |
| 20 + years | 68.4% | 21.17% | 7.12% |
Favorability: IC’s vs. People Managers
These two groups are fairly similar in their favorability of AI tools.
| Category | Favorable | Indifferent | Unfavorable |
|---|---|---|---|
| People Managers | 73.35% | 18.27% | 6.32% |
| Individual Contributors | 71.51% | 19% | 6.85% |
Favorability: Top 10 Countries
Brazil, India, and France are all most favorable to AI (> 74%); Brazil and India are also more likely to be using or planning to use AI tools than other countries in the top 10 group.
| Category | Favorable | Indifferent | Unfavorable |
|---|---|---|---|
| Brazil | 80.08% | 14.43% | 3.33% |
| India | 78.57% | 14.63% | 3.51% |
| France | 74.22% | 14.64% | 6.63% |
| Canada | 72.04% | 18.55% | 7.2% |
| Netherlands | 70.43% | 19.35% | 7.66% |
| United States of America | 68.54% | 20.25% | 8.66% |
| United Kingdom of Great Britain and Northern Ireland | 67.16% | 22.12% | 9.2% |
| Ukraine | 65.24% | 26.81% | 4.09% |
| Germany | 64.31% | 23.23% | 8.75% |
| Poland | 61.9% | 23.87% | 9.99% |
Benefits of AI tools
Question: For the AI tools you use as part of your development workflow, what are the MOST important benefits you are hoping to achieve?
Benefits: Professionals vs. Learning to Code
Those learning to code are focused on leveling up to a professional position and this most likely influenced their top choice of “speed up learning” as AI tools’ top benefit (76% v. 72% for “increase productivity”). This may be the reason those with less experience also embrace this emerging technology.
| Category | Increase productivity | Speed up learning | Greater efficiency | Improve accuracy in coding | Make workload more manageable | Improve collaboration |
|---|---|---|---|---|---|---|
| Learning to Code | 72.18% | 76.38% | 57.68% | 39.7% | 31.39% | 12.06% |
| Professional Developers | 82.59% | 60.87% | 58.45% | 28.72% | 24.27% | 7.51% |
Benefits: Top 10 Developer Roles
All of the top 10 developer roles agree that "increased productivity" is the top benefit of GenAI. For the secondary benefit of "speed up learning", data engineers and students rank it higher than other roles, and for "greater efficiency" engineering managers and students rank it higher than other roles.
Students with access to student discounts and free trials are able to utilize some GenAI tools with less cost burden than their professional peers, but it is assumed that cost is still prohibitive for the majority of GenAI tools.
| Category | Increase productivity | Speed up learning | Greater efficiency | Improve accuracy in coding | Make workload more manageable | Improve collaboration |
|---|---|---|---|---|---|---|
| Academic researcher | 76.8% | 58.88% | 57.86% | 25.99% | 25.99% | 6.61% |
| Data engineer | 82.6% | 67.02% | 60.46% | 31.45% | 23.51% | 8.4% |
| Developer, back-end | 81.48% | 61.07% | 54.22% | 27.2% | 22.42% | 6.74% |
| Developer, desktop or enterprise applications | 80.68% | 63.57% | 56.76% | 28.43% | 21.16% | 6.26% |
| Developer, embedded applications or devices | 75.04% | 64.78% | 55.42% | 20.51% | 20.51% | 3.57% |
| Developer, front-end | 82.54% | 58.07% | 60.56% | 33.83% | 25.92% | 8.23% |
| Developer, full-stack | 83.26% | 60.76% | 60.48% | 29.68% | 25.03% | 7.56% |
| Developer, mobile | 82.59% | 61.79% | 56.41% | 34.34% | 27.93% | 11.16% |
| Engineering manager | 84.03% | 58.44% | 61.69% | 28.18% | 22.08% | 6.49% |
| Student | 74.72% | 68.72% | 60.81% | 33.65% | 27.62% | 6.79% |
Benefits: Years of Experience
Later-career developers are more likely to rate "increase in productivity" higher than less-experienced developers. Developers with 1 - 4 years of experience rate "speed up learning" and "greater efficiency" higher than other experience groups. Productivity is important to all experience levels, but finding ways to save time learning or coding are valuable to those that may be spending extra time doing both.
| Category | Increase productivity | Speed up learning | Greater efficiency | Improve accuracy in coding | Make workload more manageable | Improve collaboration |
|---|---|---|---|---|---|---|
| Less than 1 year | 78.52% | 70.35% | 59.87% | 32.65% | 29.23% | 7.95% |
| 1 to 4 years | 80.99% | 64.75% | 61.64% | 30.41% | 28.75% | 8.22% |
| 5 to 9 years | 82.83% | 59.51% | 59.8% | 28.07% | 25.06% | 7.73% |
| 10 to 14 years | 83.98% | 58.25% | 56.66% | 28.05% | 22.51% | 7.64% |
| 15 to 19 years | 84.24% | 57.04% | 56.53% | 28.34% | 18.46% | 6.04% |
| 20 + years | 82.86% | 60.37% | 53.4% | 28.21% | 17.64% | 5.4% |
Benefits: IC’s vs. People Managers
People managers are more likely to rate all the listed benefits higher than individual contributors except for "speed up learning" but the percentages difference are very close.
| Category | Increase productivity | Speed up learning | Greater efficiency | Improve accuracy in coding | Make workload more manageable | Improve collaboration |
|---|---|---|---|---|---|---|
| People Managers | 84.57% | 61.08% | 59.97% | 31.05% | 26.66% | 9.65% |
| Individual Contributors | 82.69% | 61.35% | 59.63% | 28.32% | 24.5% | 7.55% |
Benefits: Top 10 Countries
India rates the less selected benefits of "improve accuracy", "make workloads more manageable", and "improve collaboration" higher than other countries. India is the top country for AI tool usage in the 2024 Developer Survey.
| Category | Increase productivity | Speed up learning | Greater efficiency | Improve accuracy in coding | Make workload more manageable | Improve collaboration |
|---|---|---|---|---|---|---|
| Brazil | 84.61% | 63.71% | 47.08% | 33.93% | 24.49% | 10.34% |
| Canada | 80.62% | 62.41% | 61.24% | 26.65% | 22.14% | 6.18% |
| France | 80.55% | 53% | 56.64% | 25.27% | 16% | 4% |
| Germany | 81.4% | 57.37% | 60.39% | 20.82% | 18.75% | 3.93% |
| India | 83.81% | 67.25% | 59.01% | 43.36% | 43.58% | 15.94% |
| Netherlands | 78.92% | 51.05% | 55.39% | 19.79% | 17.33% | 3.63% |
| Poland | 78.48% | 59.64% | 56.1% | 26.02% | 19.7% | 4.39% |
| Ukraine | 82.79% | 67.68% | 48.95% | 32.63% | 22.11% | 9.37% |
| United Kingdom of Great Britain and Northern Ireland | 80.97% | 58.77% | 59.7% | 24.94% | 19.34% | 4.04% |
| United States of America | 80.63% | 63.1% | 63.4% | 28.08% | 23.05% | 5.32% |
Trust in Accuracy of AI Tools
Question: How much do you trust the accuracy of the output from AI tools as part of your development workflow?
Trust: Professionals vs. Learning to Code
Developers learning to code have a lot less to lose if they misplace their trust in AI tools: this could be why professional developers rate AI tools less trustworthy than those learning to code (41% v. 55%).
| Category | Trustworthy | Trust neutral | Not trustworthy |
|---|---|---|---|
| Learning to Code | 54.91% | 24.59% | 20.51% |
| Professional Developers | 41.43% | 27.25% | 31.32% |
Trust: Top 10 Developer Roles
Context is key: front-end developers rate AI tools most trustworthy and embedded developers rate it least trustworthy. AI tools are better at assisting with front-end technologies.
| Category | Trustworthy | Trust neutral | Not trustworthy |
|---|---|---|---|
| Developer, front-end | 49.49% | 27.47% | 23.05% |
| Developer, mobile | 46.67% | 29.33% | 24% |
| Student | 45.47% | 22.8% | 31.73% |
| Data engineer | 45.36% | 25.11% | 29.53% |
| Engineering manager | 41.87% | 26.89% | 31.24% |
| Developer, full-stack | 41.33% | 27.87% | 30.79% |
| Developer, back-end | 41.26% | 27.31% | 31.43% |
| Developer, desktop or enterprise applications | 38.16% | 25.77% | 36.07% |
| Academic researcher | 37.19% | 21.42% | 41.39% |
| Developer, embedded applications or devices | 28.84% | 26.96% | 44.2% |
Trust: Years of Experience
Developers with > 10 years experience are likely to be equal in their rating of 'not trustworthy' but there is a spike in trustworthiness for the 20+ years of experience group: most likely this in influenced by people managers that tend to be more positive towards AI tools.
| Category | Trustworthy | Trust neutral | Not trustworthy |
|---|---|---|---|
| Less than 1 year | 45.79% | 24.36% | 29.85% |
| 1 to 4 years | 43.38% | 27.99% | 28.63% |
| 5 to 9 years | 41.48% | 27.14% | 31.39% |
| 10 to 14 years | 39.28% | 26.53% | 34.19% |
| 15 to 19 years | 37.27% | 28.1% | 34.64% |
| 20 + years | 40.44% | 24.65% | 34.91% |
Trust: IC’s vs. People Managers
People managers are more likely to find AI tools trustworthy than individual contributors (43% v. 40%), but the margin of difference is small. People managers are more likely to be responsible for disruptions in business caused by a mistake originating from their team, it makes sense that they would remain cautious.
| Category | Trustworthy | Trust neutral | Not trustworthy |
|---|---|---|---|
| People Managers | 42.46% | 27.87% | 29.66% |
| Individual Contributors | 40.35% | 26.25% | 33.4% |
Trust: Top 10 Countries
India, Brazil, and Ukraine rate AI tools the most trustworthy out of all top 10 countries; they are also the top 3 countries using AI tools.
| Category | Trustworthy | Trust neutral | Not trustworthy |
|---|---|---|---|
| India | 59.1% | 28.3% | 12.6% |
| Brazil | 45.89% | 27.78% | 26.33% |
| Ukraine | 43.6% | 36.25% | 20.15% |
| United States of America | 40.35% | 20.69% | 38.97% |
| Netherlands | 38.17% | 25.14% | 36.67% |
| United Kingdom of Great Britain and Northern Ireland | 38.03% | 21.97% | 40% |
| Canada | 37.67% | 23.68% | 38.66% |
| France | 35.6% | 33.22% | 31.18% |
| Germany | 32.82% | 24.28% | 42.9% |
| Poland | 29.78% | 29.68% | 40.55% |
Areas of Development Workflow Most Popular for those Currently Using AI Tools
Question: Which parts of your development workflow are you currently using AI tools for and which are you interested in using AI tools for over the next year?
Currently Using: Professionals vs. Learning to Code
For developers currently using AI tools, writing code remains the top benefit similar to last year. 85% of professional developers use them to write code, followed by 69% that use them to search answers. These two benefits are the focus of time-saving GenAI tooling that seeks to increase productivity.
| Usage | Learning to Code | Professional Developers |
|---|---|---|
| Writing code | 77.34% | 84.76% |
| Search for answers | 72.81% | 68.92% |
| Debugging and getting help | 66.98% | 56.74% |
| Documenting code | 32.97% | 42.43% |
| Generating content or synthetic data | 33.47% | 36.41% |
| Learning about a codebase | 42.08% | 30.36% |
| Testing code | 24.3% | 29.65% |
| Committing and reviewing code | 22.61% | 12.96% |
| Project planning | 22.31% | 11.31% |
| Predictive analytics | 7.52% | 4.9% |
| Deployment and monitoring | 7.12% | 4.32% |
Currently Using: Top 10 Developer Roles
Data engineers and engineering managers are most likely to be using AI tools for writing code if they are currently using them in their workflow, data engineers and desktop developers are most likely to be using them to search for answers, and students and data engineers are most likely to be using them to debug their code.
| Role | Writing code | Search for answers | Debugging and getting help | Documenting code | Generating content or synthetic data | Learning about a codebase | Testing code | Committing and reviewing code | Project planning | Predictive analytics | Deployment and monitoring |
|---|---|---|---|---|---|---|---|---|---|---|---|
| Academic researcher | 84.78% | 65.45% | 59.06% | 44.29% | 34.09% | 35.01% | 23.44% | 12.94% | 14% | 9.28% | 5.63% |
| Data engineer | 87.74% | 72.17% | 66.04% | 46.38% | 37.11% | 29.4% | 27.67% | 13.05% | 11.16% | 5.35% | 4.25% |
| Developer, back-end | 84.26% | 68.9% | 51.29% | 42.75% | 33.17% | 29.09% | 31.93% | 12.14% | 9.23% | 4% | 4.03% |
| Developer, desktop or enterprise applications | 83.56% | 70.44% | 49.52% | 36.98% | 27.57% | 30.51% | 20.34% | 12.17% | 8.08% | 3.52% | 2.95% |
| Developer, embedded applications or devices | 76.82% | 69.24% | 54.09% | 39.55% | 25.61% | 28.33% | 18.94% | 8.33% | 6.52% | 2.88% | 1.52% |
| Developer, front-end | 85.16% | 69.37% | 62.68% | 40.15% | 40.2% | 30.49% | 33.46% | 14.47% | 10.23% | 3.53% | 3.68% |
| Developer, full-stack | 85.43% | 68.51% | 58.38% | 40.74% | 37.84% | 30% | 29.14% | 12.87% | 12.1% | 4.3% | 4.02% |
| Developer, mobile | 84.42% | 68.51% | 54.1% | 37.27% | 35.01% | 32.16% | 26.63% | 14.99% | 13.07% | 6.03% | 5.19% |
| Engineering manager | 86.23% | 67.25% | 49.47% | 43.18% | 35.96% | 29.14% | 33.16% | 12.83% | 9.49% | 6.02% | 4.55% |
| Student | 80.23% | 69.2% | 68.02% | 37.05% | 33.84% | 33.05% | 18.52% | 15.52% | 17.73% | 5.07% | 3.9% |
Currently Using: Years of Experience
Developers with more experience are more likely to be using AI tools to write code, less experienced developers are more likely to be using AI to committing and reviewing code and project planning. Later-career developers are more adept at knowing what they don't know, and can utilize AI tools more confidently, while those new to their role have yet to learn from mistakes.
| Usage | Less than 1 year | 1 to 4 years | 5 to 9 years | 10 to 14 years | 15 to 19 years | 20 + years |
|---|---|---|---|---|---|---|
| Writing code | 81.94% | 82.81% | 85.01% | 85.84% | 86.9% | 88.05% |
| Search for answers | 73.13% | 71.68% | 68.18% | 65.23% | 65.4% | 70.42% |
| Debugging and getting help | 69.7% | 65.02% | 57.54% | 50.86% | 47.74% | 47.27% |
| Documenting code | 39.28% | 43.12% | 43.55% | 43.16% | 41.03% | 38.48% |
| Generating content or synthetic data | 39.78% | 39.07% | 36.34% | 35.03% | 32.74% | 30.4% |
| Learning about a codebase | 38.23% | 32.86% | 28.54% | 28.14% | 28.78% | 31.12% |
| Testing code | 22.38% | 26.34% | 31.88% | 31.65% | 29.1% | 28.3% |
| Committing and reviewing code | 16.12% | 14.28% | 12.46% | 11.78% | 10.83% | 11.51% |
| Project planning | 19.67% | 14.8% | 10.89% | 9.28% | 6.99% | 6.46% |
| Predictive analytics | 6.65% | 5.66% | 5.02% | 4.34% | 4.12% | 4.64% |
| Deployment and monitoring | 5.65% | 5.19% | 4.26% | 3.68% | 3.03% | 3.13% |
Areas of Development Workflow Most Popular for those Interested in Using AI Tools
Question: Which parts of your development workflow are you currently using AI tools for and which are you interested in using AI tools for over the next year?
Interested in Using: Professionals vs. Learning to Code
Having the luxury of time is hard to find for busy developers. If you are learning to code, you may have some more time available to learn and these developers want to test code and document code with AI, while professionals would also like to spend time learning how to test code with AI but also would rather learn about committing and reviewing code with AI tools, as well as using AI tools to learn about codebases.
| Usage | Learning to Code | Professional Developers |
|---|---|---|
| Writing code | 14.54% | 10.36% |
| Search for answers | 19.44% | 21.01% |
| Debugging and getting help | 22.75% | 31.62% |
| Documenting code | 47.28% | 44.48% |
| Generating content or synthetic data | 36.11% | 39.1% |
| Learning about a codebase | 41.9% | 48.82% |
| Testing code | 52.54% | 54.14% |
| Committing and reviewing code | 41.02% | 49.03% |
| Project planning | 45.04% | 36.26% |
| Predictive analytics | 42.91% | 47.83% |
| Deployment and monitoring | 43.26% | 47.74% |
Interested in Using: Top 10 Developer Roles
Testing code is a top reason to get started with AI tools for those interested in using them: engineering managers and embedded developers are the top roles interested in testing code with AI. Learning about a codebase follows as another popular option, and moreso for data engineers and embedded developers.
| Role | Writing code | Search for answers | Debugging and getting help | Documenting code | Generating content or synthetic data | Learning about a codebase | Testing code | Committing and reviewing code | Project planning | Predictive analytics | Deployment and monitoring |
|---|---|---|---|---|---|---|---|---|---|---|---|
| Academic researcher | 9.13% | 18.57% | 26.94% | 35.92% | 30.29% | 36.83% | 46.27% | 36.53% | 27.4% | 27.7% | 32.27% |
| Data engineer | 8.49% | 17.45% | 22.8% | 39.94% | 37.89% | 46.54% | 52.99% | 43.08% | 34.43% | 47.17% | 44.65% |
| Developer, back-end | 8.99% | 18.08% | 30.44% | 37.57% | 35.25% | 43.58% | 44.21% | 42.67% | 29.79% | 40.13% | 40.7% |
| Developer, desktop or enterprise applications | 11.22% | 17.21% | 31.08% | 39.45% | 38.78% | 43.44% | 52.57% | 41.63% | 28.99% | 38.88% | 36.41% |
| Developer, embedded applications or devices | 13.18% | 19.39% | 31.67% | 41.67% | 37.42% | 45.76% | 53.64% | 40.76% | 27.88% | 35.15% | 34.85% |
| Developer, front-end | 10.08% | 17.34% | 25.07% | 40.39% | 28.46% | 43.12% | 46.18% | 44.3% | 36.66% | 40.9% | 43.12% |
| Developer, full-stack | 8.37% | 18.59% | 26.19% | 40.09% | 33.59% | 41.84% | 47.94% | 42.33% | 32.17% | 42.58% | 42.5% |
| Developer, mobile | 11.73% | 18.84% | 29.48% | 42.71% | 32.24% | 40.45% | 50% | 44.97% | 34.34% | 43.3% | 40.87% |
| Engineering manager | 10.43% | 22.46% | 38.24% | 46.79% | 40.24% | 51.47% | 54.14% | 53.07% | 43.32% | 53.34% | 53.61% |
| Student | 8.11% | 17.25% | 17.9% | 39.15% | 31.84% | 37.91% | 44.36% | 32.53% | 34.12% | 32.6% | 31.42% |
Interested in Using: Years of Experience
Seasoned developers are more interested in testing code or committing code with AI tools than their less experienced peers. Again, the logic here is that the additional confidence in the development workflow and knowing where mistakes typically happen can bolster the desire to look into these reasons to use AI for efficiency. Learning about a codebase has a very consistent rate of interest among those with more than 5 and less than 20 years of experience; this experience-level may have the most to learn from codebases.
| Usage | Less than 1 year | 1 to 4 years | 5 to 9 years | 10 to 14 years | 15 to 19 years | 20 + years |
|---|---|---|---|---|---|---|
| Writing code | 9.47% | 9.66% | 9.22% | 8.53% | 8.73% | 8.49% |
| Search for answers | 14.57% | 15.92% | 19.08% | 21.34% | 21.54% | 18.43% |
| Debugging and getting help | 17.62% | 20.92% | 27.45% | 32.14% | 36.26% | 36.68% |
| Documenting code | 39.34% | 38% | 39.2% | 39.5% | 40.18% | 41.55% |
| Generating content or synthetic data | 29.2% | 30.43% | 34.35% | 36.61% | 39.05% | 41.32% |
| Learning about a codebase | 36.57% | 39.95% | 44.46% | 45.08% | 44.83% | 42.81% |
| Testing code | 44.88% | 45.32% | 47.25% | 48.67% | 52.02% | 53.58% |
| Committing and reviewing code | 32.85% | 37.34% | 43.28% | 48.12% | 49.64% | 50.45% |
| Project planning | 31.41% | 32.53% | 33.68% | 32.27% | 29.71% | 28.86% |
| Predictive analytics | 34.24% | 37.74% | 41.99% | 45.47% | 47.45% | 45.81% |
| Deployment and monitoring | 33.74% | 37.68% | 41.92% | 45.15% | 47.94% | 46.27% |
Perception of AI Tools for Complex Tasks
Question: How well do the AI tools you use in your development workflow handle complex tasks?
Complex: Professional Developers vs. Learning to Code
Again, context is key when talking about AI tool efficacy. Almost half of professional developers (45%) believe AI tools are bad at complex tasks and vice versa for those learning to code (46% believe AI is good at complex task).
| Category | Complex good | Complex neutral | Complex bad |
|---|---|---|---|
| Learning to Code | 46.31% | 20.84% | 32.85% |
| Professional Developers | 34.29% | 20.82% | 44.89% |
Complex: Top 10 Developer Roles
Mobile developers and data engineers are most likely to agree AI tools are good at complex tasks and embedded developers are most likely to believe they are bad at complex tasks. This reflects similar trends in roles that find AI more or less trustworthy.
| Category | Complex good | Complex neutral | Complex bad |
|---|---|---|---|
| Developer, mobile | 41.6% | 21.99% | 36.41% |
| Data engineer | 38.38% | 20.64% | 40.97% |
| Developer, front-end | 37.86% | 21.75% | 40.39% |
| Student | 37.24% | 18.26% | 44.5% |
| Academic researcher | 36.17% | 15.59% | 48.23% |
| Developer, full-stack | 35.08% | 21.38% | 43.53% |
| Developer, desktop or enterprise applications | 33.88% | 18.13% | 47.98% |
| Developer, back-end | 32.02% | 21.11% | 46.87% |
| Engineering manager | 31.87% | 23.23% | 44.9% |
| Developer, embedded applications or devices | 23.76% | 17.49% | 58.74% |
Perception of AI as a Threat to Developer Jobs
Question: Do you believe AI is a threat to your current job?
Threat: Professional Developers vs. Learning to Code
A majority of professional developers (70%) are not threatened by the possibility of becoming redundant at work due to AI. Those learning and not yet in their first developer role have seen tech go through tough job cuts in the past two years and have less experience to know how they can contribute value as technology evolves.
| Category | No | I’m not sure | Yes |
|---|---|---|---|
| Professional Developers | 69.5% | 18.84% | 11.66% |
| Learning to Code | 53.26% | 29.94% | 16.8% |
Threat: Top 10 Developer Roles
Academic researchers are the least likely to perceive AI as a threat to their job, followed by embedded developers and engineering managers. These roles all deal with complex processes (managing developers may be more nuanced than complex) and at least embedded developers and engineering managers also were most likely of all the top 10 roles to rate AI bad at complex code.
| Category | No | I’m not sure | Yes |
|---|---|---|---|
| Academic researcher | 78.1% | 12.41% | 9.49% |
| Developer, embedded applications or devices | 76.09% | 16.23% | 7.68% |
| Engineering manager | 75.26% | 16.49% | 8.25% |
| Developer, desktop or enterprise applications | 70.92% | 18.6% | 10.48% |
| Data engineer | 70.88% | 18.75% | 10.37% |
| Developer, full-stack | 69.16% | 18.83% | 12% |
| Developer, back-end | 68.13% | 20.39% | 11.49% |
| Developer, mobile | 63.24% | 22.13% | 14.62% |
| Developer, front-end | 60.71% | 24.39% | 14.9% |
| Student | 58.62% | 27.45% | 13.92% |
Challenges of AI at Work
Question: What are the challenges to your company/whole team using AI code assistants or GenAI tools?
Challenges: Professional Developers vs. Learning to Code
Trusting AI tools for your own work is one thing, but when it comes to company-wide or team-wide adoption, more developers agree trust is a top challenge. Similarly to trust in accuracy, professional developers are more likely than those learning to code to select trust in output as a challenge but the margin between the two is smaller (66% v. 62% compared to 31% of professional developers who find AI not trustworthy compared to 21% of those learning to code.)
| Category | AI tools lack context of codebase, internal architecture, and/or company knowledge | Don’t trust the output or answers | Lack of executive buy-in | Lack of proper training and education on new tools | Not everyone uses them | They create more work (more code/PRs to review, etc.) | We don’t have the right policies in place to reduce security risks |
|---|---|---|---|---|---|---|---|
| Professional Developers | 64.72% | 66.22% | 11.87% | 29.58% | 25.81% | 12.21% | 31.85% |
| Learning to Code | 55.4% | 62.37% | 9.48% | 37.42% | 24.39% | 19.44% | 27.53% |
Challenges: Top 10 Developer Roles
Embedded developers and data engineers are most likely to find codebase context a challenge for AI among the top 10 roles from the Developer Survey; these two roles also were most interested in using AI to learn about codebases in the coming year. The challenge here may be more about seeing a successful proof of concept.
| Category | AI tools lack context of codebase, internal architecture, and/or company knowledge | Don’t trust the output or answers | Lack of executive buy-in | Lack of proper training and education on new tools | Not everyone uses them | They create more work (more code/PRs to review, etc.) | We don’t have the right policies in place to reduce security risks |
|---|---|---|---|---|---|---|---|
| Academic researcher | 53.13% | 73.1% | 10.83% | 34.35% | 22.17% | 15.4% | 31.64% |
| Data engineer | 65.04% | 64.03% | 12.94% | 32.1% | 26.05% | 10.42% | 35.29% |
| Developer, back-end | 64.65% | 66.1% | 11.41% | 26.8% | 24.15% | 11.47% | 32.38% |
| Developer, desktop or enterprise applications | 64.77% | 67.41% | 12.63% | 31.26% | 25.87% | 11.91% | 32.28% |
| Developer, embedded applications or devices | 67.21% | 71.57% | 14.22% | 29.56% | 22.46% | 13.41% | 33.6% |
| Developer, front-end | 63.8% | 61.33% | 11.82% | 29.43% | 27.77% | 12.19% | 30.99% |
| Developer, full-stack | 65.55% | 66.69% | 12.06% | 29.89% | 26.57% | 12.38% | 31.16% |
| Developer, mobile | 60.49% | 60.21% | 10.35% | 29.16% | 28.79% | 11.76% | 31.42% |
| Engineering manager | 63.19% | 64.56% | 13.19% | 36.26% | 27.61% | 9.89% | 32.01% |
| Student | 61.67% | 72.52% | 6.66% | 33.81% | 19.61% | 19.66% | 24.27% |
