Software Engineering Daily
Episode: The 2025 Stack Overflow Developer Survey
Guests: Jody Bailey (Chief Product and Technology Officer, Stack Overflow), Aaron Yeppes (Research Manager, Stack Overflow)
Host: Shawn Falconer
Release Date: November 13, 2025
Episode Overview
This episode dives into the results and key insights from the 2025 Stack Overflow Developer Survey. Host Shawn Falconer is joined by Jody Bailey and Aaron Yeppes from Stack Overflow to discuss the trends, hot topics, and surprising findings shaping the global developer community. They focus heavily on AI adoption, developer sentiment, and evolving technologies, probing what the survey says about the state—and future—of software engineering.
What is the Stack Overflow Developer Survey?
[01:16–03:40]
- The survey has run for 15 years and is "a unique opportunity every year to get that temperature check from actual developers, from them directly" (Aaron Yeppes, [01:32]).
- Captures wide-ranging data on tools, languages, methodologies, and demographics.
- Majority (76%) of respondents are professional developers, but students and adjacent roles like PMs also participate.
- Geographic hotspots: US, Germany, India, UK, France, and Canada account for ~50% of respondents.
- Notably, 65% have more than 10 years of coding experience, showing an increasingly veteran respondent pool.
Quote:
"It was always a barometer... to see what was new and exciting, what people were thinking, and was super helpful just in terms of identifying trends as well as thinking about how I hire and create a place that people wanted to be a part of."
— Jody Bailey, [03:46]
Major Themes & Key Takeaways
1. AI Tools: Adoption, Trust, and Favorability
[05:43–11:49]
- Adoption skyrocketing: 84% of respondents use AI tools for work, up for the third consecutive year.
- Trust and favorability decline: Despite more usage, "trust took a dramatic dive this year...and favorability did as well" (Aaron Yeppes, [05:43]). More are using AI out of necessity.
- Why?
- Daily users rate AI tools higher; infrequent users are less satisfied, reflecting a steep learning curve.
- Younger and less-experienced developers are more favorable toward AI ("my bar was low already..." Aaron Yeppes, [07:49]).
- Mandated usage at companies may breed reluctance.
- High expectations make disappointment more likely; AI feels less magical as it becomes routine.
Quotes:
"More people are using these tools, but they don't like them, they trust them less."
— Aaron Yeppes, [06:08]
"Nobody likes to be told ... how to do their job. And especially developers...to have somebody tell you oh well, just generate a bunch of code and it'll be fine that it discounts the value of what developers bring to their job."
— Jody Bailey, [10:18]
2. AI in Developer Workflow: Practical Benefits and Pain Points
[11:49–14:21]
- On call support: AI is most effective for searching for answers when experts are unavailable.
- Reverse engineering & onboarding: AI is valuable for understanding/explaining complex code, especially for newer team members.
- Main pain point: 45% of developers cite debugging AI-generated code as their primary frustration.
3. Hype vs. Reality: Slow Adoption of “AI Agents”
[14:21–17:46]
- AI agents (autonomous code-writing tools) are hyped, but real-world adoption is still low.
- Barriers: unclear ROI, lack of time, steep learning curve, frequent changes in the agent landscape.
Quote:
"Building agents is not even a year old. Right. But it feels old already. Right. It reminds me of when JavaScript was taking off and there was felt like there was a new framework every week."
— Jody Bailey, [16:32]
4. Vibe Coding: Hype, Backlash, and the Future
[18:44–21:30]
- Definition: Using AI tools to write code with little domain knowledge (“AI does the heavy lifting, I just ‘vibe’ along”).
- 77% of developers don’t consider vibe coding part of professional work; it’s seen mainly as an enabler for hobbyists, side projects, or non-traditional coders.
- Backlash: Professional developers see it as a source of future tech debt; strong reactions in open-survey responses.
Quote:
"Your co-workers, your peers don't want to hear about you pushing this vibe coding idea because...that's just something I have to fix later."
— Aaron Yeppes, [21:00]
5. Other Significant Trends
[21:30–24:00]
- Cloud & DevOps tools surge: Docker, Kubernetes, and PostgreSQL see significant increases as they are being repurposed for AI and gaining ground as industry standards.
- New interests emerging: Tags like “large language models,” “MCP” (a new protocol), and “RAC” are rapidly gaining traction.
- Postgres for Everything: Industry sentiment is shifting toward consolidation around robust, mature technologies.
6. Survey Surprises & Community Feedback
[24:00–24:43]
- AI overload: Many respondents voiced frustration at the prevalence of AI-related questions; some “AI fatigue” is evident, echoing broader tech culture cycles of boom/bust interest.
7. Advice for Leaders: Productivity, Metrics & Team Happiness
[24:43–29:51]
- Experimentation over enforcement: Leaders should “give the developers opportunity to experiment with different tools in order to find what works best for them” (Jody Bailey, [25:26]).
- Focus on outcomes, not lines of code: Success metrics should measure business value or delivery, not code volume.
- Community matters: When using AI, developers still need help from peers and the larger community. Stack Overflow’s role as a knowledge hub will remain pivotal.
- Measurement challenge: No standard for measuring AI impact; DORA/Lean metrics are still the gold standard for delivery.
Quote:
"At the end of the day...what we all care about as leaders is delivering results and are we getting the results to users faster... That's really what matters, not how much code we generated by AI or how fast the code was written."
— Jody Bailey, [29:14]
8. The Future of Developer Communities & AI
[29:56–32:29]
- The role of Stack Overflow is evolving: With the rise of micro-communities (Discord, private groups, etc.), Stack Overflow is seeking ways to make Q&A more flexible and welcoming to “non-traditional” coders.
- Trust in AI: While tool reliability will improve, deep trust remains “a human emotion” that’s hard to automate.
9. Looking to 2030: Trust, Tools, and Developer Autonomy
[32:29–39:51]
- AI will get better but true trust will remain elusive—developers will continue to turn to each other for validation and creative work.
- Even as tools proliferate, developer satisfaction is not impacted by tool count but by autonomy and meaningful work:
- Job satisfaction has modestly improved (from 1 in 5 to 1 in 4 happy at work).
- Developers value autonomy over which tools and methods to use.
- Mandating tools without choice tends to erode satisfaction.
Quote:
"My hope is that they continue to love the craft, the art of creating new solutions via code or vibe coding or whatever the right phrase is... I suspect that the more we try to dictate you have to do it this way, it's going to create more and more challenges in terms of that happiness with the workforce."
— Jody Bailey, [38:19]
Notable Quotes & Moments (with Timestamps)
-
On using the survey as a hiring manager:
"It was always a barometer...was super helpful just in terms of identifying trends as well as thinking about how I hire and create a place that people wanted to be a part of."
— Jody Bailey, [03:46] -
On AI tool trust and favorability dropping:
"More people are using these tools, but they don't like them, they trust them less."
— Aaron Yeppes, [06:08] -
On forced AI adoption and developer pushback:
"Nobody likes to be told what to do or how to do their job. And especially developers...to have somebody tell you oh well, just generate a bunch of code and it'll be fine that it discounts the value of what developers bring."
— Jody Bailey, [10:18] -
On debugging AI-generated code:
"45% of developers indicated that debugging AI generated code was their top frustration with using AI tools."
— Aaron Yeppes, [21:00] -
On the proliferation of tools and developer satisfaction:
"I think developers like using a lot of tools...it's not the number of tools that's really dictating whether they're happy at work."
— Aaron Yeppes, [35:55] -
On job satisfaction and autonomy:
"My expectation, or maybe my hope is, you know, there's more. More developers are happy at work...the more we try to dictate you have to do it this way, it's going to create more and more challenges in terms of that happiness with the workforce."
— Jody Bailey, [38:19]
Timestamps for Key Segments
- What is the survey? [01:16–03:40]
- AI Adoption, Trust & Favorability [05:43–11:49]
- AI in Practical Developer Work [11:49–14:21]
- AI Agents & Hype [14:21–17:46]
- Vibe Coding & Its Backlash [18:44–21:30]
- Technology Trends (Cloud, DevOps, Postgres) [21:30–24:00]
- Survey Surprises/Feedback [24:00–24:43]
- Advice for Engineering Leaders, Metrics, Community [24:43–29:51]
- Evolving Role of Communities & Trust in AI [29:56–32:29]
- Looking Forward to 2030, Job Satisfaction, Developer Autonomy [32:29–39:51]
Conclusion
This conversation offers a nuanced look at the evolving lives of software developers in 2025. While AI's rapid rise is impossible to ignore, it brings mixed emotions—widespread adoption comes with skepticism, mandated tools, and the perennial struggle for trust. Developers remain pragmatic, seeking autonomy, a sense of craft, and community support. As Stack Overflow continues to adapt to these rapid changes, it aims to serve both traditional and non-traditional coders, helping the community collectively work through the emergent challenges and opportunities of AI in software engineering.
