Is Technology Making Life Easier or More Complicated?
84% of developers use AI tools, yet 46% distrust their accuracy. A 2025 trial found a 19% slowdown while developers felt faster. Here is how to measure it yourself.
6 min read · Last updated: October 25, 2026
Technology makes individual tasks easier and the overall job more complicated. For developers, the data shows the same split: 84% use AI tools, yet more distrust their accuracy than trust it, and a 2025 trial found experienced developers worked 19% slower with AI while believing they were faster.
Came here from Instagram, LinkedIn, X or YouTube? Here is the full source list you came for. Everything below is built from these, nothing else.
The post that started the question
The data
A Medium essay titled "Technology: Making Life Easier or More Complicated?" is making the rounds again. I could not open it (Medium returned an access error to my fetch), so I will not summarise its arguments or quote it. I will answer its question with data I did open.
That is the useful move anyway. "Easier or more complicated" is an opinion fight until you put numbers on it. Developers are the best test group, because we adopt new tools first and measure the results most openly.
Key takeaways
- According to the Stack Overflow Developer Survey 2025, 84% of respondents use or plan to use AI tools, up from 76% the year before.
- The same survey found 46% distrust the accuracy of AI tools and 33% trust it. Only 3% highly trust it.
- 66% named "AI solutions that are almost right, but not quite" as a frustration, and 45.2% said debugging AI-generated code takes more time.
- METR's July 2025 trial of 16 experienced open-source developers across 246 real issues measured a 19% slowdown with AI allowed.
- Those developers had expected a 24% speedup and still believed AI helped by 20% afterwards.
- Easier tasks plus harder judgement is the pattern. The skill that matters now is checking, not producing.
So, is technology making life easier or more complicated?
Both, at different layers. Tools reduce the effort of a single task, such as writing a function or sending a message. They also add decisions, tools to learn, outputs to verify and notifications to manage. The total effect depends on whether you spend the saved time or lose it to the new overhead.
The developer data below is the cleanest evidence I have found that this is not just a feeling.
What happened with developers and AI tools?
Adoption is high and confidence is not. According to the Stack Overflow Developer Survey 2025, 84% of respondents use or plan to use AI tools, and 47.1% of all respondents use them daily. In the same survey, positive sentiment fell to about 60%, down from over 70% in 2023 and 2024.
| Measure | Result | Source |
|---|---|---|
| Use or plan to use AI tools | 84% (76% a year earlier) | Stack Overflow 2025 |
| Use AI tools daily | 47.1% of all respondents, 51% of professionals | Stack Overflow 2025 |
| Favourable view of AI tools | About 60% | Stack Overflow 2025 |
| Distrust accuracy / trust accuracy | 46% / 33% | Stack Overflow 2025 |
| Highly distrust (experienced developers) | 20% | Stack Overflow 2025 |
| "Almost right, but not quite" | 66% | Stack Overflow 2025 |
| Debugging AI code takes longer | 45.2% | Stack Overflow 2025 |
The second data point is a controlled trial. METR, a nonprofit that evaluates AI systems, published a study on July 10, 2025. It followed 16 experienced open-source developers on 246 real issues, each randomly assigned to allow or disallow AI.
Developers expected AI to speed them up by 24%. With AI allowed, they took 19% longer. After the study they still believed AI had sped them up by 20%.
METR notes that these results are dated. It has since published updated data, in February 2026, on late-2025 tools. Treat 19% as a snapshot of one group on one kind of work, not a law.
Why does this matter to a developer?
Because the gap between how productive a tool feels and how productive it is can be wide, and you cannot see it from the inside. In the METR trial, developers misjudged their own speed by a large margin, in the optimistic direction, even after finishing the work.
Three things follow from that:
- Your gut is not a measurement. Feeling fast is cheap. Shipping fast is not the same thing.
- Verification is now the work. When 66% of developers say AI output is almost right, the cost sits in finding the part that is wrong.
- Experience changes the maths. In the Stack Overflow survey, experienced developers had the highest share of high distrust (20%) and the lowest share of high trust (2.6%). The people who can spot the wrong part are the most careful.
I am not arguing against the tools. I use them. The point is that "easier" and "more complicated" are both true, and only one of them shows up in how the work feels.
What changes in practice?
Your day shifts from writing code to reading it. The tool drafts, and you review, test and decide. That is a different skill, and it rewards people who understand the system underneath.
| Before | Now |
|---|---|
| Write the function, then test it | Review a generated function, then test it |
| Stuck means searching docs | Stuck means checking whether the answer is plausible but wrong |
| Speed came from typing less | Speed comes from rejecting bad output early |
| Learning came from struggling | Learning needs a deliberate choice to struggle |
The last row is the quiet one. If a tool removes the struggle, it can also remove the learning. I wrote about that trade in Is AI making you smarter or just more dependent?, and about the other side in how to use AI to learn programming 3x faster.
The same logic applies outside code. More devices means more interruptions. If the phone is the source of your complication, the 5-step system that fixed it for me is a practical place to start.
What should you do this week?
Run your own small measurement. That is the only way to know which side of the line a tool sits on for you.
- Pick one recurring task, such as writing tests or fixing small bugs.
- Time five instances with AI and five without. Include the time you spend checking and fixing the output.
- Write down your guess first. Compare it with the real numbers. The METR result suggests the guess may be off.
- Keep a "wrong output" list. Note each time the tool was almost right. You will learn where to trust it and where not to.
- Protect one thing you do by hand. Choose a skill you want to keep, and do it without help once a week.
- Turn off one notification source. Remove one app's alerts and see whether you miss it.
Five samples is not a study. It is enough to tell you whether your feeling matches your clock.
Does technology save time or just create new work?
It does both. A tool saves time on the task it targets and adds time on setup, review and maintenance. The Stack Overflow figure that 45.2% of respondents find debugging AI code more time-consuming is one visible example of that added cost.
Is AI going to replace developers because of this?
The data above does not say that. It says the work is changing toward review and judgement. For a fuller view, read our honest answer on whether AI will replace developers and the case that we are being sold an unfinished product.
FAQ
Is technology making life easier or more complicated?
Both. Technology lowers the effort of single tasks and raises the number of decisions, tools and outputs to manage. Whether you come out ahead depends on measuring your own results instead of trusting how fast the work feels.
Do developers trust AI tools?
Not fully. According to the Stack Overflow Developer Survey 2025, 46% of developers distrust the accuracy of AI tools and 33% trust it. Only 3% say they highly trust it, and experienced developers show the highest rate of high distrust at 20%.
Do AI coding tools make developers faster?
It depends on the developer and the task. In METR's July 2025 trial, 16 experienced open-source developers took 19% longer with AI on 246 real issues, though they believed they were 20% faster. METR has since published updated data on newer tools in February 2026.
How can I tell if a tool is helping me?
Time the same kind of task with and without it, and include the checking and fixing time. Write your prediction first. Compare the two sets of numbers after about five tries each.
Pick one tool you rely on and time it this week, then try the free tools on Art of Code or browse the latest developer jobs to see which skills employers ask for.
Written by Adithya Guttha, Founder of Art of Code.
Sign in to like and save posts.
Keep reading
Discussion
Comments (0)
Want to join the conversation?
Sign in to commentNo comments yet. Be the first to comment!