Daniel Lemire 提出 Ephemeral Testing：通过 AI 临时构建应用层评估核心代码质量
Ephemeral Testing
Daniel Lemire 提出 Ephemeral testing（瞬时测试）方法：让 AI 智能体在核心代码之上临时构建应用层进行集成测试，通过评估这些临时上层的可用性间接反映原始代码的 API 设计、错误提示与文档质量。不同 AI 智能体和任务可对同一基础库重复测试，作者表示该方法已在多个项目中应用。
Daniel Lemire is a software performance expert. He ranks among the top 2% of scientists globally (Stanford/Elsevier 2025) and is one of GitHub's top 1000 most followed developers.
Menu and widgets
Your business needs help? Get in touch, I offer private talks, training, consulting and I do sponsored open-source projects.
Support my work!
I do not accept any advertisement. However, you can sponsor my open-source work on GitHub.
Daniel Lemire started this blog in 2004. It contains 2,410 posts and 16,399 approved comments.
Daniel Lemire's blog ranks among the top 50 most popular blogs on Hacker News, a leading tech news aggregation platform.
Join over 12,500 email subscribers:
Search for:
Recent Posts
- Ephemeral testing
- Research paper overload: submissions capped at two a month
- Google’s first orbital data center is in orbit
- Parsing compressed JSON at 40 GB/s
- Transcoding UTF-8 to UTF-16 with replacement at gigabytes per second
Recent Comments
- Sebastian Good on Ephemeral testing
- Simon on Google’s first orbital data center is in orbit
- reddit on How many strings can you create per second?
- Ivan Bobev on How do you deal with people who strongly disagree with your views?
- Victor on How do you deal with people who strongly disagree with your views?
Pages
- A short history of technology
- About me
- Book recommendations
- Cognitive biases
- Interviews and talks
- My bets
- My favorite articles
- My favorite quotes
- My rules
- Newsletter
- Predictions
- Privacy Policy
- Recommended video games
- Terms of use
- Write good papers
Archives
Archives
Boring stuff
Ephemeral testing
5 October 2026 · 2 min
We have many ways to ensure software quality. Unit testing. Fuzz testing. Integration testing. And so forth.
I’d like to propose a method that was unthinkable before: ephemeral testing. (Ephemeral is a fancy word for ‘throw away’ or ‘temporary’.)
You write your code. You build your software component. Or the AI agent does it for you, it does not matter.
Then you ask an AI agent to build on it: an application, another layer, maybe several. You have it test what it built. You do not assess the original work directly. You assess how good the software built on top of it is.
It is a form of integration testing. The difference is that the software on top is entirely ephemeral. You throw it away when you are done.
A library with a clean API, stable invariants, and useful errors lets the agent produce something that works quickly. A library with hidden state, surprising defaults, or incomplete docs produces a pile of patches and failures. The failures are evidence about your code, not about the agent.
You can repeat it. Different agents, different tasks, same foundation.
In effect, instead of building the core while trying to anticipate what might be needed at the other layers, you just simulate the other layers by actually building them.
Of course, you could argue that with AI, you can rebuild everything whenever you need to. But that’s not practical. You need some form of stability.
I have been applying this trick to various projects. As I consider a new feature, I ask my AI to prototype quickly what I might later build based on what I am doing it. Ephemeral testing works for me thus far.
Daniel Lemire, "Ephemeral testing," in Daniel Lemire's blog, October 5, 2026, https://lemire.me/blog/2026/10/05/ephemeral-testing/.
[BibTeX]
Published by
Daniel Lemire
A computer science professor at the University of Quebec (TELUQ). View all posts by Daniel Lemire
Posted on October 5, 2026 October 4, 2026Author Daniel Lemire
One thought on “Ephemeral testing”
Sebastian Goodsays: October 5, 2026 at 1:00 pm I think this is at the moment one of the best ways to get velocity from AI programming. While it is very good at writing lots of code, it needs guidance on objective and modularity. It doesn’t seem to be good yet at saying “I will build this with a compiler pattern” or “at its heart, this is a distributed graph” or “FORTRAN-style vector computation”. But if you tell it, it often does a good job. Stacking software on top of your library forces these abstractions. Reply
Leave a Reply Cancel reply
Your email address will not be published.
Comment *
Name *
Email *
Website
- Save my name, email, and website in this browser for the next time I comment.
Δ
You can also subscribe by email to this blog (non-commercial, no ads, weekly email).
How to post code (C, C++, Java, Python, etc.):
Wrap your code in backticks, like this:
int main() { return 0; }
Post navigation
Previous Previous post:Research paper overload: submissions capped at two a month
来源：Hacker News · lemire.me