burtenshaw commited on
Commit ·
00186a7
1
Parent(s): 46cc191
docs: update article content
Browse files
app/src/content/article.mdx
CHANGED
|
@@ -36,14 +36,12 @@ import codeAgentSlop from './assets/image/code-agent-slop.png';
|
|
| 36 |
|
| 37 |
In March 2026, a `transformers` maintainer posted a screenshot in Slack. The Transformers repository was receiving more pull requests per day than a human team could review. Most of them came from AI coding agents. The maintainer's message was blunt: "I think we need to do something drastic."
|
| 38 |
|
|
|
|
|
|
|
| 39 |
This problem is not unique to Transformers. Every large open source project on GitHub has been dealing with the same thing. Coding agents like Codex, Claude Code, and their wrappers make it trivial for anyone to fork a repository, point an agent at an issue, and open a pull request in seconds. Sometimes even without direct instruction from humans.
|
| 40 |
|
| 41 |
The volume is unprecedented and unfortunately the quality, on average, is low. And the maintainers reviewing these contributions are the same teams they have always been. Honest folk.
|
| 42 |
|
| 43 |
-
<Image src={honestWork} alt="Meme: I have 2 PhDs in computer science and now I read through 500 AI generated PRs an hour. It ain't much, but it's honest work." />
|
| 44 |
-
|
| 45 |
-
<Image src={codeAgentSlop} alt="Screenshot of the huggingface/transformers pull request list showing dozens of PRs labeled 'Code agent slop', most closed within minutes of each other." />
|
| 46 |
-
|
| 47 |
To understand this, we started a project at Hugging Face to study this problem on Transformers. We didn't set out to solve it, but we wanted to understand what was coming, whether any of it was useful, and whether there was a way to extract signal from the noise.
|
| 48 |
|
| 49 |
Ultimately, we don't want to block or discourage people from contributing to transformers, even if the code is sub par, we pride ourselves on educating our community. And we want to be ready for the next era of AI development, with the same culture that has brought us this far.
|
|
|
|
| 36 |
|
| 37 |
In March 2026, a `transformers` maintainer posted a screenshot in Slack. The Transformers repository was receiving more pull requests per day than a human team could review. Most of them came from AI coding agents. The maintainer's message was blunt: "I think we need to do something drastic."
|
| 38 |
|
| 39 |
+
<Image src={codeAgentSlop} alt="Screenshot of the huggingface/transformers pull request list showing dozens of PRs labeled 'Code agent slop', most closed within minutes of each other." />
|
| 40 |
+
|
| 41 |
This problem is not unique to Transformers. Every large open source project on GitHub has been dealing with the same thing. Coding agents like Codex, Claude Code, and their wrappers make it trivial for anyone to fork a repository, point an agent at an issue, and open a pull request in seconds. Sometimes even without direct instruction from humans.
|
| 42 |
|
| 43 |
The volume is unprecedented and unfortunately the quality, on average, is low. And the maintainers reviewing these contributions are the same teams they have always been. Honest folk.
|
| 44 |
|
|
|
|
|
|
|
|
|
|
|
|
|
| 45 |
To understand this, we started a project at Hugging Face to study this problem on Transformers. We didn't set out to solve it, but we wanted to understand what was coming, whether any of it was useful, and whether there was a way to extract signal from the noise.
|
| 46 |
|
| 47 |
Ultimately, we don't want to block or discourage people from contributing to transformers, even if the code is sub par, we pride ourselves on educating our community. And we want to be ready for the next era of AI development, with the same culture that has brought us this far.
|