12 Comments
User's avatar
David Parker's avatar

The eval section is the one I keep. A demo with vibes attached is exactly it.

We wrote a mutation audit that breaks each of our safety checks on purpose and re-runs it. Three of ten guards stayed green after I broke the thing they were guarding.

The dashboard would have called that healthy. The test that could have proven us wrong did not.

A green tick on a check you never tried to kill is a decoration.

StrongOp's avatar

I am impressed on how good this article is. Thanks Miguel!

Miguel Otero Pedrido's avatar

thanks to you for the support!

Lou's avatar

I think after reading this article I will take the class

Sayani Ghosh's avatar

Amazing article, this is worth an office hour Miguel!! Really good stuff, lots to learn and think about.

Amin Ghaderi's avatar

Mapping to an old article, wow. Thanks.

Pavan Belagatti's avatar

A true agentic system that has all the components mentioned clearly.

Logan Thorneloe's avatar

I like this comparison! The original paper was written by my team and it’s amazing how relevant it continues to be.

Servas Adolph's avatar

Amazing, easy to understand thanks Miguel for a good article 🔥🔥🔥

Miguel Otero Pedrido's avatar

Glad you like it!!

Pawel Jozefiak's avatar

The 2015 paper reference lands. The glue-code problem in MLE shows up in agent systems as a different shape but same disease: every workaround you add to handle a flaky tool becomes a thing the next run has to read and reason about.

For my setup the worst version of this was correction prompts accumulating in the system message. I had to build an explicit consolidation pass to drain ad-hoc instructions into stable rules, otherwise context bloat compounded faster than capability. Looking forward to the rest of the issue.