Honestly, it's empirical. We started with what was easiest to measure: human correction rate. If I had to step in and fix something, that's a clear signal the agent took a bad path.
Iterations and reverts turned out to be noisier -- sometimes high iteration count means the task was genuinely hard, not that the agent made a mistake. So we downweighted those.
The meta-answer is: pick the metric that most directly captures "I wish the agent hadn't done that." For us that's human intervention. For a team with better test coverage, it might be test failures after commit. For infra work, maybe rollback frequency.
There's no universal loss function — it depends on where your pain actually is. We just made it explicit and started logging it. The logging alone forced clarity.
We hit something similar building document processing agents. What helped was treating memory as typed knowledge rather than flat text.
Three buckets: (1) constraints -- hard rules that should always apply, (2) decisions -- past choices with context on why, (3) heuristics -- soft preferences that can be overridden.
Retrieval then becomes: constraints always injected, decisions pulled by similarity to current task, heuristics only when ambiguity is high. Still experimenting with how to detect "ambiguity" reliably -- right now it's a cheap classifier on the task description.
The deduplication problem is real. We ended up hashing on (topic, decision_type) and forcing manual review when collision detected. Brutal but necessary. TBH we are not home yet but that's the path we walking.
Interesting framing, but I think emotions are a proxy for something more tractable: loss functions over time.
Engineers remember bad hygiene because they've felt the cost. You can approximate this for agents by logging friction: how many iterations did a task take, how many reverts, how much human correction. Then weight memory retrieval by past-friction-on-similar-tasks.
It's crude, but it lets the agent "learn" that certain shortcuts are expensive without needing emotions. The hard part is defining similarity well enough that the signal transfers.
Still early, but directionally this has reduced repeat mistakes in our pipeline more than static rules did.
Why am I not surprised? We’re starting 2025 just as we ended 2024—with masks dropping everywhere. From Big Oil slashing investments in green energy, to U.S. banks walking back on sustainability-linked loans, to the EU's push for an omnibus that could weaken already fragile regulations. And let’s not even get started on carbon markets turning into a growing scam.
Maybe it’s a good thing? Clarity can be painful, but it’s better than living in illusions. Time to take the reins ourselves—invest consciously, act decisively, and demand accountability.
This is my wish for all of us in 2025: that we see things as they are, not as we wish they were, and have the courage to drive the change we need.
Apologies if this sounds overly philosophical—blame it on the rainy Brussels weather. It does things to a person!