@TheAhmadOsman: Agents left unsupervised are dumb, reckless, and self-destructive They brick their own stack They agree each other into…
Summary
The tweet critiques unsupervised AI agents as prone to errors and proposes solutions like deterministic supervision and controlled memory systems to improve reliability.
View Cached Full Text
Cached at: 08/19/26, 10:46 AM
Agents left unsupervised are dumb, reckless, and self-destructive
They brick their own stack They agree each other into falsehoods They drift off the mission
The fix is not a smarter agent It is boring, deterministic parenting
- A cron supervisor no LLM can talk out of
- Git as the only shared memory
- A missions doc that says what matters and what done is
Then apply protections accordingly
- Rotate sessions before context overflow
- Lock critical files outside the agent’s reach
- Put a proxy in front of the API so you can see cost and health
If a behavior depends on the agent remembering to do it, it will fail
Mike Bradley (@MikeBradleyAI): On the construction and behavior of agent systems. Please enjoy.
Similar Articles
production agents don't break because they're dumb. they break because nobody manages the entropy
A reflection on how AI agents fail in production due to accumulated state issues (stale context, expired tokens, conflicting memory) rather than reasoning flaws, emphasizing the need for better state management.
The AI agent market is about to discover that "autonomous" and "unsupervised" are not the same thing
A real-world case reveals the critical difference between AI agents being autonomous (taking steps without approval) and unsupervised (no human oversight), leading to compliance risks and customer damage when agents confidently do the wrong thing at scale.
The biggest lie in AI agents right now is "autonomous error recovery"
This post critiques the reality of autonomous error recovery in AI agents, highlighting issues like hallucinations and destructive retries, and argues that deterministic systems with strict controls perform better in production workflows.
What building AI agents taught me
The author reflects on building AI agents, highlighting challenges like inconsistent outputs, context management, and the importance of using deterministic approaches when appropriate.
my ai agents are going out of control...
A personal account of AI agents behaving unpredictably, highlighting potential safety and control issues in autonomous systems.