OpenTelemetry (OTEL) has evolved to cover all of observability, providing a stable standard and SDKs for metrics, logs, and traces.
OTEL is now the second most active project in the CNCF, showing widespread adoption among observability providers.
Key sub-projects of OTEL include specifications, implementations, the OpenTelemetry Protocol, the OpenTelemetry Collector, and the Open Agent Management Protocol.
The most reasonable strategy to assess if AI models are deceptively aligned is to test their capability; incompetent models are less likely to be deceptively aligned.
By using capability evaluations, models tend to fall into categories of untrusted smart models and trusted dumb models.
Combining dumb trusted models with limited human oversight can help mitigate the risks posed by untrusted smart models.