adding a deterministic retrieval layer made model choice much less important
大多数人认为在AI应用中，选择更强大的模型是提高准确性的关键，但作者认为添加确定性检索层比模型选择更重要。这一反直觉观点表明，在生物数据处理领域，基础设施的改进可能比模型升级更能解决问题，这与AI领域普遍追求更强大模型的趋势相悖。
6 Matching Annotations
- Jun 2026
-
www.anthropic.com www.anthropic.com
-
- Apr 2026
-
-
Non-deterministic software breaks the contract.
传统软件的确定性承诺被AI的非确定性打破，这是当前AI产品体验“滑溜”的根源。作者敏锐地指出，这并非单纯的模型缺陷，而是设计契约的失效。我们需要用界面设计来重塑这种可靠性，为不可控的输出建立护栏。
-
- Nov 2022
-
stackoverflow.com stackoverflow.com
-
In general, I've found Selenium to be unreliable and not deterministic as laid out here although I still don't entirely know why.
-
- Feb 2021
-
www.quora.com www.quora.com
-
The non-determinism is why cache invalidation — and that other hard problem, naming things — are uniquely and intractably hard problems in computer science. Computers can perfectly solve deterministic problems. But they can’t predict when to invalidate a cache because, ultimately, we, the humans who design and build computational processes, can’t agree on when a cache needs to be invalidated.
-
- Oct 2020
-
medium.com medium.com
- Aug 2015
-
-
deterministic
Is appropriate the use of a deterministic model here? It seems to me that it will be less accurate than a probabilistic one, and that that could influence the results of the comparison of the different aggregation levels biasing the conclusions.
-