Water company wasted $200k on bad answers from an AI so built slop filtering (opens in new tab)

(theregister.com)

11 pointsbeardyw7d ago3 comments

3 comments

> The tool passes answers through a verification layer designed to detect and correct errors and hallucinations

Call us back when the verification layer is verified to detect errors and hallucinations.

and when the verification layer verifier is verified

“Not being PhDs in the space, we read relevant academic papers and used LLMs like Grok and ChatGPT to validate our findings…”

So exactly the wrong way to use LLMS? Cool story bro.

j / k navigate · click thread line to collapse

> The tool passes answers through a verification layer designed to detect and correct errors and hallucinations

Call us back when the verification layer is verified to detect errors and hallucinations.

and when the verification layer verifier is verified

“Not being PhDs in the space, we read relevant academic papers and used LLMs like Grok and ChatGPT to validate our findings…”

So exactly the wrong way to use LLMS? Cool story bro.

j / k navigate · click thread line to collapse