Hacker News new | ask | show | jobs
by larubbio 686 days ago
I remember this article from 2017 about the google snippet for how long it takes to caramelize an onion.

https://gizmodo.com/googles-algorithm-is-lying-to-you-about-...

The gist of the story is the author wrote an article debunking the recipe myth that you only need 5 minutes. In their tests it was 25 - 45. Google snippet would report "about 5 minutes" and link to the article as a reference.

I think the information summaries at the top of searches have been wrong for a while.
2 comments
daedrdev 686 days ago
I think I actually ran into that exact search issue and was confused why my onions were not cooking well, lmao
link
pradn 686 days ago
It's certainly wrong, but its not out of place w/ the sheer number of recipes and blogs saying the same thing. It's parroting "common wisdom" which is wrong.

It's not as egregious as some of the generative AI results spewed out by the lowest-cost-inference LLMs they hooked up to Google search a few months back.

link
nox101 686 days ago
It was still a common occurrence for google to display incorrect results that weren't common wisdom.

I had tons of such experiences. One I remember was searching for "healthy body weight for x of age y" where x was man,woman and y was my age. Google said 50lbs higher than I expected. I clicked through to the article. The article said the average was this high number, not a healthy weight.

link
pradn 686 days ago
Yes, totally. Automated, fallible systems.

Ultimately it's a cost issue. If they wanted to, they could fact check the top million question queries - but they don't want to spend on that.

link
kevindamm 686 days ago
How familiar are you with the knowledge graph update process? There is actually a significant amount of spend going into reviewing the wrong answers to top queries and updating to fix/tweak them.
link
pradn 683 days ago
Yes, there's humans in the loop for the knowledge graph. I'm unsure if snippets are reviewed in the same way.
link
refulgentis 686 days ago
Your assumption basis is completely wrong, not only on when it came about. Bizarrely, Google spelled out that the AI answers weren't from an LLM in post-debacle PR, like it was a good thing
link