Remix clone Hacker News

It's interesting they mention charcoal purification, there is this plant that Japanese people eat (like wild green tubes with leaves) and they soak it in water with wood ash from a fire.

▲

vasusen a day ago | parent [-]

I believe (more likely, chatgpt believes) this plant is Warabi. I didn't know it was processed with charcoal purification!

The process sounds similar to how I use something acidic to process the toxins from Taro leaves before eating them.

▲

ge96 a day ago | parent | next [-]

Yeah that looks right, long tubes. Another one I see a guy foraging/eating is a big leaf called "strawberry something" and he batters it up/fries it, interesting.

▲

bigiain a day ago | parent | prev [-]

> I believe (more likely, chatgpt believes)

Possibly the most depressing thing I've read all week... :sigh:

▲

vasusen a day ago | parent [-]

Sorry instead of "googling" this plant up, I used ChatGPT to tell me candidates on what this plant might be.

▲

esperent 21 hours ago | parent [-]

Just don't tell people where you got the info until it's normalized to ask AIs.

Honestly, while LLMs hallucinate, they still return more reliable info than using Google and getting the info from an SEO spam blog post that played Chinese whispers with already wrong info from ten other blogs.

The only difference is that using Google is normalized, so you wouldn't have felt the need to announce it.

Of course, unless you personally feel the info is reliable, you probably shouldn't post it no matter where you got it from.

	▲	bigiain 20 hours ago \| parent \| next [-]
		> getting the info from an SEO spam blog post that played Chinese whispers with already wrong info from ten other blogs. Didn't you just perfectly describe the ChatGPT training set?
	▲	foobarian 20 hours ago \| parent \| prev [-]
		It's like a hash table. You get a result quick, but you still have to check that the key is actually equal to what you asked for.