Why satsuki

[Deep Sea Diver]

Wow folks! There’s all that and we seem to now be getting a bit off point.

There is no doubt that many misconceptions and deliberate obfuscations have pervaded the evergreen azaleas world once these marvelous plants were imported into the Western World.

For example western collectors brought in azaleas from Japan and renamed them so these might appeal better to Westerners… and sometimes renamed them again! Much to the dismay of hobbyists even now.

One example is Kazan, a satsuki flower sport of Kozan. It was renamed Rukizon Go figure! Why would Rukizon appeal more?

In fact, I was advising a local nursery who wanted to develop a satsuki collection on prospective new purchases from Van Veen nursery. Van Veen’s listing still lists Rukizon vs Kazan…. Stuck!

This is a really good thread folks could use as a reference over years if we can work together to continue to develop the best data possible.

Thanks.

As an aside, I would like to mention that @Glaucus new azalea database is well on the way to becoming the best azalea knowledge center in the Western world.

It’s at:

Pictures, descriptions and parentage of azalea varieties

A database with over 1500 evergreen azalea varieties, including photos and parentage. Important cultivar groups like satsuki, Kurume, obtusum, Hirado, Belgian indica and American varieties are listed. Wildtype species are also included.

azaleadatabase.com

Cheers
DSD sends

 

[shohin_branches]

Sspoken like a true Luddite. You may think it’s wrong but you’re in a shrinking minority. I known how AI works. I have been writing about tech for 35 years.

Just pointing out that your crusade to “correct” what has become ingrained in so many places is a bit futile. Go right ahead and tilt at that windmill Don.

I'm a software test engineering manager and LLMs are wildly inaccurate. We also have a common saying in software engineering and that is "garbage in garbage out." AI contamination or pollution is making models even less accurate because it's learning off of inaccurate information created by other LLMs. Best practice is to use datasets up to 2022 to avoid AI pollution when training. There are prompts that worked four months ago that don't even get close to working now. I use an enterprise license of copilot everyday for development work. Sometimes it's good sometimes it's garbage. Always confirm information coming from LLMs with a trusted source before utilizing it. ChatGPT even has a disclaimer right under the prompt field: "ChatGPT can make mistakes. Check important info."

It doesn't really understand plants and I see it spitting out a lot of inaccurate and sometimes dangerous plant info (I got an associates in horticulture before going back to school for computer science)

People calling out that LLMs are inaccurate doesn't make them luddites. Let's stay away from name calling and personal attacks or assumptions on this site and focus on the discussion topic.

 

[snox]

Just toilet folk know have two young satsukis on the way and sure will have more questions along the way . Thanks for the replies and info .