The Weekly Update

[HereOnThePrimalEdge]

Hanakaoʻo Park. It was definitely in that area.

I think that would make the island in the photo Lanai.  Molokai is north, to the right but out of frame of the photo.  Nice shot!

Great photo of the east side cliffs in Puna wahine!   Something to get up for in the morning!

[TomK]

HOTPE - yes, the island is definitely Lanai.

PT - Whenever I hear that song I think of the movie "The Dish". If you haven't seen it I highly recommend doing so. It's a brilliant movie and you will love the soundtrack!

Wao nahele wahine - thanks for posting that. I think you were in almost the same spot as where I took the following picture, although mine was just before sunrise, so I was able to use some long-exposure techniques.  I've posted it before, but your picture reminded me of it!

[TomK]

Having never before taken the advantage/time to look at AI imagery before I wonder.. do you all tell the thing to create images that look that way or is this an AI thing? They all look the same. They all have a certain quality.. cartoonish.. but cartoonish in a certain way.. that really isn’t pleasant. In fact I find it unpleasant. Seriously, if this is the “state of the art” I’d say it’s a failure. Kind of like a trailer park compared to a housing development. Trash.

And yeah, I get it, there’s a lot of trailer trash out there. Just look at the love fest over on the Bicker Board the last few days.. that’s intellectual traitor trash. Look at the internet of the early days, all that crap folks adorned their sites with.. that’s trailer trash. People stooping to some school yard throwing sand in each other’s face level of existence.. no expectations just acceptance of whatever is being shoved in front of them. Commercialization for animals.

Anyways, it’s another beautiful morning on a dying planet filled with trash.. I am sure there’s some species somewhere that would thank you for your contributions.. if it wasn’t being trashed to death.

You just open an AI app, tell it to create a picture describing what you want, and ask it to change things if the first picture isn't what you want. You may not like the pictures here, but I doubt anyone cares. I lost count of how many times you mentioned trailer trash, but obviously, that's some connection you have made with AI. After that, you claim the planet is dying, and then you tell someone not to forget their antibiotics. I don't know what you're thinking, but it's not a good look.

[kalianna]

Although this article is not about Hawaii, I think it can be useful for us here since many of us seem unaware of what all our quick AI cartoons cost in terms of water, land, electricity and the effects on farmers.

https://www.npr.org/2026/06/12/nx-s1-578...ico-desert

[MyManao]

I lost count of how many times you mentioned trailer trash..

If the shoe fits.. and omg, Tom, you're the quintessential example of.. they look perfect on you.

I am impressed how quickly the world has traded their concern for the environment for their love of all things AI. Fools, every last one of them.. selfish narcissistic idiots.. of which the pics posted here are perfect examples..

And no, the planet has not stopped warming, and every one of those data centers you all relish like children on Christmas morning, besides using an exorbitant amount of water and electricity, are emitting a phenomenal amount of heat. Which of course, is the real problem.. heat.

From.. https://thebulletin.org/2026/06/an-area-...ures-rise/

An area of sea ice as big as France is gone from Antarctica’s west coast, as temperatures rise

Which begins..

Antarctica’s west coast is missing an area of winter sea ice the size of France, sparking concerns for threatened penguins, other marine life, and global sea levels.

One expert said the loss of ice in the Bellingshausen Sea was “depressing” and the failure of ice to form could have intensified a heatwave over the continent’s peninsular last week that saw daytime temperatures peak at 15.4 degrees Celsius (about 59 degrees Fahrenheit) which is more than 20 degrees Celsius above average. (It is normally about 24 degrees Fahrenheit at this time of year.)

It’s winter in Antarctica, when sea ice expands rapidly around the continent peaking in September.

But satellite observations showed the Bellingshausen Sea—on the west side of the Antarctic peninsular and which by June would usually be covered by ice—was almost completely ice-free.

Scientists said the region was missing about 650,000 square kilometers (250,000 sq miles) of sea ice, compared with the average between 1991 and 2020. That is an area about the size of France, and almost 10 times the size of Tasmania.

Kalianna's shared article doesn't come close to pointing out how sick our species is.. and how detrimental to the planet's well-being man's desires for self gratification are. But yeah, this thread is a classic. Look at me.. I'm a monkey too! And no, Q, just because the planet will 'recover' doesn't justify the sin of our willingness to destroy life. Of our happiness gained on the traumatic demise of others. A sin is still a sin.

[Punatang]

You better stop streaming movies then 'Mao. This article's agenda is downplaying the effects of streaming and it still shows how much more resource intensive it is.  You're killing polar bears!

Based on your last few posts, you should apply for a religious exemption from using AI.

https://tnfarmbureau.org/transforming-ag...of-farming

https://andthewest.stanford.edu/2025/thi...-the-west/

https://futurism.com/artificial-intellig...on-ai-work

[kalianna]

Streaming movies is an entirely different process. You are accessing one file. AI searches through every image that has ever been uploaded to the internet to decide what the silly cartoon is going to look like.

[Punatang]

Streaming movies is an entirely different process.  You are accessing one file.  AI searches through every image that has ever been uploaded to the internet to decide what the silly cartoon is going to look like.

We are talking about resource consumption of one vs the other.  Yes, it is a different process but they both utilize data centers and use the same resources.  Which one uses more to do what it is doing?  That is the issue.

ETA:  Grok says you are wrong:

   


No, this is not true.Modern AI image generators (like DALL·E, Midjourney, Stable Diffusion, Grok's image tools, etc.) do not search through every image (or even any images) on the internet when generating a new "silly cartoon" or anything else. Here's how it actually works:Training Phase (Done Once, Long Ago)

• The model is trained on a massive but fixed dataset of millions or billions of images + text captions scraped from the internet (LAION-5B for many open models, proprietary datasets for others).
  
• During training, the AI (usually a diffusion model) learns statistical patterns: what shapes, colors, styles, compositions, and concepts tend to go together. It doesn't store or memorize individual images like a database.
  ibm.com
  
• It learns things like "cartoons often have bold outlines, exaggerated features, and vibrant colors" or "a silly cartoon cat might have big eyes and a goofy expression."
  

This training is computationally expensive and happens before the model is deployed for public use.Generation Phase (What Happens When You Prompt It)

• You type a prompt ("a silly cartoon of a dancing pineapple").
  
• The model starts with pure random noise (static, like TV snow).
  
• It then iteratively denoises that noise over dozens of steps, guided by your text prompt. At each step, it predicts what a slightly less noisy version should look like based on the patterns it learned during training.
  altexsoft.com
  
• The result is a brand-new image synthesized from those learned patterns — not a collage, copy, or direct retrieval of existing images.
  

This process is called inference. The model is not looking anything up in real time.
cloud.google.com
Common Misconceptions

• It feels like it's "searching" because the output often resembles styles or elements from its training data (which came from the internet).
  
• Some older systems or niche tools might use retrieval-augmented generation (pulling reference images), but mainstream generators like the ones people use daily do not.
  
• No live internet access is involved for standard image generation.
  

Bottom line: The AI isn't browsing the web or scanning uploaded images on demand. It's applying compressed knowledge from its earlier training to create something new from noise. This is why the same prompt can produce different results each time (due to randomness) and why outputs can sometimes include unintended elements or "hallucinations" from imperfect pattern learning.

[HereOnThePrimalEdge]

ETA:  Grok says you are wrong:

Yes.
Let’s ask AI whether AI uses more power than other services.  That should provide a totally objective answer. And which AI better to ask than an AI service connected with a recent IPO, the share prices dependent on potential buyers believing they are investing in a benign, even beneficial robot overload and it’s trillionaire Obergruppenführer?

[Punatang]

Still haven't dehydrated those grapes?