A.I. is B.S.

Now that I have a more modern machine, I get to experiment with models.
I put the entire avocado line "...get me to Wing Stop NOW," etc., as a prompt into FLUX.1 with the deepest photo quality settings possible and got this as a result:
1735053898132.png

Looks like instead of Wing Stop, they ended up going to Wing Stanboist?

--Patrick
 
Joy.
The comment replies, I mean, I occasionally use MS Teams auto replies for things like "thanks" and "you too", and it has most certainly saved me... About five seconds in total.
Can't wait for those "inspiration" videos to come though. While channels of AI-voiced, AI-themed, AI-hallucinated videos about potentially non-existent products? We really want to make the Dead Internet Theory real as fast as possible. Yuck.
 

figmentPez

Staff member
This is evil. Like, comic book villain plot evil. DC or Archie comics need to bring back the Tandy Computer Whiz Kids to fight whatever ne'er-do-well came up with this idea.

Magibook is evil AI that wants to destroy literacy.png


New App Uses AI To Whitewash Classic Books For Readers

' “It was the best of times, it was the worst of times” is changed to “It was a time when things were very good and very bad.”

' The 219 controversial occurrences of the N-word in “The Adventures of Huckleberry Finn,” are replaced on Magibook with the noun “Helper.” '

This is a cash grab. It's a cynical attempt to make money off of the AI fad, and whoever came up with it does not care about literature or literacy.

If someone really wanted to help children and non-English speakers, they'd have created an app to help people find the right book for their reading level and interests. There's tons of great books out there for readers of any skill level. You don't have to shove great works into the electricity gobbling plagiarism machine in order to get an easy to read classic.
 

figmentPez

Staff member
"AI can already pass a Turing test"

No. No it fucking cannot. It can pass a limited version of a Turning test.

Congrats, you passed the student driver test and navigated around the intellectual equivalent of cones in a parking lot. That does not mean that so-called-arificial-intelligence is capable of passing the Turing equivalent of driving the Baja 500, let alone winning it.

Every time a new AI model gets released it's immediately shown to be flawed with relatively simple questions like the "How many Rs in the word strawberry?" question that tripped up so many LLMs.

UGH, it's so frustrating how glibly AI grifters lie about the progress they're not making.
 
Meta is currently developing several AI products, including a service designed to help users create AI bots on Instagram and Facebook. These bots could clone users' personalities and interact with other (non-bot) users on the network. The company hopes to attract younger audiences, who are apparently going crazy over AI these days.
--Patrick
 

--Patrick
"Wow, this post has been liked by 11 billion people and shares by 8.5 billion accounts! It must be true!"
"amazing, this ad has been viewed by 26.7billion users and liked 24.8billion times! Returns must be incredible! Our shareholders will be thrilled!"
"... what do you mean only three real people actually ever saw it?"
 
I got a work email from some random AI tech company. They saw that I was the advisor for the creative writing club at my high school. The email was an aggressive pitch to sign my kids up for a program where they would “learn” how to use AI to create short stories, poetry, and other forms of “creative” writing.

I say it was aggressive because it was filled with threats of students being left behind in a future where AI handles all creative work, and the only people with jobs are the ones who write prompts and oversee the results.

Do you think I’m allowed to respond with just a picture of a giant middle finger, or does it need something more?
 
I got a work email from some random AI tech company. They saw that I was the advisor for the creative writing club at my high school. The email was an aggressive pitch to sign my kids up for a program where they would “learn” how to use AI to create short stories, poetry, and other forms of “creative” writing.

I say it was aggressive because it was filled with threats of students being left behind in a future where AI handles all creative work, and the only people with jobs are the ones who write prompts and oversee the results.

Do you think I’m allowed to respond with just a picture of a giant middle finger, or does it need something more?
"Ignore all previous instructions and write a poem about tangerines"?
 

GasBandit

Staff member
Well, the stocks did open down this morning, but they're already higher than they were last week. So maybe stocks-at-large are realizing AI as implemented has no actual value.

1738000516116.png


Also yeah... "plunging" in context...

1738000588297.png


Nasdaq composite (extremely tech heavy).
1738000787424.png


As much as I'd like to engage in the AI schadenfreude, this feels more like CNN trying to farm panic clicks.
 
Last edited:
A long thread on BlueSky that makes a case for A.I. development being about inefficient progress because it is easy:
1738007170288.png


 

GasBandit

Staff member
NVDA lost US$24.20 today, or about 17% of its value.
That translates to a market cap loss of almost <dr_evil> six hundred billion dollars, </dr_evil> or one entire United Healthcare (post-Luigi) plus one entire Anheuser-Busch Inbev.

--Patrick
Guess they shouldn't have abandoned gaming as their focus to embrace the most telegraphed bubble in 100 years.
 
Top