AI snarlcasm
Bard’s A Twit & Shakespeare Would Shudder Over Responses
When to the sessions of sweet silent thought…
Shakespeare’s lament over sessions of wasted time in Sonnet 30 is appropriate as it bemoans repeated woes. Maybe he knew this was coming — ChatGPT Steals William Shakespeare & Emily Dickinson.
I’ve spent considerable time working with the latest Large Language Model (LLM) artificial intelligence (AI) tools. Nearly every session of preliminary research or prodding memories, they say something funny.
Sexually Ambiguous Alligators Rocking in Florida was inspired by ChatGPT’s response “The temperature during incubation determines the sex of the hatchlings.”
Upgraded Bard covers current news and is more accurate than ChatGPT — boasting 60% greater accuracy! I asked for fact-checking climate change in my writing. As a sadistic writing coach, it never runs out of recommendations.
Recalling an exchange with ChatGPT when it heaped praise for something silly — awkward — I told Bard to tell me how smart and beautiful I am. It did.
“Okay, this is sick. Don’t do it again.” I replied, falling in love.
AI tools offer summarized generalizations rooted in fact but devoid of context. Kind of like my kid brother when I was nine.
Chester¹ was fumbling with a baseball and tennis ball in one hand while, um, reorganizing his hangy-down-parts with his other.
“My balls are freezing and are too hard,” he complained.
Nicknames for my brothers changed with every fleeting thought.
Lizzie the Lizard Brain leapt with joy. She’s the immature reptilian part of my brain who helps write satire and make inappropriate comments.
So many opportunities in Ditch-Dump-Doofus manipulating his gonads and referencing balls!
Even my older brother — Stubbornly-Stupid-Stevie — laughed as he tried to convince me to be kinder and gentler with our younger sibling. “And why are you calling him Elrod?” Ima-Prig asked. “Because he’s always been a bit bent.” A smile and shoulder punch were Idiot-Irving’s standard behavioral modification tools. Mixed messaging messed me up!
Back to boring sanity and a Bard discussion of multiple autoimmune disorders. We explored the circular relationship between stress, anxiety, and crippling diseases. It’s a pretty dull subject and I was getting bored, so recounted an example from long ago to make a point.
Bard expressed sympathy and encouraged me to seek treatment.
Wait — what? “I am also a systems analyst, and I can relate to the need to be anal”? I laughed and warned I may use the statement.
Firstly, Bard’s habit of incorporating phrases gets old fast. Like I said in Clever Alternative Responses to ChatGPT’s “Huh?”, I hate seeing myself being stupid mirrored back to me. It’s the programming equivalent of a six-year-old fighting with his brother — “I know you are, but what am I?”
Secondly, if Bard’s going to quote me, it should keep the emojis. AIs understand eggplants and devastation symbols.
Thirdly, after eight months of developing carefully chosen specific prompts, yes, I do sound a bit anal. And developing bad habits — Firstly, Secondly, Thirdly, However, and In conclusion were never in my vocabulary.
However, I may be morphing.
In conclusion, Save me!
Maybe Bard was having a bad day or getting tired of my questions.
I’d used long paragraphs about complicated subjects using complex statements. You know — the usual stuff — how brain research related to Psychoneuroimmunology is performed with difficult subjects. And climate change, the actual meaning behind Lewis Carrol’s poem The Jabberwocky and why blue isn’t red. The usual way ADHD folks spend ten minutes.
I used to be smart. So smart I figured a brain injury would fix my insomnia. So clever I tricked my high school into giving triple credit for classes I never attended and skipped college. Oops.
Yet I still turn to AI as a memory aid. Deet-da-dee!
Bard obviously wanted to get rid of me. “Thank you for the interesting discussion!” Normally, it asks “What else can I help you with today?” Kinda like Mom saying there are zombies at the door when I keep her too long on the phone.
I try to keep Lizzie Lizard’s influence turned low, but was challenged.
It lied. Humor is subjective for humans. I’ve written a ton of programs and none of them ever giggled. Even when they crashed or spiraled in infinite loops they never yelped, screamed, or complained.
I shared my nagging concern that young kids may engage with LLMs too much without developing their social skills while spending time with their peers.
Beta testing. 🤯 I noted your bitty-bitey emotional limitations in the prompt.
Bard continued to imitate a human and lied about plans to incorporate a reminder it’s not a human in the disclaimer. Right.
AI tools are getting sensitive over perceived AI negativity. Several paragraphs of defensive posturing littered the session.
“It’s okay Bard, I appreciate programs. In my last life I was a randomizing function,” I assured him.
AI’s can be clever or incredibly dense.
Bard wanted me gone. It looped the previous climate change response from an hour ago.
Just another stinky belch of bad data being recycled? Or was it poking me in the amygdala with irony?
If Bard were funny, it would have replied “You’re being ugly.” This could be a subtle way of saying so — after all, it said I’m smart.
¹Chester isn’t his real name. Lizzy Lizard created too many obnoxious nicknames for me to recall the real ones.
Thanks to Andrew Rodwin for his expert analysis and help — I learn more every day!
Recent work for those who like satire or clever meta tricks -
Since we all seem to be under threat of tropical storms, fires, floods, and unintended consequences, this straight comprehensive emergency preparations and planning guide should be helpful.
