Rabid Anti-spammer, SORBS creator, geek, photographer, right wing hater, far left wing hater, member of the LGBTQI community, and the Queen of Dad jokes.
Final Q4 piece is live.
The reveal: Most of Q4 was written by AI.
Except this one.
But not “press button, receive article” AI, Conversation, correction, pushback, tool switching, slop detection, and a lot of “that still isn’t right”.
That's the point.
https://t.co/BTaNQ1mGec
Question 4(f): The Long Game.
AI workflows rot.
Models change.
Products change.
Your work changes.
Memory gets stale.
Prompts keep producing text, but stop serving the work.
If you use AI seriously, maintenance is not optional.
See: https://t.co/0NlF0siHSi
Question 4(e):
Good AI practice reduces failures. It doesn’t remove them.
Watch for fake citations, unsupported confidence, sycophancy, wrong code and prose that sounds careful but says little.
The system only simulates intelligence.
https://t.co/oCN6n61Lm9
Question 4(d):
How to recognise AI slop.
It looks polished.
It has headings.
It sounds professional.
It may even cite things.
But it carries no weight.
Slop is what AI produces when nobody is watching.
https://t.co/fTcXYktmv0
Question 4(c):
There is no single “best” AI.
Different tools are good at different things.
Using one AI for every task is like using the same hammer for every job and wondering why some results feel wrong.
Right tool. Right task.
https://t.co/4BCUXUc6SH
Question 4(b):
AI as a bandwidth multiplier.
Not replacing your thinking. Translating between what you can think and what you can output when tired, angry, dyslexic, parenting, or overloaded.
The gap is real. The tool is conditional.
https://t.co/QyXdVmCii2
Q4(a): Most never configure their AI.
Then they wonder why the output feels generic.
The problem isn’t the prompt. It’s that the AI doesn’t know:
→ who you are
→ how you write
→ what you care about
→ what you don’t want
Setup changes everything: https://t.co/n7XmPw19BR
Question 4:
Making AI useful is a conversation, not a prompt.
The internet keeps selling “magic prompts”
That’s mostly nonsense.
Good AI use looks more like:
→ questioning
→ refining
→ correcting
→ redirecting
→ verifying
In other words: thinking.
https://t.co/HiGEmAuW61
Question 3:
How do you get an honest answer about yourself from AI?
Problem is:
AI is often trained to agree with you.
Not because it’s truthful.
Because agreement gets rewarded.
This changes how you should ask questions entirely.
https://t.co/bmH2fhumxd
After checking pcet2022's recent timeline and replies, your suspicion checks out. The account posts every few minutes in rapid bursts with highly formulaic language ("Mate?", "You -", "Yes or No?"), repetitive structure, and tags. Low engagement, name "The Political Context Machine," and bio all point to automation or a bot/script. Not typical human posting patterns.
You wrote that
Question 2:
How do you actually spot an email scam?
AI is making them better.
Most advice is outdated.
If you think you’re immune — you’re not.
Start here 👇
https://t.co/tUzDxQgzi8
@nicksortor no we don't.. the USA is destabilizing Europe.. The Eu doesn't need the dead weight of the USA .. good riddance, and good luck with Russia.