You must log in or register to comment.
First, men trained AI.
Now AI has trained a man.
AI destroys man. AI creates dinosaurs.
Hold on
The advancement of science waits for no one!
Scientists in the background, gluing toppings onto pizza
I wish I understood the reference…
(Granted, this is not the scene where he says the famous line, but it is the imagery I was referencing)+
Holy shit, that’s hilarious
AI, uh, finds a way.
Missing em dash
How would one add em dashes in real life?
Emm pause?
I think we need SI metrics for the duration of different dashes when spoken out loud, so we can use the em dash accurately in speech. Quick, someone call a nerd
Dunno. You shout morse code real loud?
Funny thing is that I have noticed for some folks deep in the LLM chat, conversationally they are a lot more obnoxiously verbose than they used to be.
As a person who can be obnoxiously verbose I don’t know how I feel about this…
Judging by your username I don’t think that’s a problem in your profession.
Yeah! I was already concerned that people’s eyes would glaze over and they’d skim the latter 3/4 of my message and miss some important connected dots.
Now we’ve all been trained to detect and reject AI slop for a couple years, so the whole “here’s the whole situation described in a single message” strategy is right out the window.
I even like using em dashes sometimes. :/
At least I don’t use emojis in place of bullet points. The more you know. ✨
lol same, I don’t think I come off as AI but AI and I have more similar writing styles than your average user
AI comes off as us. It was 100% trained on the small subset of internet users that express complex thoughts and type in complete sentences.
Given how many wordy reddit comments I made over a decade – replete with em dashes – I can’t help but see myself in these little LLM guys. Brings a tear to my eye
I, too, have struggled with verbosity.
LLMs are really stupid and have to have their hand held at every step. Lots of instructions included.
I recently learned the difference between “Justification” and " Rationalization." Justification is laying out the legit reasons for doing something, while Rationalization is explaining why it was okay. Justification is considered positive, while Rationalization is generally negative. Justification is reasons, and usually comes before, while Rationalization is excuses, and usually comes after.
In the case of this cartoon, Rationalization would have been a better word than Justification.
Just sayin’.
It’s a justification because he talks in AI word soup in real life. That’s the punchline in the comic. His defensive justification is that he sounds like an AI in his email because he just happens to sound like an AI all the time. There’s no rationalization required if he didn’t use AI to write emails and the justification for him sounding like AI in his emails is that actually sounds like AI all the time.
I’d always used justification as convincing someone else that your action is right, while rationalization is convincing yourself. For honest intent I usually use explanation
Yeah its like scientific testing results …what is the validity and what is the reliability.
You can get really reliable results on a test that is invalid.
I can’t help but feel that one of my real takeaways from Corporate America is that people don’t read good.
I’m not going to lie. In a professional scenario having LLM feed me responses would be really beneficial, it could even do some irrelevant small talk for me. That thing is significantly better than i am in it and most of this type of talk already consists of predetermined phrases.
Feel free to point out flaws in it.
You’ll never develop your own skills in that type of communication, it’s incredibly obvious when someone uses LLM output in that manner, and many people consider it incredibly rude to use it that way.
Fair, yeah i agree it would hamper communication skills development, but even in a professional scenario? Where most of the communication already consists of fixed responses. Technically not even an LLM would be needed for it, just automated scripts and keywords, just a simple bot would already be enough.
Or does even this almost automated communication have some underlying layers im missing?
Though yeah while i might find basic day to day small talk kinda pointless, it does serve a purpose and using an LLM for it might be already pushing it.
Ive worked in a lot of “professional scenarios” and I can’t think of very many where a “fixed response” would be OK for anything internal comma. Maybe very formal responses to a boilerplate ticket or something.
I’ll yield to your higher experience as my main experience has been just reiterating why the boss is right or why his ideas are great, for which LLMs fit perfectly.
Sounds sad, to be fair
Why’s that?
In my opinion, if main experience is buttering the boss, it isn’t a pleasant or rewarding experience
As the recepient of those responses, they are unwelcome. If I wanted an LLM response, I would ask the LLM myself.
If you want to have your interactions more concise, then you can steer things that way. Don’t meet unwelcome verbosity with more unwelcome verbosity.
Wait, you guys can do that?
When you meet the guy, LLM are trained on
Autismmmm 🤤🤤
The best part is that they no longer used AI in their response. Nowhere does that character sequence appear in their response!
deleted by creator
If it makes you feel any better, the real way to get ahead in the job market is the same as it’s ever been: know the right people.
Yep but if you use too much AI theyll reject you. My careers advisor also said AI scanners aren’t programmed to recognise neurodiversity in applications which puts diverse people at a disadvantage
Do people really talk like that?
That’s the joke. AI talks like that.
i think that basically writing texts in bullet-point style is just better for readability. nobody likes long walls-of-text.
the thing is just that until now, you couldn’t convince people. suddenly people see AI output lots of long but still easily-readable texts in bullet-point style. people realize how convenient that is and start copying it.
Imagine what its like talking to someone who uses Emacs Doctor as their AI
