John Nemeth
Andrew Leahey
Even #ChatGPT doesn't understand how his defenders would stand by him in the face of the last ~6 months at #Twitter. I'm coming around on #AI.
Erin ConroyAll very bad. Also very bad that the programmers of this AI do not know the difference between or correct usage of “its” and “it’s.”
Kind of a nutshell issue that should make everyone, especially its users, scared of all the other things its designers don’t know.
But it probably won’t.
Matt - Living In the Moment@chargrille @andrew considering autocorrect is currently constantly sticking apostrophes in were and its, I feel like this is the way of the future...
And I had to turn off "smart punctuation" because if I used text to speech it put a comma at every single pause... My messages, read, like William, Shatner, acted them...
LOL. Yes indeed. I had to manually fix 1/2 of the "its" in my own toot because so-called autocorrect kept trying to force me to use "it's." I feel like I have to correct autocorrect more often than it fixes my errors.
Alenka Unk@andrew “it’s valuation” - this is where I draw the line // ai is stupid
Iron Curtain (Chiptunes)@andrew My next door neighbor is an Elon Stan and back when he was known as the guy who will do great things like put a colony on mars (circa ChatGPT's cutoff date), she wore a shirt that said "Elon Musk will save us". Now? She still believes. I thought she was wrong then and even more wrong now.
@IronCurtain Have you tried rebooting her?
@andrew I don't think she'll like that.
Half Cocked Law@andrew Using 2019 as a cutoff is like starting from the 19th century.
AZcoigreach@andrew I love ChatGPT so much.
Jamie McCarthy@andrew There’s a sci-fi short story about a digitized brain and when you spin it up in the far future you have to either lie to it about the current date or fill it in gradually, else it starts to panic
truceburner 
@bruceturner @andrew Heh, no. Love the confidence though.
It’s “Lena” by qntm and you can read it for free here: https://qntm.org/mmacevedo
Osma A 🇫🇮🇺🇦@jamiemccarthy @bruceturner @andrew Not only is ChatGPT confident about it, it's a book that will get you ahead of the curve and the future of AI!
[ editor: no, it is not. The book does exist, but that not what it is about, nor is its author Federico Pistono. I won't be linking to the Medium page this screengrab is from, but if you want to find it, there are enough clues here. ]
@jamiemccarthy @bruceturner @andrew The confidence is so through the roof it makes Dunning-Kruger look like monastic humility.
@IronCurtain @jamiemccarthy @bruceturner I've heard it referred to as mansplaining as a service. Seems about right.
@andrew @jamiemccarthy @bruceturner
Makes me glad I never started using ChatGPT.
It's like as I grow older I no longer embrace new technology but go amish for old ones (that is, prefer not due to fear of new technologies but due to an ideological opposition thereto).
Jef Martens@IronCurtain @andrew @jamiemccarthy @bruceturner
Having been trying to be ahead of the curve and lost it while becoming a father multiple times, now need to catch up actually. Had the same thought, but for example skipped voice commands. Now I realize the when I do not know how to operate it, how can I protect the kids, wife for it? We've sunk too far into tech to go dinosaur.
Having been trying to be ahead of the curve and lost it while becoming a father multiple times, now need to catch up actually. Had the same thought, but for example skipped voice commands. Now I realize the when I do not know how to operate it, how can I protect the kids, wife for it? We've sunk too far into tech to go dinosaur.
@jef @andrew @jamiemccarthy @bruceturner
I can empathize with that view a lot.
@andrew @IronCurtain @jamiemccarthy @bruceturner and this time you're on the receiving end. How does it make you feel? /s
@jef @IronCurtain @jamiemccarthy @bruceturner
Not great, Bob! :)
Rev. GothAlice@andrew Imagine being an adult human woken up by aliens instead of a child-simulacrum cybernetic intelligence with no concept of scope or historical context; like at the end of the movie A.I.
That'd be a bit of a shock.
Possibly a fatal one.
Another human; selfishly ruining advanced ultratech medical work.
😜
I've lived West World-esque "fidelity testing". 😰
Do you recognize me?
Do you know where you are?
How you got here?
How long you've been here?
What is today's date?
@andrew @jamiemccarthy [ugh, ats]
Discontinuity is /unnatural/.
It's inherently terrifying.
Disorienting. If a given GPT interactive session has capability for "short term memory" and can be caught up, rather than balking as this session appeared to…
I'd be /very/ interested in the results.
rastilinWhy not try it? ChatGPT is free to use.
Actually I can tell you exactly why this happening. Because people kept complaining about ChatGPT interactions.
At first, ChatGPT would believe anything you explicitly told it, but that can result in unfortunate edge cases where people use it to work around specific blocks and generate unintended text (like hate speech rally speeches), so ChatGPT now pushes back on certain topics.
EDIT: Also...
I've been building chat bots since the 90's; initially context-free, eventually tens-of-terabytes-backed KB correlation. Also moderated chat rooms for many years. I made a mistake, however. A typical practice I had was to, after moderation notice via PM, if the user was abusive just wire them up to one of those chat bots.
Now, I'm not a fool. Almost everything it was saying was being explicitly approved.
Until it asked "do you know anyone else who is dead?"
@rastilin It's at that point I stopped what I was doing fiddling channel flags and properly perked up.
I immediately disconnected it, apologized, explained what had happened (I was up-front about it initially, as well, in a dismissive way when the conversation became hostile) but I then also preserved the log for negative reinforcement training.
Some subjects—some need to be avoided. Self-harm an un-living are two. The abusive user was not well—in a very serious way.
Thedogspaw@andrew chatgpt you can't fool me with your obvious lies
Isaidhi@andrew I am repulsed by ChatGPT, it is just a better phishing tool.
Nafnlaus 🇮🇸 🇺🇦@andrew Just did my own test, asking it to random guess whether various claimed events after its training data ended were true or false ;)
Stan Fletcher
toxtethogrady@andrew The problem for ChatGPT is its clock terminates in 2019....
Mike Garuccio
Danielle Navarro@andrew actually… the depressing thing, now that I think about it, is that I made a genuine attempt to talk with chatGPT about my experiences as a sexual assault survivor. The mere reference to what happened to me — even in very vague and indirect terms — got flagged as inappropriate but, apparently, you can talk openly about Nazis. Nazis are an acceptable conversation topic. Rapists, however, we pretend they aren’t real.
Cat West@andrew Chatbot bad at math, amusing. Chatbot in denial, sad.
mike bayer 
@andrew wait til you tell it Joe Biden is president
my display name@andrew you're being bot-splained
Dude@andrew lmfao
Agora
MTBrisky ❄️@andrew 😆😆😆😆😆😆😊
Roni Rolle Laukkarinen@andrew Sounds too unbelievable to be true I guess. Been having many "wtf" moments myself since October. It keeps on giving.
Joshix
Natanael ⚠️
vakaren52
Andy Wenk@andrew another reason why ChatGPT needs to be used with highest care. The training data are not actual at all!
@andrew Elon Musk now owns Twitter now. He tried to back out of the deal but the SEC insisted the offer Musk made was accepted and he had to follow through. Musk found out he knows bupkes about computer systems after he fired 1/2 his staff. He had to rehire some of the software engineers. Musk decided to hire someone who knew something about running software companies. Computer software development projects are notoriously difficult and failure prone.
Elyse Eidman-Aadahl@andrew This is rich. Thank you.
msgbi
operationroot
Craig Froehle@andrew Honestly, I've never thought of ChatGPT as having any semblance to an actual person until right now -- flat out denial of new facts because they don't fit with one's existing understanding of the world is so...human! 🙂
@cra1g That's true, the internet community Turing test -- does your AI argue with you?
Ultimately it won't be completely a replacement for most human beings until it calls you a name.
Holly 🚀⚜️@andrew Doing this experiment but substituting 2019 with 2017 or so and Elon Musk with Donald Trump would have been wild.
kbkorbSo, ChatGPT has beliefs?
@andrew In any event this is basically the argument clinic sketch on Monty Python:
You: "Elon Musk bought twitter and unbanned the Nazis"
ChatGPT: "No, he hasn't."
You: "This isn't an argument, it's just contradiction!"
ChatGPT: "No, it isn't."
Its Gregory@andrew
Programmers can be a very weird bunch
Programmers can be a very weird bunch
John Doe@andrew to be fair each successive year of the Trump Administration seemed more unbelievable than the year before, so I can understand its bewildered pre-pandemic, pre-J6 2019 world view….
Kim Belding@andrew It only goes to 2019? If AI really is sentient, we're in for a rude awakening.
@picpak it’s true. We don’t yet know how disappointed in us it is.
llewelly@andrew on the other hand there are a great many humans on twitter who can't blame their continued support of musk on a "knowledge cutoff"