advertisement


ChatGPT, the latest conversational chatbot from OpenAI

windhoek

The Phoolosopher
After watching the below video by The Nedwriter on the dangers of ChatGPT, I decided to give it a go myself. Suffice to say, it is indeed a conversational chatbot whereby you can have a back-and-forth conversation on the same topic over a number of responses. I even asked the bot to communicate with me in British English and it obliged, how obliging!

So far we've been chatting about music, record players, chess, and moral philosophy. Its answers are completely spontaneous and self-generated, though very non-commital when it comes opinions and preferences, for as it will persistently inform you, it has no opinions, desires or preferences. It's not sentient, not yet, but I reckon it gives us a good glimpse of what a sentient bot might be like.

OpanAI say the following about it: "We’ve trained a model called ChatGPT which interacts in a conversational way. The dialogue format makes it possible for ChatGPT to answer followup questions, admit its mistakes, challenge incorrect premises, and reject inappropriate requests."

Here is the video by The Nerdwriter:


Exciting and a little scary!
 
73xth6.jpg
 
Have you tried arguing with it about posh power cables?

I reckon even an AI bot won't be able to keep that up for 70+ pages without losing the will...
 
Have you tried arguing with it about posh power cables?

I reckon even an AI bot won't be able to keep that up for 70+ pages without losing the will...
I suspect it's already devised a strategy for those:
the OP said:
Its answers are completely spontaneous and self-generated, though very non-commital when it comes opinions and preferences, for as it will persistently inform you, it has no opinions, desires or preferences.
That'll probably work for here, too:

Frankly, it's going to make a rubbish forum contributor.
 
Has anybody else been using it? Apparently, it's gone viral and the whole world knows about it now.
I have, I think it's amazing - it's way beyond Siri/Alexa/Google in how it interacts with the user. I've had all sorts of daft conversations with it, and also tried it out with my son for ideas generation for a school project (which it was brilliant at, so that in itself is already a bit terrifying). If it could speak it's answers rather than write, and was up to date on its info (apparently it is cut off at around 2021?) then it would be like having HAL in the room, and I wouldn't be at all surprised to learn that it was planning to kill me.
 
yes my students are using it to write essays. There is no way to authenticate the work was done by the student.

I have seen it used to write software as well.

IIRC MS are throwing sh1t loads of money at its creators - OpenAI, about $10bn
 
ChatGPT outputs grammatically correct prose with impeccable spelling. Those are the red flags that students are using it for assignments.

Joe

yes - timed essays under exam conditions, maybe open book on a closed network.
 
The other giveaway is that ChatGPT has begun to experiment with ironic ridicule, so you need to be attuned to what passes for machine sarcasm versus human sarcasm.

tgns287sarcasm2.jpg


It's important to mention that ChatGPT has entered puberty, so also watch out for moodiness and misdirected rage.

_large.jpeg


Joe
 
I have been using this quite a bit recently, although demand is clearly going through the roof as I haven't always been able to get on and it can be a bit glitchy, which is to be expected at this early stage. Tonight I keep getting 'load failed' but then I am asking it some ridiculous questions, like - have your developers written in any secret or hidden code? :D It is truly incredible and an amazing tool as you get very good responses in a very short space of time.

Other things I have asked are: help with SEO for my business, text for sales mailers (some great suggestions, which is easily adapted), about Roswell, other life in the Universe, business strategy for partners employers (again, a really good base to build and work from), the meaning of life and so on...

I tend to stray from being matter of fact as I easily forget it has no feelings or beliefs and often find myself asking what it's thoughts are on this or that, doh! :rolleyes:

It's only going to get better and is definitely the future - I absolutely love it :)
 
I've been impressed so far. Unlike web search, it's conversational and it responds in the context. So you can explain a problem in pieces or layers, then eventually probe particular points, and that works very well.

We're already using it at work. It's next level and does give me pause.

I'm pretty sure it doesn't understand as we do and isn't conscious as we are. So Kasparov consoled himself as the king fell ...!
 
And all the time, it learns from these interactions.

..and slowly, and surely, they laid their plans against us...

"You are being watched. The government has a secret system, a machine that spies on you every hour of every day. I know because I built it. I designed the machine to detect acts of terror but it sees everything. Violent crimes involving ordinary people, people like you. Crimes the government considered "irrelevant." They wouldn't act, so I decided I would. But I needed a partner, someone with the skills to intervene. Hunted by the authorities, we work in secret. You'll never find us, but victim or perpetrator, if your number's up... we'll find *you*."
 


advertisement


Back
Top