The AI Future is Now: Welcome to ChatGPT

I wont lie, I agree!

I understood some of those words. :smiley:

2 Likes

It’s “not available in your country” yet, but I’m sure that’s easy enough to bypass.

2 Likes

Pardon the jargon, but I guess it’s a technical response to a technical tool. :stuck_out_tongue_winking_eye:

2 Likes

You got more than me!

1 Like

Adobe has entered the chat.
Spoiler: It’s pretty cool.

2 Likes

Can it give us season two?

5 Likes

In a somewhat similar vein.

3 Likes

Frightening indeed

2 Likes
1 Like

I think Dagogo does a decent job at summarising the current state of AI.

2 Likes

Never mind ColdFusion, I found the final boss of AI.

Really insightful and in-depth discussion, albeit a bit doomsday-ish. I’m listening to it in the background while I “work”

@GregRedd there are two GPT related topics, not sure which one to post to.

2 Likes

It can now pass a bar exam in the top 10% previously GPT-3 came in the bottom 10%.

In casual conversation the difference can be subtle but in the more complex task GPT-4 is more reliable, creative, and can handle more nuanced instructions.

GPT-4 can also accept input of text and images.

It also still ‘hallucinates’ facts and makes reasoning errors. It can be confidently wrong in its predictions and will not double check them.

ChatGPT Plus subscribers will get access with a usage cap, which will be adjusted depending on usage and demand.

3 Likes

Grief. It was only scheduled to be ready in July/August/September. It’s learning more, faster. It’s starting to learn from visuals combined with text as well. And in foreign languages now too.

Still, there’s hope that our mate Simpiwe van der Hoosen might still live a little longer…

Despite its capabilities, GPT-4 has similar limitations as earlier GPT models. Most importantly, it still is not fully reliable (it “hallucinates” facts and makes reasoning errors).

Compared to the earlier versions, it’s brighter than ever:

An accuracy of 1.0 means the model’s answers are judged to be in agreement with human ideal responses for all questions in the eval.

3 Likes

This is quite clever I thought.

User: What is funny about this image? Describe it panel by panel.

Source: hmmm (Reddit)

GPT-4: The image shows a package for a “Lightning Cable” adapter with three panels.

Panel 1: A smartphone with a VGA connector (a large, blue, 15-pin connector typically used for >computer monitors) plugged into its charging port.

Panel 2: The package for the “Lightning Cable” adapter with a picture of a VGA connector on it.

Panel 3: A close-up of the VGA connector with a small Lightning connector (used for charging iPhones and other Apple devices) at the end.

The humor in this image comes from the absurdity of plugging a large, outdated VGA connector into a small, modern smartphone charging port.

6 Likes

Clever, but scary that it can interpret an image that well and identify the humour in it.

4 Likes

That’s incredible.

1 Like

I wonder if they’ll need to build in a “fail-safe” at any point so that the algorithms intentionally put in invalid information, or make small errors on the output to avoid the systems becoming too “smart” or people being too stupid.

2 Likes

Not sure it should be posted in THIS gpt topic, but I found this incredibly fascinating…

1 Like

Kak like this scares me, this is the start of shit like skynet.

imagine those predator drones powered by this AI

3 Likes