Artificial Intelligence (AI)

[bamberfishcake]

There are plenty of similar ones on the net lol..

Dark Zuckerberg!

 

[Guntis]

This is very popular right now; the video is a bit old, but it shows many details:

 

[bamberfishcake]

Have used MidJourney a few times, but just messed with it a little.

Nothing like this.

Was interesting to see the guy paste an image and ask for a prompt, then use that prompt to generate an AI-generated image of the image.

And that FaceSwap is amazing, but a little scary. Not sure you can trust any photo now.

 

[Guntis]

Have used MidJourney a few times, but just messed with it a little.

Nothing like this.

Was interesting to see the guy paste an image and ask for a prompt, then use that prompt to generate an AI-generated image of the image.

And that FaceSwap is amazing, but a little scary. Not sure you can trust any photo now.

Not sure we'll be able to trust anything soon, lol

 

[Kroffe]

You could probably make good money if you create a fake person and use it for a Onlyfans account.
Would that be breaking any laws?

Everybody knows that the future world is a world thats full of love, and AI generated Onlyfans accounts.

 

[dunover]

What AI did to our favourite slot:

Bonanza is a slot gambling game from the company Big Time Gaming, found at many online casinos. It has a mining subject and shows pictures of gems and card symbols with wilds appearing only on the top reel. You need 2 or 3 consecutive matching symbols from the left to win an amount back. It explodes winning symbols and allows new ones to fall from above for possible multiple wins within one play. It will pay back players on average 96c per one dollar staked.

It can sometimes, rarely, offer a bonus free games feature by landing 4 or greater gold bars giving 12 or more spins where wins are affected by an increasing multiplier. I have generated an image and video of a 14400 times bet unit win on YouTube as a representation of the capability that players can believe in.

Bonanza is a popular game among gamblers, the mentally ill and disillusioned individuals. Afflicted persons can be seen in gambling subject-videos and internet gambling forums or chatrooms. After many years the game has attracted online, in majority these listed adjectives or nouns: shitter, bastard, filth, disgusting, fucker, cunt, despicable, tight, unbelievable, twatting, vomit, wanker, fuckpig, rigged, fixed, pre-scripted, bollocks and bullshit.

 

[incrediblestuff]

You guys are lagging behind!

Now on the menu are 1: Sora, which promises even more fun in the age of false information, and 2: 'AGI' - more or less being used as broad term for AI fully capable of performing a lot of Human tasks much more efficiently than Humans. Both of which, at the current rate of unforeseen leaps in progress, should be a thing in 2030, or thereabout..

I keep saying it: better be nice to your Smart-fridge!

Edited to add: forgot to mention that spurt in unforeseen growth is highlighted by the following: they now had AI (GPT4) use EUREKA - a hybrid gradient architecture, to optimally make use of a reward based scheme, to rewrite and improve code, that could/will significantly make Robot controllers more advanced, and increase their dexterity. This can also be applied to program other AI-models faster, better and more advanced than Expert Humans previously could: remember that ChatGPT5 is going to be around 10x more powerful than GPT4, and i'm being on the modest side with that estimate

 

[Kroffe]

You guys are lagging behind!

Now on the menu are 1: Sora, which promises even more fun in the age of false information, and 2: 'AGI' - more or less being used as broad term for AI fully capable of performing a lot of Human tasks much more efficiently than Humans. Both of which, at the current rate of unforeseen leaps in progress, should be a thing in 2030, or thereabout..

I keep saying it: better be nice to your Smart-fridge!

Edited to add: forgot to mention that spurt in unforeseen growth is highlighted by the following: they now had AI (GPT4) use EUREKA - a hybrid gradient architecture, to optimally make use of a reward based scheme, to rewrite and improve code, that could/will significantly make Robot controllers more advanced, and increase their dexterity. This can also be applied to program other AI-models faster, better and more advanced than Expert Humans previously could: remember that ChatGPT5 is going to be around 10x more powerful than GPT4, and i'm being on the modest side with that estimate

I would assume that whatever we plebs get to play with is already pretty old by time we get access to it.
Anyway, we all know where we will end up, there has been several documentaries about it already.

 

[dunover]

You could probably make good money if you create a fake person and use it for a Onlyfans account.
Would that be breaking any laws?

Everybody knows that the future world is a world thats full of love, and AI generated Onlyfans accounts.

I got that double album 'Keeper of the Seven Keys' back in the 1990's along with a live double and Chameleon album. Good stuff. I don't need to watch the video to remember the lines lol "Our future lives will be glorious...come with me to future world"

 

[Guntis]

In case anyone needs it, here is a decent resource to keep up with all AI-related stuff. They also have a free subscription where they send different insights, trending AI tools, etc.

You do not have permission to view link Log in or register now.

 

[goatwack]

Whilst humans like to get giddy with new tech and trends, the 'A.I becoming Skynet' or gaining sentience is of course hyperbolic nonsense.

There's no denying that in its application, 'A.I', i.e cleverly programmed software, is just that, software and code based on the users' input, to create what we have now: a glorified search engine optimized in a streamlined way to rearrange information in great ways.

Yet based on human input they certainly are, not to mention the database of 'what's already out there', much like ChatGPT only going up to Sep '21 in its archives, from recollection.

Which makes the bias of Google's Gemini all the more laughable as they are caught bang to rights, effectively forced to slink off as they 'rework' their A.I after being exposed. And when the programmers' bias is so heavily skewed one way, then it becomes clear that results are manipulated to the nth degree, and certainly cause for defamation suits!

Who can forget the classic:.....

 

[L&L-Jan]

Been trying to let Gemini roast a celebrity or tell a joke, but she wont. Shame. Useless tool.

 

[Guntis]

Whilst humans like to get giddy with new tech and trends, the 'A.I becoming Skynet' or gaining sentience is of course hyperbolic nonsense.

There's no denying that in its application, 'A.I', i.e cleverly programmed software, is just that, software and code based on the users' input, to create what we have now: a glorified search engine optimized in a streamlined way to rearrange information in great ways.

Yet based on human input they certainly are, not to mention the database of 'what's already out there', much like ChatGPT only going up to Sep '21 in its archives, from recollection.

Which makes the bias of Google's Gemini all the more laughable as they are caught bang to rights, effectively forced to slink off as they 'rework' their A.I after being exposed. And when the programmers' bias is so heavily skewed one way, then it becomes clear that results are manipulated to the nth degree, and certainly cause for defamation suits!

Who can forget the classic:.....

View attachment 193981

This is that image.



Gemini can access information from the real world through Google Search, and as they say 'It's our most capable and general model'. But it sucks even against Perplexity whose basic version is GPT 3.5 based.

This is Perplexity's output, and it's also based on real-world data, plus it includes links to sources. It even picked the latest about this topic but says 'AI chatbots' instead of Gemini.



The next is GPT 4 with no internet access, but it has a system prompt for how always to respond.



Gemini compared to those two sounds like a vague piece of terms in that casino, meh

 

[incrediblestuff]

To add to my earlier post. if you have 23 minutes to spare check this out, it's exactly what i tried to highlight in it, but more detailed and with examples of it functioning in real-time..
As scary as it is fascinating!

Ignore the click-baity images and title please
Meet Devon:

 

[justdoit]

Soon this will happen in UK Resisting AI by Dan McQuillan

 

[Guntis]

@dunover might like this for his next BTG slot reviews

 

[Kroffe]

Skynet next.