Apple Shares First Example of Image Playground in Action, and It's Based on Craig Federighi's Dog

[UliBaer]

I just have a Déjà-vu from not so long ago:
People were pissed off by bugs and required workarounds in iOS and instead of getting the bugs fixed, they got presented a bunch of new cute emojis... 🤣

 

[MNWildFan]

Instead of this could you have done more with AirPods Max than just change a port

Yeah, for real

I mean, who is even asking for this AI image generation stuff?

 

[spazzcat]

Sorry but this is a gimmick will be fun for a demo that’s it

You have no idea how busy the AI services that people use to create AI images are ... they run nonstop.

 

[Sasparilla]

This goes along with whats on the Pixel 9, so their keeping parity. Over on the Pixel it'll drop in anything you want into an existing photo and as long its not something too complicated, it'll look perfect (not an animated scene like the example above). While that can be fun / ridiculous stuff, it can also be stuff to trash someone else's reputation. Was reading an article saying we (as a society) are not ready for the responsibility that's coming with this tech, but its coming anyways.

 

[spazzcat]

Instead of this could you have done more with AirPods Max than just change a port

Two completely diffenet teams and what more do headphones need to do?

 

[spazzcat]

Yeah, for real

I mean, who is even asking for this AI image generation stuff?

You seem to be living under a rock?

 

[turbineseaplane]

I just have a Déjà-vu from not so long ago:
People were pissed off by bugs and required workarounds in iOS and instead of getting the bugs fixed, they got presented a bunch of new cute emojis... 🤣

...and now we can "make emojis from photos or prompts", I guess?

Yay? I guess?
😐

 

[cyb3rdud3]

I love it. Like that it is a local model, seems to be nicely integrated, and can be fun.

Look this implementation is for on the phone, it's not replacing full image generation in Photoshop or something. Nor is there a need to blow it up to cover a whole towerblock.

 

[Frantisekj]

That's a pretty disappointing level of quality.

People, what do you think. Will Apple try to bring generative images to level you can hardly distinguish from real photo? Or they will try to stay in distance from being accused of producing dangerous fake images?

 

[Dawn of Individual Merit]

Wow, that's ugly.

 

[CorvusFox]

Like all "AI" images, these are fully soulless and creepy. No thanks. The only people who would use this feature (boomers) won't even know it exists or how to find it, thankfully.

 

[Jonnod III]

There are several generative AI apps out there that surpass this. One is Draw Things which can even run the new Flux model.

Which AI tools are producing better images with models running on smartphones?

Models running is fairly niche, have you tried a google seaerch?

 

[sunapple]

Definitely migrating to the US for these awesome AI features I’m missing! (Sarcasm)

 

[neuropsychguy]

There are several generative AI apps out there that surpass this. One is Draw Things which can even run the new Flux model.

I've used Draw Things a lot on my Mac. Is it producing better images than that on iPhones?

 

[Octavius8]

A super strong reason to buy a new iPhone and make my phone slower!.
Why Apple?.

 

[peneaux]

Not even meme worthy.

Shame, Shame, Shame.

 

[neuropsychguy]

Models running is fairly niche, have you tried a google seaerch?

I have not, but have done a Google search.

I use Draw Things on my Mac. I use Automatic1111 on my Windows/Linux desktop. I was simply asking if people were generating better images on their iPhones with purely on-device (iPhone) models.

 

[klasma]

I want to exchange "dog" by "Craig Federighi" in the prompt.

 

[trusso]

I’ll take the DMA features from the EU over this steaming pile of AI crap any day, thankyouverymuch.

Apple has lost the plot, along with the rest of Silicon Valley.

 

[attila]

I'm glad it says "AI Generated Image" because everybody and their dog (pun intended) sure cannot tell when an image is generated or not

 

[Mr_Ed]

I just can’t conceive of anyone I personally know thinking to themselves:

“Gee. I’d really like to be able to generate an image from a text prompt on my phone. In fact, I think that would be so useful I would be willing to buy a new phone and spend even more for the privilege. And I would be totally OK Ith whatever limitations Apple puts on what can be generated even though I’m paying a premium for my phone.”

But I suppose anything is possible.
I’ll be sure to report back when one of my acquaintances utters anything close to the above.

 

[AirunJae]

I want to exchange "dog" by "Craig Federighi" in the prompt.

But why does this look like Craig Federighi AS a dog? 🐶🤣

 

[klasma]

But why does this look like Craig Federighi AS a dog? 🐶🤣

I don't think the hair is right.

 

[terminator-jq]

This is definitely a feature that is targeted towards a certain demographic. That being said, to call it useless might be a step too far too early. Even as an artist (who previously had negative opinions towards Ai art) I can see some useful cases such as:

1. Having the ability to generate an image to help explain a concept. This becomes particularly useful when the person you’re speaking to is still learning English, is a child or a very elderly person. Google images may help you find something close but sometimes you need something more specific to what’s in your imagination.

2. Creating reference images

3. Creating images to help visualize a scene in a book

4. Playing with your kids

When it comes to “Genmoji”, even as a grown man I can see myself using it to joke around with my friends.

Apple is playing it safe by keeping everything cartoony and even then, I’m sure they will have major restrictions put in.

 

[cocky jeremy]

Instead of this could you have done more with AirPods Max than just change a port

Because the people working on AI features are the same people working on audio devices.