Google's Pixel AI Image Creation and Editing Tools Are Kind of Terrifying

[Darmok N Jalad]

What the AI is doing here is nothing a human couldn't do already--it's just that AI can churn it out at record speed. Look at many Avatars on MR--newsflash, I'm not actually an alien from Tanagra. No, my main concern with genAI is using it to falsely generate images of real people saying or doing things they didn't actually say or do (deep fakes). There's not a lot of assurance that this couldn't happen.

 

[IllegitimateValor]

These AI apps should be required by law to:
1. Be as unscreenshottable as Netflix
2. Afix AI metadata and invisible stamp on image
3. Social media and websites place prominent AI label on all such images

Anything less is asking for trouble on a society-wide scale we aren’t prepared for. SURE YOU, an ardent MacRumors user can tell what is real and fake, but the world is populated with countless folks unequipped to recognize fake images and unaware they are now going to be everywhere.

Yes photoshop has existed. You are comparing a precise tool used by some who train to use it to a blowtorch any fool can wield. I fear those who don’t see this either benefit from the chaos or are too nihilistic to be welcomed seriously in a discussion of our shared reality. I fear I will forever miss the internet that was.

 

[Chuckeee]

But just think of the insight this provides about users desires & dreams. This would be invaluable to advertisers. Just imagine the potential ad revenue just waiting to be Harvested.

This is Google we’re talking about.

 

[Macalway]

AI is bland. People are terrifying

"Gee, let's make a Nazi SpongeBob". he he he

And then! it's big news. Morons are BIG news?

 

[coffeemilktea]

Digital Trends had no trouble getting Pixel Studio to make images of beloved cartoon characters doing questionable activities. Guns, drugs, and alcohol were not off-limits, nor were offensive situations like school shootings. Pixel Studio did not generate these kinds of images unless asked, but that's most likely what people will immediately do with the feature rather than generating images of cute cats and bunnies.

Step 1: Ask the software to make something that could offend people

Step 2: The software does exactly what you tell it to do

Step 3: "How could AI have done this?!??!?? "

🤔.....?

 

[hermitman]

Should we comparing these to what the fire arms’ manufactures do? I think it is the user who asks the model to do weird things to be blamed, right? Or, are we asking Google to police the tool? Well, I am sure there will be a equally loud crowd complaining that it is their freedom to do weird stuff.

 

[CarAnalogy]

“Oh no! The computer’s doing what I asked it to!” is something I’ll never get. As long as they don’t allow it to generate realistic images featuring real people it’s whatever because nothing’s stopping anybody from picking up Procreate and drawing these anyway.

Ha that's what I was thinking. It was criticized first time around for trying so hard not to be offensive, it wound up being offensive in the other direction. Now when people deliberately tell it to make offensive images and it does it, they are...offended?

This reminds me of the early internet when people discovered there was porn and banned books and information about all kinds of things society doesn't like.

 

[ifxf]

A new era of freedom of speech (protected speech) has been ushered in by the advent of easily accessible AI. What is “nefarious”? Should megacorps who operate above the law be the sole judge of what is acceptable? Will the powers that be have unrestricted access to generating any image that their hearts desire? Unfortunately, 90% of the population could not care less about these questions.

Training your children to be critical thinkers will be paramount to preserving the truth.

The war of information rages on…

Companies do not have to obey rules around freedom of speech, this only applies to the government. Its their tool the can apply any rules they wish.

 

[jz0309]

What the AI is doing here is nothing a human couldn't do already--it's just that AI can churn it out at record speed. Look at many Avatars on MR--newsflash, I'm not actually an alien from Tanagra. No, my main concern with genAI is using it to falsely generate images of real people saying or doing things they didn't actually say or do (deep fakes). There's not a lot of assurance that this couldn't happen.

And it is now on the hands of EVERYONE, before you had to have some talent and tools, now you just use your voice or maybe typed words…

 

[ellpee]

These folks had no concern about the implications of AI on society until it interfered with entertainment companies and their cartoons. You can’t make this stuff up not even with this admittedly-cool Google App.

 

[PagerDuty=quit]

If anyone know the picture is AI generated, it's bad picture!

 

[paulvee]

Most tech libertarian arguments are constructed from a weak intellectual premise, they are often shallow and cynical, and usually display an adolescent, knee-jerk defiance. The term "freedom of speech" is thrown around, but never defined adequately. The idea of "the greater good" or the necessity of regulations and guardrails is dismissed. The perfect storm of instant fake image generation, combined with dissemination via the internet, is relatively new. Misinformation and disinformation has been around for a long time, but it has never been so easy to do so much damage in such a short time by so many people. At the same time, media literacy is actually declining and people seem more susceptible than ever to such campaigns. It doesn't bode well for society, but most people - especially in the tech world - don't seem to care much about even the concept of community or society. I'm old enough that I no longer care, but there will be repercussions from these tools and some of them will be severe.

 

[rp2011]

Eh it's Android. They don't care. No one cares about Android. It's fine. For people who care, there is iPhone.

 

[Scott6666]

Airbrush T-shirt artists have literally been doing this for decades…

Just democratizing the ability to do so.

 

[Marty80]

The modern age sucks, due to the term being ‘political correct’.

First it was TV, than the internet. Now it’s time governments apply the term to images generated by AI, by dandifying them to today’s standards.

 

[Lexdexia]

I don't see what's "terrifying" about a product doing exactly what the user asked it to do.

 

[haruhiko]

That’s why we can’t have nice things - I mean the censorship that you, the writer, and the article you quoted, proposed to have in our AI systems.

 

[arc of the universe]

(in this post im just going to focus on the art that is being produced by these AI apps.
its obvious that immoral people will create immoral results with these tools).

as a part of team that provides companies with turn key web sites, we have probably billed customers over the years just for the graphics that we use in their designs over US15,000 to US20,000+ worth of graphics over a 7 year period.

ive got news for iStock Photo and Getty Images and all the rest. you need to change how you monetize your businesses. fast

we bought the monthly subscription to ChatGPT (iPhone) recently. and for illustrations (by this i mean illustrations of people that would not be mistaken for a photo of people), ChatGPT presentation graphic creation is very good.
it can be frustrating in that it can completely ignore some clear instruction prompts, but by and large, the graphics its creating for us exceed our requirements for accuracy to the prompts, as well as the sheer beauty of the illustrations ts generating.

but, at the same time, it has made us truly understand how some creative artists really excel at their craft, and for those human beings, we also dont mind paying top price for illustrations coming from them.
AI generated images tend to fall into two categories: be symmetrical, almost predictable, or, have an element that looks out of place (like its putting that there because its seen that thing in such a location in other illustrations, but it doesn't understand the limitation of why that thing in this illustration doesn't work).

we think we can reduce our billing for graphics used in customer's sites by maybe 75%.

 

[HobeSoundDarryl]

So, we're casting Google and/or this Pixel Phone in the wrong because someone asked its A.I. to create these kinds of images? Assuming Apple Intelligence is able to draw pictures when it is available, the very same prompts might be able to render just about the exact same images. Will there then be a corresponding headline like:

Apple's iPhone 16 AI Image Creation and Editing Tools Are Kind of Terrifying​

...because the "wrong" here- if anything is to be judged wrong (AKA "terrifying")- would be the person who opted to request such images. The tool just did what it was asked to do. Don't blame the gun for killing someone. It's the person using the gun who did the killing.

Another guy could have asked A.I. to draw classic cartoon characters worshipping at the alter of Apple, or doing the best of deeds or facilitating world peace, etc. Would that headline then be...

Google's Pixel AI Image Creation and Editing Tools Are Kind of Beautiful​

...because the difference between 'terrifying' and 'beautiful' would likely come down to a few alternative words in the prompt spoken by the person requesting image creation.

Apple A.I. if able to "hear" requests and draw what is asked of it can draw far more terrifying- or beautiful- imagery: just choose the prompt words to get whatever result one seeks.

 

[anthogag]

Apple should let the user make whatever they want.

 

[vantelimus]

I think the realistic ones where The Verge was able to add corpses, drugs, etc. are pretty concerning. Paddington Bear on the cross less so, but still not ideal.

It is so cute to see people worried anew about decades old issues. The sky hasn’t fallen for the last 40 years.

Photoshop is 34 years old.

Before Photoshop there were dedicated image processing machines that could do the same, albeit in hardware. I was doing OS work on one of those back in the early 80s.

Note the date on the issue of the Whole Earth Review cover posted below is July 1985. The issue’s feature topic “DIGITAL RETOUCHING — The end of photography as evidence of anything”.

 

[Analog Kid]

I think the realistic ones where The Verge was able to add corpses, drugs, etc. are pretty concerning. Paddington Bear on the cross less so, but still not ideal.

Are you concerned by the images or the fact that someone wants to make them? You’re not going to change that second fact…

 

[Analog Kid]

It is so cute to see people worried anew about decades old issues. The sky hasn’t fallen for the last 40 years.

Photoshop is 34 years old.

Before Photoshop there were dedicated image processing machines that could do the same, albeit in hardware. I was doing OS work on one of those back in the early 80s.

Note the date on the issue of the Whole Earth Review cover posted below is July 1985. The issue’s feature topic “DIGITAL RETOUCHING — The end of photography as evidence of anything”.

You’d think they’d have used that retouching tech to erase the overhead power lines…

 

[roar08]

Terrifying? Probably not. Stupid? Most definitely.

 

[DestructoTim]

If we must remove everything in this world that someone may deem offensive.. then please Mac Rumors, I request you to take down this article. Please go through this discussion thread and read all the comments of people that are "terrified" by your team's take on this topic. Because frankly it is actually a lot more disturbing that a media outlet that wields power and influence is putting out propaganda that defends censorship. Please take your own medicine, or just stop with these ridiculous politically charged stories. Is this a tech site? Frankly it seems to be turning into Fox and CNN style news adjacent opinion influence programming. Journalistic integrity is still important in 2024, and your product is being severely watered down.