Dark_Omen
macrumors 6502a
Isn't Apple's place to prevent "offensive" content from the tool. Let people do what they want, it's their device.
Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
These Two Apple Intelligence Features Not Expected Until iOS 18.2
- Thread starter MacRumors
- Start date
- Sort by reaction score
Volvo’s shares plunged recently because of releasing a vehicle which is already delayed and with functions ”may” come with future updates. It’s not in my plan to buy iPhone 16 but for those who does, there is nothing to be excited about. Unless 16’s hardware is >30% better than 15.
I’ll be replacing my 15PM only when Apple changes the design.
I’ll be replacing my 15PM only when Apple changes the design.
jz0309
macrumors P6
i
They key for “Apple Intelligene” to be successful in the long run is to provide real life value to the average consumer (who is also the average iPhone buyer), what they announced thus far - way off from that… Watch this space 😉
when it gets to that point, I’ll watch the Terminator movies again 😉
They key for “Apple Intelligene” to be successful in the long run is to provide real life value to the average consumer (who is also the average iPhone buyer), what they announced thus far - way off from that… Watch this space 😉
I've been using Office 365's AI feature for a while. It's such a mixed bag and I'm left feeling I really don't think AI is the big breakthrough that people want it to be. It's very hard to describe why.
On the one hand, it can do amazing things, and is ideal for some narrow circumstances. Want the outline for a report? Word will create it instantly, tailored to your subject area. Ditto it can create presentations. In Excel it can create pivot tables and charts by analysing data for interesting tidbits. The main Copilot chat tool can search all your docs and chats to find when something was mentioned.
It's very impressive. But it's somehow not actually that useful.
The first problem is that we need features in apps to be cast iron reliable. And AI just isn't. It's pretty good. But it has a habit of letting the user down. It's like having a ********ting friend. You might enjoy being in their company, and you may love what they say. But at the end of the day, you're a fool if you trust or rely upon them.
The second issue is that in the rush to get the AI products to market, nobody has really understood what they are.
Computers until now have been do-it-with-me. Steve Jobs said a computer was like a bicycle for the mind, and this was perfect. A computer enhances human intelligence. You can do the same things but better, and faster.
AI is a different type of computing. It's do-it-for-me. To use the bicycle metaphor, the computer will now steer the bike. All you have to do is point it in the right direction and pump the pedals to make it move. Hmmm...
It's reducing the necessity of the human's abilities, but the human is still right there. This is the fundamental issue.
Computers work great when they do-it-with-us.
They're problematic when they do-it-for-us. Nobody has yet realised this and I suspect a bubble is about to burst.
On the one hand, it can do amazing things, and is ideal for some narrow circumstances. Want the outline for a report? Word will create it instantly, tailored to your subject area. Ditto it can create presentations. In Excel it can create pivot tables and charts by analysing data for interesting tidbits. The main Copilot chat tool can search all your docs and chats to find when something was mentioned.
It's very impressive. But it's somehow not actually that useful.
The first problem is that we need features in apps to be cast iron reliable. And AI just isn't. It's pretty good. But it has a habit of letting the user down. It's like having a ********ting friend. You might enjoy being in their company, and you may love what they say. But at the end of the day, you're a fool if you trust or rely upon them.
The second issue is that in the rush to get the AI products to market, nobody has really understood what they are.
Computers until now have been do-it-with-me. Steve Jobs said a computer was like a bicycle for the mind, and this was perfect. A computer enhances human intelligence. You can do the same things but better, and faster.
AI is a different type of computing. It's do-it-for-me. To use the bicycle metaphor, the computer will now steer the bike. All you have to do is point it in the right direction and pump the pedals to make it move. Hmmm...
It's reducing the necessity of the human's abilities, but the human is still right there. This is the fundamental issue.
Computers work great when they do-it-with-us.
They're problematic when they do-it-for-us. Nobody has yet realised this and I suspect a bubble is about to burst.
Gemoji is the only AI feature on my iPhone that I want. Everything else is a waste of resources on tech that only benefits billionaires.
glindon
macrumors 6502a
This I think is such a lame reasoning. Why can’t we all just be adults and make whatever we like? How about giving those who want a guardrail the option to choose it instead? They could opt to not generate or receive such “offensive” (whatever the hell that means) images.
Being moralized by some computer generated entity is dumb.
v0lume4
macrumors 68030
If I could like this ten times, I would.
diego.caraballo
macrumors 6502a
A thought on Genmoji and Image Playground.
As the way Apple implement it, both platforms aren't capable of producing infinite results.
Just as with Memoji today, since you need to use some of the built-in designs, you can have
a big amount of possible combinations, but not an infinite one.
(According to Apple description on the Memoji app for the Watch, there are over a trillion possible
combinations if you include skin tones and accessories)
So, what's exactly the point of having an AI to generate them in real-time?
Seems a waste to me since you are only choosing between a pre-existent combinations of elements.
Apple could just render all possible combinations once, and after that it's just a big image repository...
I'm missing something here?
As the way Apple implement it, both platforms aren't capable of producing infinite results.
Just as with Memoji today, since you need to use some of the built-in designs, you can have
a big amount of possible combinations, but not an infinite one.
(According to Apple description on the Memoji app for the Watch, there are over a trillion possible
combinations if you include skin tones and accessories)
So, what's exactly the point of having an AI to generate them in real-time?
Seems a waste to me since you are only choosing between a pre-existent combinations of elements.
Apple could just render all possible combinations once, and after that it's just a big image repository...
I'm missing something here?
Bungaree.Chubbins
macrumors 65816
I'm assuming a similar delay to their deployment in Sequoia too. Even so, that's a big 'meh' from me.
As loudly as I have been deriding 'AI' (specifically LLMs), Apple Intelligence actually has some useful stuff for me. I like the proofreading tools, and image clean-up. I still don't like generative AI much, though I have found minor uses for the new chat bots.
I do appreciate that Apple's use of AI is to make better tools, and not just 'Yay! Chatbots!', a focus on creating fake photos, and the such.
As loudly as I have been deriding 'AI' (specifically LLMs), Apple Intelligence actually has some useful stuff for me. I like the proofreading tools, and image clean-up. I still don't like generative AI much, though I have found minor uses for the new chat bots.
I do appreciate that Apple's use of AI is to make better tools, and not just 'Yay! Chatbots!', a focus on creating fake photos, and the such.
FelixDerKater
macrumors 68040
Earendil
macrumors 68000
I agree for general purpose AI. On device “AI” that can help me with my own personal data will be incredibly helpful. I generate more data than my brain can remember and process in a timely manner, and yet the data can be useful. I also have mild Anomic Aphasia, and hope that “AI” will make my life easier in time.
I’ve only personally run into one iOS bug in the last couple years that I can remember, and it wasn’t a blocking bug.
So for me personally, I’ll take the advances in data organization and processing.
WarmWinterHat
Suspended
I have the same reaction. What's the point in these "AI" features if they're all so gimmicky.
“In this week’s Power On Newsletter, German writes to remind us the Power On Newsletter exists. Amazing details are offered on an iPhone that’s over a year away, but the date of the event in 2024 is wrong.” 🙂
Smartwatchlover
macrumors regular
The day there will be an Apple event without this incredibly revealing sentence is yet to come… and won’t be today I fear!
Algr
macrumors 6502a
There is no conceivable message that no one will find offensive. It is really hopeless for Apple to be trying to do this.
Here is an image generator that works on iPhones and Macs now. Apple will never surpass it.
- Free
- Local rendering
- Unlimited
drawthings.ai
Here is an image generator that works on iPhones and Macs now. Apple will never surpass it.
- Free
- Local rendering
- Unlimited
Draw Things: AI-assisted Image Generation
drawthings.ai
Algr
macrumors 6502a
Do you really want the Apple event where they announce that the phones are getting worse?
Will it help us code, like ChatGPT can? Will there be an API so we can hook AppleAI into applications that accept ChatGPT account key?
haruhiko
macrumors 604
How dare the Apple Magic Keyboard can be used to generate offensive texts. The Apple Pencil can be used to generate obscene images. They all should be cancelled until Apple figures out a way to automatically disallow these tools to generate unpleasant text and images. The browser should close itself before displaying p**n sites too.