Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
It’s coming. Alex Ziskind did a video of Tiny ai’s Pocket AI device connected to a MacBook Neo.

This is great! I just ordered a MacBook Neo for travel. Something like this is a great addition if I decided later I needed the capacity to use an AI agent for work I want to do that would either not work at all on the Neo or would be a crawl.
 
I have ZERO trust with any and all contacts with Google. User's information is Google's product, has been and always will be. The bonus now is our surveillance state will love their total access.

Sure hope there is an actually working OFF switch for this AI stuff. Unfortunately a digital slide on-off switch is NOT proof of being actually inoperative.

Not sure if even a purely simple non-digital cell phone could not be programmed to to spy on its user. Location data is constantly supplied to the powers that want to track any person with a cell phone.
 
I have ZERO trust with any and all contacts with Google. User's information is Google's product, has been and always will be. The bonus now is our surveillance state will love their total access.

Sure hope there is an actually working OFF switch for this AI stuff. Unfortunately a digital slide on-off switch is NOT proof of being actually inoperative.

Not sure if even a purely simple non-digital cell phone could not be programmed to to spy on its user. Location data is constantly supplied to the powers that want to track any person with a cell phone.

What exactly are you so afraid of?
 
Is there a way I can use this right now? Like open up a chat via Apple in some way, and get Gemini to do something? I’m comfortable with th echat interface, so … make it so, Apple, or shut it.🍸😺🪊
 
> > Siri, directions to Whole Foods
One option I found is Dunkin’ Donuts on South Milton Road, about 5 miles to your north, and it gets 68% positive ratings. Does that one sound good?
> > No
The second is Fat Bagels on Historic Route 66, about 7 miles to your east, and it gets 2.9 stars on Yelp. Do you want that one?
Siri: "You did ask for foods with holes in them, didn't you?"
 
  • Like
Reactions: jillownsamac
  • ARC-AGI-2 (Logic): Gemini 3.1 Pro (77.1%) > Opus 4.6 (68.8%) > GPT-5.2 (52.9%).
  • GPQA Diamond (Science): Gemini 3.1 Pro (94.3%) > GPT-5.2 (92.4%) > Opus 4.6 (91.3%).
  • MMLU (Understanding): Gemini 3.1 Pro (92.6%) > Opus 4.6 (91.1%) > GPT-5.2 (89.6%).
  • Humanity's Last Exam: Gemini 3.1 Pro (44.4%) > Opus 4.6 (40.0%) > GPT-5.2 (34.5%).

AI benchmarks are nonsense because the responses can be random. Even if you average the scores out they are still nonsense. They are not like benchmarking a CPU. The only resemblance they have to traditional benchmarking is they are used for deceptive marketing gimmicks.

When it comes to using language models their communication style is the most effective way to judge their usefulness. If they gaslight you, are overly censorious, can’t handle trading insults and are too politically correct then they are as useless as dating a nun or talking about evolution with a mullah.
 
What exactly are you so afraid of?
Aside from "what I do, who I see, what I say, view, buy and where I go" is absolutely none of your (Google, Meta, etc.) business, I also might not want the government to know more about me than I explicitly give them permission to know. As someone who lives in Minneapolis, and just watched my neighbors rounded up and gunned down in the street, I have reason to fear the government knowing anything more about me than I actually give them permission to know. When Google, Meta, etc. collect all this kind of information, and then the Department of Homeland Security sends them "Administrative Subpoenas" (translation: "give us that information because we said so") and Meta then turns over that information (which just happened), yes, I have reason to not want that to happen. Asking, "what are you afraid of?" is like the cop who pulls a guy over for no obvious reason. The guy asks, "why did you pull me over, was I doing anything wrong?" and the cop answers, "I don't know, WERE you doing anything wrong?" There's a really good reason we have a Fourth Amendment to protect our privacy against unwanted intrusion.
 
Post above also my reasoning. Think of the Miranda ruling, what you say (do, attend, associate with, go to etc) can and will be used against you.

Best advice is to read George Orwell's 1984 (written in the mid 1950s) as we are probably 90% there now. The television monitoring bits described then are actuality capable of happening now which is amazing...... The exercises of the main person are critiqued in real time over the TV speaker by a remote government viewer.

Seriously think about that!
 
Post above also my reasoning. Think of the Miranda ruling, what you say (do, attend, associate with, go to etc) can and will be used against you.

Best advice is to read George Orwell's 1984 (written in the mid 1950s) as we are probably 90% there now. The television monitoring bits described then are actuality capable of happening now which is amazing...... The exercises of the main person are critiqued in real time over the TV speaker by a remote government viewer.

Seriously think about that!
1984 was finished in 1948, and published in 1949. “1984” came from reversing the last two digits of the year Orwell finished writing the book.
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.