Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Yep, forcing it in there is a problem. A little while ago my manager said something to the effect of "I loved XYZ, it was a great product, but now they've rammed AI into it and we can no longer use it, because sending this data to their servers violates our privacy policy". This sort of thing can't go on.
This.

I was looking for a good, one time purchase or free open source, cross platform (mac and windows is a must, iphone and ipad good to have) pdf editor. PDFgear recommended by several people and by several websites but then again it has an AI feature which you cannot disable. This is a red flag as I’m not paying for it so it means my data will be mined. (Still searching by the way, if anyone has a good recommendation, I’m open)
 
That would be a glorious sight to see. Someone did a video of ChatGPT and Gemini arguing back and forth with each other. It was hilarious to see each one trying to debate and rationalize their lies and propaganda.
AI Justice, AI Plaintiff argues with AI Defendant and Judged by AI Jury and Judge. 😝
 
No, I’m genuinely excited too. And genuinely frightened. AI is not overhyped—it’s going to change everything.

I do agree with those who don’t want it in their beach balls and dinner plates though.
AI underwear, I think you have **** your pants. 😝
 
  • Like
Reactions: drrich2
No, I’m genuinely excited too. And genuinely frightened. AI is not overhyped—it’s going to change everything.

AI is simply a tool. And like any tool, it's only as useful as the person wielding it. People have to remember that fact, and not blindly rely upon it replacing common sense, and to always verify it's output. Nothing AI says should be taken at face value when inherent bias are intentionally introduced into it's datasets.
 
AI is simply a tool. And like any tool, it's only as useful as the person wielding it.

Sure. You could say the same about the combustion engine, or the lightbulb—yet they completely changed the way we live.

People have to remember that fact, and not blindly rely upon it replacing common sense, and to always verify it's output. Nothing AI says should be taken at face value when inherent bias are intentionally introduced into it's datasets.

True. AI makes plenty of mistakes to be sure, yet in my experience, it tends to answer in a more knowledgeable, sophisticated, and dare I say it, wise way than most humans I know. And keep in mind, we’re at the very beginning of the AI revolution. It’s nuts that it’s already as good as it is. What I also find incredible is that so many people don’t seem blown away by it like I am!
 
Wouldn't CES be a great place for Apple to drop a new device they want to sell to 5-7M customers a year?
 
Good video. Nice to see the new Samsung TVs and TriFold. Waiting to hear more on the pricing of these new TVs. Some of the new accessories by Belkin are also very good. The O Station battery charger is another nice product.
 
  • Like
Reactions: mganu
These AI enabled smart TVs… I doubt for a second that all of that is done on device.
That means the already “always connected gadgets” are even more hard core connected right? Sending current film images and receiving responses from the AI processing datacenters?

Seems like they would go obsolete quick too… the moment their AI updates to an incompatible model, server being pinged shuts off in a few years, etc
 
AI makes plenty of mistakes to be sure, yet in my experience, it tends to answer in a more knowledgeable, sophisticated, and dare I say it, wise way than most humans I know.

The crux of the problem AI is a lie. It's (A)rtificial, but not (I)ntelligent. It simply answers based on a set of probabilities and inference. It has no idea if it's right or wrong (hence the lack of intelligence), instead simply parrots it's training data. AI's strength is in pattern matching, like searching a MRI scan for anomalies.
 
The crux of the problem AI is a lie. It's (A)rtificial, but not (I)ntelligent. It simply answers based on a set of probabilities and inference. It has no idea if it's right or wrong (hence the lack of intelligence), instead simply parrots it's training data. AI's strength is in pattern matching, like searching a MRI scan for anomalies.

The problem for it is if the patterns are wrong or contradictory, its answers will be bad as a result. The noise an overwhelm the signal, so to speak. It really doesn't reason so it can't decide if an answer makes sense. I have used it to help answer code questions, and I'd say 10% or more of the time the answer is not correct; and this is with a data set that ought to be fairly robust. It appears it has old documentation, reddit answers, etc. that it draws on and as a result makes the wrong inferences when for whatever reason they seem to have higher probabalities.
 
  • Like
Reactions: Ursadorable
The problem for it is if the patterns are wrong or contradictory, its answers will be bad as a result. The noise an overwhelm the signal, so to speak. It really doesn't reason so it can't decide if an answer makes sense. I have used it to help answer code questions, and I'd say 10% or more of the time the answer is not correct; and this is with a data set that ought to be fairly robust. It appears it has old documentation, reddit answers, etc. that it draws on and as a result makes the wrong inferences when for whatever reason they seem to have higher probabalities.

Exactly. It's a glorified search engine.

One day it will eventually become intelligent, but until then, it may as well be called (A)rtificial (I)diot.
 
  • Like
Reactions: jlc1978
AI is simply a tool. And like any tool, it's only as useful as the person wielding it. People have to remember that fact, and not blindly rely upon it replacing common sense, and to always verify it's output. Nothing AI says should be taken at face value when inherent bias are intentionally introduced into it's datasets.
Our relationship with AI is that of master and servant. Problems arises when we forget we are the master and AI is there to do the grunt work for us, not the thinking.

I'm beginning to see student lose what little critical thinking ability they've cultivated over the years. Now they just regurgitate whatever CheatGPT spits out. If you have to read a few dozen AI written essay, you'd be ready to pull the plug too.😖 Of course you can cheat and let AI read the AI generated essays.😏 There's no point being a teacher at that point.🧐
 
The crux of the problem AI is a lie. It's (A)rtificial, but not (I)ntelligent. It simply answers based on a set of probabilities and inference. It has no idea if it's right or wrong (hence the lack of intelligence), instead simply parrots it's training data. AI's strength is in pattern matching, like searching a MRI scan for anomalies.

And yet, incredibly, it still does many complex tasks better than humans can.

I’m not arguing with you that AI is a tool, or that it can be misused. My point is, it’s an incredibly useful tool, and, like it or not, it will change the way humans live from this point in history—dramatically. Do I find this scary? Yes. Do I worry about what life might be like for this next generation of kids? You bet.
 
  • Like
Reactions: drrich2
And yet, incredibly, it still does many complex tasks better than humans can.

To an extent. It is very good at pattern recognition, what it appears to be good at (yet) is determining the validity or usefulness of those patterns. Computers certainly an do complex calculations much faster than humans, but are still subject to GIGO. Humans are as well, but can still sense if something doesn't make sense.

I’m not arguing with you that AI is a tool, or that it can be misused. My point is, it’s an incredibly useful tool, and, like it or not, it will change the way humans live from this point in history—dramatically. Do I find this scary? Yes. Do I worry about what life might be like for this next generation of kids? You bet.

I agree it will make some significant shifts in how we live as it integrates into our lives; I just think right now we are in the hype phase, much the same as PCs were years ago.
 
I agree it will make some significant shifts in how we live as it integrates into our lives; I just think right now we are in the hype phase, much the same as PCs were years ago.
Astute observation and brought to mind a key example. With personal computers, it started out with desktop computers where people word processed, played Solitaire, etc. But that opened the way for the Internet, a whole other thing. Provided the infrastructure that enabled a new platform. And now, computing is so integrated that calling a business and getting a human on the phone without going through an automated menu is almost surprising.

But another major paradigm shift came with the smart phone. There are many people who were never going to be 'computer nerds,' sitting for hours at a PC at home in their discretionary time, but who took to smart phones avidly. After all, smart phones laid the infrastructure to aid the rise of social media - Facebook, TikTok, texting, sharing selfies spontaneously, etc. We went from computer nerds to some people who felt seen as 'less than' because their Android phone didn't produce blue text message bubbles when texting iPhones.

So now, A.I. is rapidly infiltrating the technological landscape and impacting our culture (e.g.: may lead to automation, layoffs, etc.). My question is then what will A.I. itself be the infrastructural foundation for? And will it have more impact in the wearable space, with people who aren't 'techies' or 'nerds?'

We live in interesting times.
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.