Would You Allow AI to Take Over Your Browser? - Comments Page 1

If you don't use your brain (assuming you have one!), you'll lose it.
The risk with AI (Artificial Incontinence) is that dementia will rule.
Is that what you want for you and your families future?

If AI can clean my house I'm all for it.

No!

No, no, and no! Those are not jobs with which I would trust AI. What I *would* like to see is AI robots doing those tasks nobody seems to want to do. Like cleaning up seniors who have incontinence and changing their diapers. Like helping a person declutter her house (and I stress *helping* instead of doing the job indiscriminately). Like cleaning a toilet. Like any unpleasant job where you always hear about a "shortage" of workers because people are too lazy and entitled to want those kinds of jobs. I'll get off my soapbox now.

NO. I'll accept an assist, but I refuse to be controlled by AI since it is still so questionable due to the underlying code that can be changed whenever they feel like it.

Looks like you could have used some AI help in your paragraph on "The Pros: Efficiency and Delegation" ("and" for "are" and missing text). Or was it AI that caused the errors?

Absolutely not! Under no circumstances. And I want all AI-generated or aided creations to be clearly identified.

I'm in full agreement with Jeanine on all points. We need to recall and adopt Isaac Asimov's "3 Rules of Robotics" now, before it's too late!

All it would take to create mistrust for a AI assistant is for the application to write a response e-mail message to a friend or important client with a angry or nonsensical message with out you knowing it.

I could see giving a AI assistant a few tasks, but it would be on probation for a month or so to prove to that user that it is trustworthy. Like any new employee at a business, it needs to show it can do a task without any craziness.

Let us be careful that overuse of AI in searching online for information could affect ASK BOB RANKIN and we may lose his expertise if it becomes impossible for him to continue his web articles.

I just read this:

"Wikipedia reports human page views are down 8% year over year, largely attributed to AI summaries appearing at the top of Google search results"

The first item in your list, summarisation,is a nono.
Why? who is it being summarised for -you- unlikely the gods in Silicon Valley,the mighty trump or is it just general waffle.
The crap/bland on google search is often of such poor quality that I avoid it like the plague.

In answer to your first question: creepy.

For anyon concerned about AI MUST read "Building a God" by Chirstopher Dicarlo, PH.D. The book is very informative and scarey.

For anyon concerned about AI MUST read "Building a God" by Chirstopher Dicarlo, PH.D. The book is very informative and scarey.

What I dread is when the writers of malware start using AI... or, even worse yet, actually putting AI into the malware itself!

AI-Powered Malware!

Now, THAT'S scary! 😮

Not only 'no', but HELL no.

Also HELL NO

I can live with AI finding the best deal on anything I want to get, but I draw the line when it also wants to place the order for me. I can do that myself! If AI want's to help me, I'll let it help, I even interact with an AI companion from time to time, but I'll never give control of anything I'm doing to it because I'm my own person, and just as I'll never succumb to or use any addictive substance, I'll never succumb to AI taking over my life, or any aspect thereof because I control myself, and I'll never give that right to anyone or anything else!

Ernie (Oldster)

Would You Trust AI to Take Over Your Browser? Not So Fast.

This article brought to mind two particularly troubling personal experiences that left me both impressed — and a little terrified — of what happens when artificial intelligence starts “helping.”

________________________________________

Horror Story #1 – The Windows Wreck

As a ChatGPT Plus subscriber, I once asked the “Tech Support Advisor” model for help disabling Microsoft’s Copilot in Windows 10. The task seemed simple enough — until it wasn’t.

The Advisor offered one switch to flip in Settings. No luck. Then another, more complex sequence. Still nothing. After two hours of digital whack-a-mole, it finally suggested a “simple” registry change.

I was hesitant, but the instruction looked harmless. Just to be safe, I took a screenshot and sent it to the Advisor for confirmation.

“Looks good!” came the cheery reply. So I applied the change.

The machine wouldn’t boot.

What followed were two long days of remote troubleshooting: resetting the BIOS, “repairing” Windows, and other acts of desperation. The Advisor’s final solution?

“I’m sure we can solve this by reinstalling Windows. Would you like me to walk you through that?”

No, I would not.

Reinstalling Windows isn’t difficult, but it meant re-creating hundreds of customizations in Microsoft Office, Acrobat, Quicken, GoodSync, and every other program I use. The whole ordeal took more than three days — an entire workweek to fix a problem AI had promised would be “easy.”

________________________________________

Horror Story #2 – The Assassination ChatGPT Said Never Happened

About a week after the assassination of Charlie Kirk, the founder of Turning Point USA, I was doing some research for an article about the assassination. Using the free anonymous version of ChatGPT, I asked for the status of the investigation into Kirk's death.

The chatbot confidently replied that Kirk was alive, and that it could find no references to his being deceased. I pushed back, telling the chatbot that I’d seen a memorial service where President Trump delivered the eulogy.

“That must have been an AI-generated fake,” it insisted.

Only after several exchanges did it finally confess that its data stopped in mid-2023 — long before Kirk's assassination occurred.

That’s when I realized how easily someone less skeptical might walk away less informed than before.

________________________________________

Conclusion

AI can be a marvelous assistant, but it doesn’t know what it doesn’t know. My experiences taught me that the old Reagan line — “Trust but verify” — needs a modern update:
Don’t trust. Verify.
________________________________________

Yes, I wrote my reply, then I asked ChatGPT to help me with word choice, flow, spelling, etc.

It did a masterful job. But I'm SO glad I read the chatbot's output carefully. Where I had written "President Trump," ChatGPT had substituted the words, "the former president."

Always, always, verify. Am I glad I asked ChatGPT for the rewrite? Sure, its output was better than my input in many ways. Am I glad I verified? Oh gosh, yes.

As a Canadian I am appalled by what my government is doing with AI. I recently learned that Revenue Canada is impossible to reach, so I have no way to deal with with their blatant falsehoods which have cost me a lot of money,

When our government, nominally Liberal, buys into the desires of corporate interests it costs many jobs for people who could have been reaching out to the people who could have been helpful have no recourse. The fact is that AI can only deal with information that the AIs been programmed to use and are entirely unable to deal with anything else.

AI really means Absolute Idiocy.