ChatGPT Gains Support for Reminders and Tasks

[dannys1]

and I just don't care.

AI? Yawn....

Your replies and your signature suggest otherwise.


Also as you're "not interested" in AI, you'll be asking your questions to the free version of ChatGPT which is 'fine' but it's not o1.

I'm not sure how you can be totally "not impressed" by LLMs though, it's pretty impressive what we've come up with even if YOU can't think of a use for it.

 

[dwsolberg]

You're saying the presented use case isn't realistic, and then describing how the use case is realistic. I could tell ChatGPT "Hey, I do groceries on Tuesdays, write me a meal plan and grocery list to make it happen every week" and it'll do it unprompted going forward? That's hugely convenient.

The best use case would be knowing the ingredients you have at home right now, and then suggesting a meal based on that and your likes.

 

[dwsolberg]

People laughed at the first cars because they weren’t faster than horses. You’re missing the bigger picture of where AI is heading and mocking a seed for not being a tree

Be careful of your expectations for the tree.
A part of the reason AI is so helpful over a Google search is that you don’t need to wade through all the ads, filler to make room for more ads and “optimized” results. At some point, they’ll monetize it and the AI will start giving you “promoted” answers with a few ads inserted.

 

[TechnoTiger3000]

I'm not convinced, at all, that the current AI companies are heading anywhere near the future you're proposing. Until accuracy is put at the forefront, it'll be limited.

In the news today: “AI Expected to Replace More Than 200,000 Wall Street Jobs”

LLMs are not equivalent to AI (at all); they are just one component of the broader field of artificial intelligence.

AI Expected to Replace More than 200,000 Wall Street Jobs

Discover insights from a Bloomberg Intelligence report predicting that global banks may cut 200,000 jobs in the next few years due to AI advancements.

www.eweek.com

 

[eifelbube]

These are baby steps. AI is simply the next technology that can be used to amplify the abilities and values of mankind. Those who embrace it and learn how to use it to multiply their effectiveness will continue moving into the future. Those who don't will be left behind. Smack your face all you want, but you'll be doing so in the past.

And if i were to embrace it, how would i be able to trust that MR forum comments were conceived by a creative, insightful and original human mind rather than by an algorithm plagiarizing and blending previously expressed ideas and opinions into a generic post that is not attributable to an individual author? No, thanks. I prefer to be left behind until the the AI bubble bursts.

 

[vantelimus]

And if i were to embrace it, how would i be able to trust that MR forum comments were conceived by a creative, insightful and original human mind rather than by an algorithm plagiarizing and blending previously expressed ideas and opinions into a generic post that is not attributable to an individual author? No, thanks. I prefer to be left behind until the the AI bubble bursts.

Here's your wake up call: Human responses are just plagiarized blendings of previously expressed ideas processed by the neural networks in a human brain instead of by an artificial neural network.

 

[pksv]

The amount of people completely unaware of the power of AI is incomprehensible. I feel like I'm in the early days of the Internet again. "The Internet is useless, it's a fad."

 

[TechnoTiger3000]

In the news today: “AI Expected to Replace More Than 200,000 Wall Street Jobs”

AI Expected to Replace More than 200,000 Wall Street Jobs

Discover insights from a Bloomberg Intelligence report predicting that global banks may cut 200,000 jobs in the next few years due to AI advancements.

www.eweek.com

Be careful of your expectations for the tree.
A part of the reason AI is so helpful over a Google search is that you don’t need to wade through all the ads, filler to make room for more ads and “optimized” results. At some point, they’ll monetize it and the AI will start giving you “promoted” answers with a few ads inserted.

You’re right that monetization poses a threat to the current ChatGPT experience, but AI extends far beyond just LLMs.

 

[julzmeister]

Why are we calling this reminders feature "A.I."? There's nothing "A.I." or revolutionary about it.

 

[pksv]

Why are we calling this reminders feature "A.I."? There's nothing "A.I." or revolutionary about it.

Are you aware that it uses LLM to achieve this functionality? LLM is AI.

 

[eltoslightfoot]

I find that it's a matter of how people see the world and their capacity for adaptation to new circumstances.

To some, often those who embrace novelty, it's immediately obvious how incredibly useful these things can be in augmenting your existing skills with new ones or accelerating your work with a partner that you can delegate to.

Others fear change, probably have had the same job forever and aren't capable of adapting to new roles or just simply like the way things are. They have a revolting feeling against anything that'll disrupt it so they refuse to see its usefulness — or simply can't.

The latter group is going to have a wake up call much sooner than they imagine when they finally realize that they're being left behind and have to learn how to use AI while the first group are by then experts in leveraging AI and already way ahead.

It’s more like I can’t believe people have trouble with this stuff and need AI to do it—sometimes incorrectly. Just often incorrectly enough to wonder if it can really be trusted.

It’s fine if I have a file pile of PDFs now that the some AIs are smart enough to source their results.

It’s fine if I need it to produce code in a vacuum.

But when I need to research something new? I think those that trust AI for that new information will:

1. Not get enough depth on the topic to truly understand said topic.
2. Not know what to connect it to.
3. Should always wonder if it is correct anyway.

Also, there is the fact of what to do when the AI fails at the task in question. This is why I think Agents are going to be hilarious when they get going and people start getting boxes from Amazon that they didn’t even want. (As a for instance…)

So, no, I just don’t think it is good enough to do anything real. It is, however, novel enough for bosses to think it can to real things, fire their staff, and then pikachu-faced shock when it all blows up in their face.

But fear? Nah, I have been messing with SLMs and bots for python. Not afraid. Just understand that it’s all a little silly for most of these personal things.

I mean, how hard is it to go, “Hey Siri, remind me to wash the clothes at 5 PM today?” Not hard at all.

 

[eltoslightfoot]

Wait hold up folks. Is this a LLM text-based front end for the Fantastical app (which already has natural language input)? Wow! And it only costs an additional $20 month! Game changer!

I think someone who really understands AI should explain to the rest of us why AI seems to be great at programming and pretty much ***** at everything else.

It’s not really great at programming beyond the basics. Once you get into creating something useful out of it, its out of its depth.

 

[macfacts]

The amount of people completely unaware of the power of AI is incomprehensible. I feel like I'm in the early days of the Internet again. "The Internet is useless, it's a fad."

No one important ever said the internet is useless.

 

[SBlue1]

The amount of people completely unaware of the power of AI is incomprehensible. I feel like I'm in the early days of the Internet again. "The Internet is useless, it's a fad."

Or when old folks were telling me 40 years ago that my Amiga computer was just for gaming and nothing else. It started my interest in computers and my creativity exploded.

AI and LLMs are here to stay. This is not a fad like 3D TV or curved monitors, this is more like the first personal computers. Most didn't know what to do with them and thought they were just useful for people working in offices, but some saw the potential and knew that the world would change forever.

 

[SBlue1]

No one important ever said the internet is useless.

40 years ago the internet was just for hackers and nerds and people thought it would be useless for everybody else. Heck, even in the 90s people were saying „Why would I need it?”. Many people struggled to see how the internet would fit into their daily lives, especially if they didn’t work in a tech-related field. Skepticism about the internet being a passing trend or a niche tool was prevalent, with some viewing it as something only for scientists, academics, or hobbyists. There was apprehension about using it due to a lack of knowledge. Concerns about privacy, scams, and viruses were common. Many older adults and less tech-savvy individuals felt overwhelmed by the idea of learning to use computers and navigating the internet. Some people thought it was unnecessary because their current methods (like newspapers, libraries, and phone calls) worked just fine.

 

[pksv]

No one important ever said the internet is useless.

Clifford Stoll
In 1995, Clifford Stoll, an astronomer and author, wrote an article in Newsweek titled “The Internet? Bah!” in which he stated:
“The truth is no online database will replace your daily newspaper, no CD-ROM can take the place of a competent teacher and no computer network will change the way government works”.

Paul Krugman
Paul Krugman, Nobel Prize-winning economist, wrote in 1998:
“The growth of the Internet will slow drastically… By 2005 or so, it will become clear that the Internet’s impact on the economy has been no greater than the fax machine’s”.

Robert Metcalfe
Robert Metcalfe, the inventor of Ethernet, made a bold prediction in 1995:
“I predict the Internet will soon go spectacularly supernova and in 1996 catastrophically collapse”.

 

[MiG007]

Here's your wake up call: Human responses are just plagiarized blendings of previously expressed ideas processed by the neural networks in a human brain instead of by an artificial neural network.

Human responses are just plagiarized blendings? Thought, comprehension, (real-world) experience and reasoning are blendings?

LLMs don't know what the words mean. Per ChatGPT today:

"My responses are generated based on statistical patterns in language rather than reasoning or understanding in the way humans do.

So, while I can simulate intelligent behavior and offer useful information, it's all rooted in algorithms and data, not true awareness or reasoning."

 

[MiG007]

Are you aware that it uses LLM to achieve this functionality? LLM is AI.

I would respond - marketing aside - LLM is LLM.

 

[MiG007]

The amount of people completely unaware of the power of AI is incomprehensible. I feel like I'm in the early days of the Internet again. "The Internet is useless, it's a fad."

For me it is not the rise of the machines/LLM that I have an issue with. It is that, as usual, the tech industry never fixes anything, they just move onto the next shiny object -- 5G, Google Glass, VR, AR, NFT...

Why are we still dealing with spam email? Why hasn't that been fixed at the root level instead of monetizing it with service subscription after service subscription that is supposed to help? Or to put it simply, blaming the user. sort of like asking me to put yet another lock on my door.

How many deadbolts does my ISP/Apple/Google/Microsoft want on my door?

Why are we still dealing with spoofing on cell phones?

Why are we still dealing with malicious texts?

I wonder if LLMs will make it even more difficult to figure out truth from fiction.

 

[vantelimus]

Human responses are just plagiarized blendings?

Pretty much yes at the neural level. The main differences are 1) biological neurons update in real time, whereas GPT's neurons are fixed between lengthy training phases; 2) the brain contains ~86 Billion neurons, whereas GPT has on the order of hundreds of millions of neurons; and 3) the structure and connections in the brain are more complicated and only partially understood, whereas the neurons in a GPT are connected in more orderly patterns allowing us to express their actions as algorithms.

Concepts such as thought, reasoning, experience, and awareness are emergent properties of the action of biological neurons. Once those are understood sufficiently there is nothing in principle that would stop us from being able to implement them artificially. We are not currently at the level of understanding of the human brain to be able to do that and our computational capabilities would not allow it to happen in real time. However, these are merely technological limitations of capacity and magnitude. It should amaze you what ANNs can accomplish given these limitations. You should use all of your capabilities of thought, reasoning, experience, and awareness to imagine what happens when those limitations are lifted.

 

[MiG007]

Pretty much yes at the neural level. The main differences are 1) biological neurons update in real time, whereas GPT's neurons are fixed between lengthy training phases; 2) the brain contains ~86 Billion neurons, whereas GPT has on the order of hundreds of millions of neurons; and 3) the structure and connections in the brain are more complicated and only partially understood, whereas the neurons in a GPT are connected in more orderly patterns allowing us to express their actions as algorithms.

Concepts such as thought, reasoning, experience, and awareness are emergent properties of the action of biological neurons. Once those are understood sufficiently there is nothing in principle that would stop us from being able to implement them artificially. We are not currently at the level of understanding of the human brain to be able to do that and our computational capabilities would not allow it to happen in real time. However, these are merely technological limitations of capacity and magnitude. It should amaze you what ANNs can accomplish given these limitations. You should use all of your capabilities of thought, reasoning, experience, and awareness to imagine what happens when those limitations are lifted.

So as a young kid, when I touched the hot burner on the stove and learned not to touch it again, that was a plagiarized blending?

 

[vantelimus]

So as a young kid, when I touched the hot burner on the stove and learned not to touch it again, that was a plagiarized blending?

We were talking about the generation of ideas. You want to bring in other aspects of being human. That's fine.

The situation you describe was part of your training in which autonomic functions in your body gave you a pain input in response to a particular action. A neural network was modified in your brain to keep you from doing that again. You learned.

Some people, those with congenital insensitivity to pain, wouldn't learn from the immediate experience. It would take longer to train them. They might need to learn more elaborate procedures to keep themselves from harm. That they don't have the same inputs as you doesn't diminish their ability to think or produce responses to questions.

That you have more senses for input than a GPT and that your training systems are more real-time and honed by evolution doesn't mean that your neural networks give rise to fundamentally different abilities than those of artificial neural networks trained in analogous fashions.

 

[eltoslightfoot]

40 years ago the internet was just for hackers and nerds and people thought it would be useless for everybody else. Heck, even in the 90s people were saying „Why would I need it?”. Many people struggled to see how the internet would fit into their daily lives, especially if they didn’t work in a tech-related field. Skepticism about the internet being a passing trend or a niche tool was prevalent, with some viewing it as something only for scientists, academics, or hobbyists. There was apprehension about using it due to a lack of knowledge. Concerns about privacy, scams, and viruses were common. Many older adults and less tech-savvy individuals felt overwhelmed by the idea of learning to use computers and navigating the internet. Some people thought it was unnecessary because their current methods (like newspapers, libraries, and phone calls) worked just fine.

I was there and no one said this. No one.

 

[eltoslightfoot]

Pretty much yes at the neural level. The main differences are 1) biological neurons update in real time, whereas GPT's neurons are fixed between lengthy training phases; 2) the brain contains ~86 Billion neurons, whereas GPT has on the order of hundreds of millions of neurons; and 3) the structure and connections in the brain are more complicated and only partially understood, whereas the neurons in a GPT are connected in more orderly patterns allowing us to express their actions as algorithms.

Concepts such as thought, reasoning, experience, and awareness are emergent properties of the action of biological neurons. Once those are understood sufficiently there is nothing in principle that would stop us from being able to implement them artificially. We are not currently at the level of understanding of the human brain to be able to do that and our computational capabilities would not allow it to happen in real time. However, these are merely technological limitations of capacity and magnitude. It should amaze you what ANNs can accomplish given these limitations. You should use all of your capabilities of thought, reasoning, experience, and awareness to imagine what happens when those limitations are lifted.

This is quite entertaining. That people think the probabilistic based AI will be able to do these things on the basis of artificially defined "neurons" and comparing them to the brain when we barely understand the brain.

 

[SBlue1]

I was there and no one said this. No one.

I was there too. People thought I was a hacker when I was telling them how convenient the internet was.