Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

What Is AGI-Hard

lspace.swyx.io

Post reply on HN
Print view View original

111–120 of 129 posts

Re: What Is AGI-Hard

#111

Post by swyx »

ben7799 wrote: post #100

I find it interesting that everyone is arguing about AGI-hard or AI-complete or whatever and trying to find where those boundaries are. Those are interesting problems for sure. But that's not what he's talking about, it's not the point. The scary thing and the important point in the article is that most of the press and non-technical people are being sold the idea that nothing is AGI-hard or requires Human Intelligen…

> But that's not what he's talking about, it's not the point. The scary thing and the important point in the article is that most of the press and non-technical people are being sold the idea that nothing is AGI-hard or requires Human Intelligence, whatever that is.

wow, you phrased it better than i did. thank you!

Re: What Is AGI-Hard

#112

Post by vlovich123 »

tlb wrote: post #37
vlovich123 wrote: post #31

Here are simpler definitions I think: An AI system is General and “adult” level if it can improve its own performance on some task without human support to be taught how to do that. In other words, can it correctly decide when it needs practice (skill acquisition) vs knowledge acquisition (filling information gaps) on its own and figure out how to acquire that (or make clear, explicit, detailed, requests for what it…

Any reinforcement learning system meets your first definition. RL systems don’t typically use separate practice sessions, though you could create one that did. But the need for deliberate practice maybe be just a biological brain thing. Instead, RL systems usually have a parameter for how much to explore new options vs. taking what seems to be the best. This lets them learn continuously as they go.

RL systems lack the autonomy aspect I lay out in terms of finding new problems, figuring out how to improve on the task, or communicating that information in any way. What your describe is necessary but not sufficient.

Re: What Is AGI-Hard

#113

Post by bmitc »

mannykannot wrote: post #80
bmitc wrote: post #60

Earlier quoted context omitted.

> Humanity has developed technology to improve humanity. I would argue that the vast majority of technology has hurt humanity and that a large proportion of technology is simply solving problems caused by other technology.

I am not entirely convinced you really hold this view, but if so, you could reduce technology's pernicious effects on yourself by not participating in this discussion.

Is this a discussion, or are you just being spiteful? Because that's a little much, isn't it?

Re: What Is AGI-Hard

#114

Post by bmitc »

jcelerier wrote: post #105
bmitc wrote: post #60

Earlier quoted context omitted.

> Humanity has developed technology to improve humanity. I would argue that the vast majority of technology has hurt humanity and that a large proportion of technology is simply solving problems caused by other technology.

feel free to go back live 2500 years ago

What is that an argument for?

The amount of technology developed by humans is highly skewed to the past couple hundred years. In that couple of hundred years, more people are starving today, right now, than there were people alive just a couple hundred years ago. One has to ask, why is that? I would argue that our technological explosion and industrialization are key components to this. For example, there's the idea that technological development feeds inequality.

Re: What Is AGI-Hard

#115

Post by Retric »

bmitc wrote: post #114
jcelerier wrote: post #105

Earlier quoted context omitted.

feel free to go back live 2500 years ago

What is that an argument for? The amount of technology developed by humans is highly skewed to the past couple hundred years. In that couple of hundred years, more people are starving today, right now, than there were people alive just a couple hundred years ago. One has to ask, why is that? I would argue that our technological explosion and industrialization are key components to this. For example, there's the idea…

> more people are starving today, right now, than there where people alive just a couple hundred years ago.

This is false, more people are in extreme poverty today but that isn’t starvation as witnessed by these people going not dying off quickly. 99% of people living a couple hundred years ago where living in what we would consider extreme poverty so that’s basically complaining we have more people today.

People without access to modern medicine, safe drinking water, sanitation facilities, and education used to more or less just be people.

Re: What Is AGI-Hard

#116

Post by mannykannot »

bmitc wrote: post #113
mannykannot wrote: post #80

Earlier quoted context omitted.

I am not entirely convinced you really hold this view, but if so, you could reduce technology's pernicious effects on yourself by not participating in this discussion.

Is this a discussion, or are you just being spiteful? Because that's a little much, isn't it?

It is, and for that I apologize. I must admit that I strongly doubted you truly hold this view, and suspected you were just saying it here in order to be argumentative, but now I see you do. I realize that it is not strictly inconsistent to both take advantage of technology and hold that its net effects are harmful.

Even so, it is beside the point - humanity has demonstrably been applying its intelligence to improving technology's capabilities, which is a fact regardless of whether it has been beneficial, and that is enough for it being a candidate measure of machine intelligence.

I realize that there's an argument saying that if the machines start improving themselves of their own volition, it would be the beginning of the end for humanity, and I do not think that it can be summarily dismissed. If it turns out to be correct, you would be vindicated! - though neither of us want that.

Re: What Is AGI-Hard

#117

Post by bmitc »

Retric wrote: post #115
bmitc wrote: post #114

Earlier quoted context omitted.

What is that an argument for? The amount of technology developed by humans is highly skewed to the past couple hundred years. In that couple of hundred years, more people are starving today, right now, than there were people alive just a couple hundred years ago. One has to ask, why is that? I would argue that our technological explosion and industrialization are key components to this. For example, there's the idea…

> more people are starving today, right now, than there where people alive just a couple hundred years ago. This is false, more people are in extreme poverty today but that isn’t starvation as witnessed by these people going not dying off quickly. 99% of people living a couple hundred years ago where living in what we would consider extreme poverty so that’s basically complaining we have more people today. People wit…

It is not false. There’s data to back it up. I did use the word starving, which has a gradient of meanings (see below for separate numbers for the gradients), and I did not mean dying of starvation. The latest numbers are (definitions of state in the links):

* 828 million people are hungry

* 345 million are experiencing acute hunger

* 924 million experience severe food insecurity

* 2.3 billion people experience moderate or severe food insecurity

* at least 9 million people die of starvation every year

https://www.fao.org/newsroom/detail/un-report-global-hunger-...

https://www.concernusa.org/story/world-hunger-facts/

https://www.who.int/news/item/06-07-2022-un-report--global-h...

https://www.un.org/en/chronicle/article/losing-25000-hunger-...

World populations:

* 1850: 1.2 billion

* 1800: 959 million

* 1750: 629 million

https://www.wolframalpha.com/input?i=world+population+1750

Wealth concentration has dramatically increased since the 1970s and 1800s.

Re: What Is AGI-Hard

#118

Post by bmitc »

mannykannot wrote: post #116
bmitc wrote: post #113

Earlier quoted context omitted.

Is this a discussion, or are you just being spiteful? Because that's a little much, isn't it?

It is, and for that I apologize. I must admit that I strongly doubted you truly hold this view, and suspected you were just saying it here in order to be argumentative, but now I see you do. I realize that it is not strictly inconsistent to both take advantage of technology and hold that its net effects are harmful. Even so, it is beside the point - humanity has demonstrably been applying its intelligence to improvin…

I do hold the opinion, and I’m continually working it out and finding information that quantifies it. I at least posted some numbers in an adjacent comment.

I know it sounds extreme, but I can’t see any way around it. Stress levels, depression and anxiety, food problems, wealthy inequality and concentration, loneliness, and more are hitting all time highs by no small margin. And a lot of this has occurred over the past 100-200 years. I think that even without further investigation, one has to admit a strong correlation to industrialization. Societal changes, I would argue, are largely driven by technological developments.

And this ignores environmental and non-human animal suffering. Orcas in the Pacific Northwest are riddle with toxins and starve directly due to human technology. The waters off the cost of India and Thailand are so hypoxic that fish must swim at the top surface of the water, which has led to dramatic changes in whale feeding behavior. This is again due to technology. Again due to technology, we have decimated forests. These numbers are from memory, but I believe we have decreased worldwide forest coverage from around 75% in the early 1900s to around 25% today (these are rough numbers from memory).

The numbers paint a pretty grim picture. Yes, a lot of medical technology is wonderful, but from everything I can see, industrialization is generally not causing harmful trends to go down. The argument isn’t that all technology is bad.

Re: What Is AGI-Hard

#119

Post by Retric »

bmitc wrote: post #117
Retric wrote: post #115

Earlier quoted context omitted.

> more people are starving today, right now, than there where people alive just a couple hundred years ago. This is false, more people are in extreme poverty today but that isn’t starvation as witnessed by these people going not dying off quickly. 99% of people living a couple hundred years ago where living in what we would consider extreme poverty so that’s basically complaining we have more people today. People wit…

It is not false. There’s data to back it up. I did use the word starving, which has a gradient of meanings (see below for separate numbers for the gradients), and I did not mean dying of starvation. The latest numbers are (definitions of state in the links): * 828 million people are hungry * 345 million are experiencing acute hunger * 924 million experience severe food insecurity * 2.3 billion people experience moder…

You misunderstood the statistics.

“As many as 828 million people were affected by hunger in 2021”

This is not equivalent to saying 828 million people are experiencing hunter at the same time. Your conflating the two is exactly the way these statistics get severely distorted.

“Figures on actual starvation are difficult to come by, but according to the Food and Agriculture Organization, the less severe condition of undernourishment currently affects about 842 million people, or about one in eight (12.5%) people in the world population.[6] “ https://en.wikipedia.org/wiki/Starvation

Inequality isn’t the same as starvation, in the last 40 years inequality grew while starvation dramatically declined. “In 40 years, the proportion of malnourished people in the developing world has been more than halved. The proportion of starving people has decreased even faster.”

Re: What Is AGI-Hard

#120

Post by MichaelRazum »

tgv wrote: post #99
MichaelRazum wrote: post #92

Earlier quoted context omitted.

Agree, solved is not well defined. Let's just say it becomes hard to distinguish it from a human. For robotics, I just think of a "simple" tasks like picking up an object and moving it to some place. Haven't seen yet something like this in robotics, that looks human or animal like.

Not to just contradict, but computers playing chess or video games is also distinct from human play. I'm absolutely not expecting strong AI (as "AGI" used to be called) to be realized within 100 years, but when it does, it will have distinctly non-human qualities. Intelligence is shaped by the constraints of the hardware, so to speak, and a computer's constraints will be different. It's been predicted that the first…

Fore sure self-aware AI will be weired. I was rather thinking on a simple metric where AI seems to be far behind in "skills". We see this with FSD and also home robotics haven't took off, because real world object manipulation and navigation is in my opinion where AI currently fails.

If you solve it, then you could maybe claim, ok the AI is more "clever" than a mouse. Currntly wouldn't agree on that. Of course in this area, it is also a hardware issue.

Post reply on HN