>So obviously, error correction with inputs/outputs is not the way we get to intelligence.
This doesn't seem to follow at all let alone obviously? Humans are able to reason through code without having to become a completely discrete computer, but probably can't reason through any length of assembly code, so why is that requirement necessary and how have you shown LLMs can't achieve human levels of competence on this kind of task?
Yes, this absolutely appears to be the main reason. Both in practical terms through birth control, but also through cultural terms in that it's now seen as a choice rather than as an obvious thing you do. To change this course, we probably need to change the culture first so that a birth control ban will be supported. That's currently not looking likely, so population collapse it is
Well you see that take over and over because that's what people actually believe in and feel and it's almost weird it has to be repeated over and over. Most protestors are not solidarity protestors. Most protestors show up when they're angry, when they feel like there's something obvious that can be changed and when people around you refuse to do the obvious thing. If you don't have these factors, you don't really get big protests.
For example I don't believe the US saw particularly large scale anti Germany protests surrounding WW2. Before the US joined the war people didn't really know what to do, while after they joined the war there was little disagreement. The Vietnam protests were much larger, because you have the internal conflict and something obvious to do: stop fighting.
People showed up for Gaza protests because they were angry and because they felt people around them, and particularly their governments were complicit in events. People do not show up for Iran because everyone agrees it's terrible but no one really knows what to do, so who are you going to be yelling at on the streets and what would you yell? Additionally events in Iran, relatively speaking probably triggers more hopelessness/confusion than anger, these are not exactly the best emotions to inspire protest
Advanced intelligence may have evolve multiple times, but wouldn't the origins of simpler intelligence lie much deeper in the evolutionary tree? If Octopi use neurons too, it seems obvious to me that rudimentary intelligence must have originated in or before the common ancestor of vertebrates, octopi and squids: flatworms. Or going back even further, perhaps even all the way to single cellular life which often seems to be able to react in complex ways to stimuli. Even our brains seem to still make use of forms of processing within the cell, isn't there intelligence in those cells? Or do we have some agreed on definition of intelligence that excludes these simpler forms?
You probably also need at least:
- Y does not appear when X does not
- We need an overwhelming sample size containing examples of both X and not X
- The experiment and data collection and trivially repeatable (so that we don't need to rely on trust)
- The experiment, data collection and analysis must be easy to understand and sensible in every way without leaving room for error
And as another commenter already pointed out: You can't really eradicate the existence of an unknown Z
And even if you do know there's causality (eg: the input variable X is part of software that provides some output Y), the exact nature of the causality can be too complex to analyze due to emergent and chaotic effects. It's seldom as simple as: an increase in X will result in an increase in Y
I'm curious if we'll see a world where computers could solve math problems so easily, that we'll be overwhelmed by all the results and stop caring. The role of humans might change to asking the computer interesting questions that we care about.
I'd like to recommend the author mentioned briefly in the article on this topic: Joel Mokyr. Unlike how this article paints him, Joel doesn't really point to a sole cause for the industrial revolution, but highlights a broad range of contributing factors, I thought it was very insightful.
While it's certainly not the only cause, high wages as a contributing factor to innovation in productivity does still seem like a plausible factor behind the industrial revolution. I suspect that these days in the west, labor is relatively so cheap compared to how much capital is around, that capital ends up being rather inefficiently used. Or at least, capital doesn't primarily go to production increases anymore. Perhaps there's avenues for gains here today
There's so much ambiguous context that goes into your average '=', even when just talking about your standard functions on real numbers. You'll see it being used for anything from:
- We've found these to be equal
- We're hypothesizing this to be equal
- These are approximately equal
- These are defined to be equal
- This is a way to calculate something else, whether it's equal is up to your philosophy (a^2+b^2=c^2)
- I'm transforming my function into something else that looks different but is exactly the same
- I'm transforming my function into something else that is the same for some of the stuff I care about (but for example does not work anymore for negative numbers, complex nrs, etc.)
- I'm transforming my function into something else, but it's actually a trapdoor, and you can't convert it back.
- This is kind of on average true within an extremely simplified context or we know it's not true at all, but we'll pretend for simplification (looking at you physics)
- We are trying to check if these two are equal
- This is equal, but only within a context where these variables follow some constraints mentioned somewhere else entirely
- This is equal, but, we're not going to say whether you can or can't replace the variables with functions or whether it supports complex nrs, negative nrs, non-integers, etc.
A lot of this is usually kind of clear from context, but some of these differences are a nightmare if you want to code it out
I wonder if there'd be any good way to confirm or rule out an ancient Martian civilization. Would Martian surface structures be able to last a few billion years? Or would Martian processes be able to make all of them disappear? And how about underground structures?
It's just a preference thing, I loathe the small program chaining style and cannot work with it at all. Give me a python script and I'm good though. I can't for the life of me imagine why people would want to do pseudo programming through piping magic when chaining is so limited compared to actual programming
To my untrained eye, the eastern United States and Uruguay + surroundings also seem like great places for civilizations to thrive. But for some reason those areas did not seem to have particularly large civilizations before the colonization period. Does anyone know of a good reason why these areas were relatively unpopulated compared to the areas you mention but also compared to say Mexico?
It's a cut through the scroll, with the time dimension in this video representing the location of the cut along the scroll lengthwise.
As you can see from the mess it's far from trivial to find the surface of any of the sheets in the scroll, often they layers are blended together messes.
You may be thinking of the scans they used of an unwrapped sheet, those were as you describe and were used to help figure out methods for the real challenge.
We need supermarkets to provide places where people can donate or sell their excess reusable shopping so other people can pick them up and use. That should put a big dent in the necessity for people to make use of new bags if they didn't bring their own reusable bags
>but the catastrophist discourse that there's a deep slowdown in productivity compared to the previous decades and that today we're mostly relying on things we already did in the past maps pretty well with my perception.
Mine too. Any ideas why productivity has slowed down so much? Is this a software online observation, or more of a wider pattern
With software I guess part of the problem is that there's a lot of diminishing returns, with it being much easier to just crank out some initial prototype that does a lot, then it is to adjust some spaghetti mess that users deeply rely on. But that should only really affect individual projects, not necessarily the industry as a whole
This doesn't seem to follow at all let alone obviously? Humans are able to reason through code without having to become a completely discrete computer, but probably can't reason through any length of assembly code, so why is that requirement necessary and how have you shown LLMs can't achieve human levels of competence on this kind of task?