I'm sorry you took my response as an attack, it was meant only as a rebuttal.

*Luck* has little to do with the development of LLMs, any more than luck has a role in the development of any major invention. There is a fanciful notion that inventors have a lightning bolt moment. But most of the time, including with LLMs, the invention comes about through a process of many people working steadfastly towards a goal. We had early neural networks decades ago. Today's LLMs did not burst on the scene, they've been coming for a long time.

Actually, I make a little less than average for my position. My house is paid in full, I have no debts. I work from home, so my commute is walking up the stairs to my home office. I live in a quiet, safe, tree-covered community where kids play in the streets and walk themselves to school and neighbors help each other. My coworkers respect me and allow me to do my job. I start work at 8 and stop at 5, and I don't check my emails or Teams messages when I'm not working, including on vacation. In my book, that's a great job.

I did not say you have to move to an urban city center to get a good job, or commute an hour each way. What I said was, you might not find your perfect job in the small town where you grew up. Everybody has a different idea of what a "good" job is, some are indeed located in city centers, some are in suburbs, some are in outlying areas. But if you're not willing to go *there* (wherever "there" is), you are limiting your options.

It's available only on YouTube.

Real research is published in written form, so that it can be reviewed and inspected, and sources and methods examined.

I'm sure there's some good, well-researched material on YouTube. But I have yet to see a case where good, well-researched material is *only* available on YouTube.

It's not if, but when.

Unfortunately, all those data center machines will be obsolete by then, and probably just head for the landfill.

That might be the pitch, but it's far from clear that the cost of AI will be less than...paying wages. AI products are NOT CHEAP, despite all the "free" loss leader AI chatbots like ChatGPT. For business use, each purpose-specific AI requires steep subscription fees, often paid per token.

Your surmises seem reasonable, although you left out the possibility that we simply haven't found the records yet.

As for "straining the world's land use" this is actually false. Despite the world's increasing population, our need for agricultural land is actually decreasing, not increasing. https://ourworldindata.org/pea... And, the world's human population is expected to peak in the 2080s and then start to decline. https://en.wikipedia.org/wiki/... Given those trends, it's likely we will *not* run out of land needed to feed humanity, even if there is a major disaster such as this civilization experienced.

There is *always* a better job. It might not be easy to find, it might require some training, and it might not be in the small town where you live, but it exists. The question is, how much effort are you (or the person in question) willing to spend locating and pursuing that job?

I moved 700 miles from my parents' home at age 18 to find that job, and moved at least that far again three times in my career, in pursuit of a better job. I have a *great* job! Some people have reasons they don't want to move 700 miles to find that better job. That's OK. But that's their choice, that's not the fault of the evil capitalist system. People have a right to choose what's important to them: stay put, or have a great job. If they choose to stay put, they should stop complaining about the consequences of their own decision.

Every Jain wishes more people were Jain. Maybe you should become one. You might not like the dietary restrictions, but go for it.

Attributing human strife to religion, is getting it backwards. People are religious and hold different views and fight with each other, because that is the nature of humans, not because they are religious. Nations and rulers that are atheistic, such as Marxism, have not been less warlike than those who are religious. So let's place the blame where it belongs, on the *human* desire to control other people.

I don't dispute that the droughts may have been a significant factor in the disappearance of the civilization. What remains a mystery, and is not answered by this study, is why it wasn't recorded.

As for unmodified lands being "only" in the "worst" places, makes me think you live in a big city and don't go out into the wilderness much. The US alone has vast areas of wilderness and untamed land. I live in Texas, where you can drive for hours without seeing a single structure or plowed field. The land isn't bad land, it just hasn't been harnessed yet. The western US, even outside the Rockies, is largely unused land. The US government owns almost half of that land, much of that half is not appropriated for agriculture or any other use.

It's easy to think the world is full, when you live in a city.

That makes sense, though we still don't know why this particular civilization disappeared without a record of what happened. Long droughts would not prevent records from being recorded and preserved.

It's easy to think that the world is "full" now. But the reality is that only a small percentage of the earth's surface has been "modified" by humans. https://www.weforum.org/storie... If we needed to move, there are still vast untouched tracts of land that could be tamed.

Anything as complex as history, has many, many inputs. I accept that the droughts might have contributed. But if the droughts hadn't happened, can the researchers really say with confidence that the civilization would still be with us today?

THIS is the central point.

We *do* understand LLMs well enough to build all kinds of things from that understanding. It's not just AI chatbots. There are LLMs tailored to coding, generating images, generating videos, finding security vulnerabilities, screening resumes, taking notes in meetings, the list is endless. It's not necessary to understand every nuance at the deepest level, it's only necessary to understand enough to do useful things with it, and to mitigate risks. That principle holds true for *all* of engineering.

I don't disagree. Poor people do have more trouble paying for necessities. It's kind of what it means to be...poor. But that doesn't invalidate the CPI.

Consider housing, for example. When I was young and poor (and yes, I *was* poor), I made ends meet by...getting a roommate. That's still a thing. My boys both did that as they were starting out, just a handful of years ago. It's *not* a travesty to be forced to get a roommate. You don't want a roommate? Get a better job! That's a good thing! Each person can decide how much effort they want to put into getting a better life. And if they put the effort in, it's still very possible to get that better life.

Reminds me of a video of a Chinese CNC supplier having to reposes machines when the customers stopped making payments; and the supplier is having trouble selling off this lightly used inventory.

You have hit on the key point here.

Both your version of "simulation" and mine, correctly fit the definition of the word *simulation*, which is "imitation of a situation or process." https://www.google.com/search?...

What you are getting at, is that there are vastly better and worse imitations that might be called simulations. The one you are referring to is apparently a high fidelity, large-scale simulation, while my carnival ride is a low-fidelity, low-accuracy simulation. It's just enough of an imitation to give the audience the effect of rain. But imitation it certainly is.

What you are also apparently asserting, is that AI is an example of a large-scale, high-fidelity simulation. That is certainly not the case. If you have used AI at all to do non-trivial tasks, such as GitHub Copilot, you know that the quality and depth of the simulation is very, very low, though it is quite impressive...kind of like the carnival ride.