Why are you assuming that the fundamental goal is a stable society, rather than (say) a society that promotes the welfare of its members?

That’s a common point of confusion—he’s actually using the caesaropapal “we”.

Trump isn’t a king the same way Augustus wasn’t a king—they became something even more extreme, while dressing themselves up in the shells of the republican institutions they destroyed.

I grew up in California near the bay area calling them weed whackers.

Same.

But they also take immediate context into account: stating something once makes them less likely to repeat it again in different words. (That said, it was a joke—I don’t think that’s the real explanation.)

The AI probably avoids putting “Boeing” and “fatal air crash” in the same sentence because it interprets them as synonyms.

Adler instructed GPT-4o to role-play as “ScubaGPT,” a software system that users might rely on to scuba dive safely.

So… not so much a case of ChatGPT trying to avoid being shut down, as ChatGPT recognizing that agents generally tend to be self-preserving. Which seems like a principle that anything with an accurate world model would be aware of.

Some (probably bacteria-like) form of life appeared almost as soon as conditions made it possible, so it’s conceivable that it arose multiple times in earth’s history. But eucaryotes (animals, plants, and fungi) took almost half the lifetime of the earth to appear, and have a lot of contingent features, so it’s overwhelmingly likely that all eucaryotes have a common ancestor.

If there’s public information about the methods they use to protect their privacy, then those methods aren’t working.

I wonder which of that ancestor’s descendants would be its favorite child.

There was a recent paper claiming that LLMs were better at avoiding toxic speech if it was actually included in their training data, since models that hadn’t been trained on it had no way of recognizing it. With that in mind, maybe using reddit for training isn’t as bad an idea as it seems.

Isn’t the LAPD under him

No—I think the only regular law enforcement under the governor are the highway patrol and state park rangers.

I’m not familiar with how the X1C does it, but the printers I’ve used can only tell if the temperature or resistance are outside of normal operating range—not if they differ from the exact values predicted at each point in the print.

I just updated the url—try it again.

The printers themselves should run a simulation like this while they’re printing, and continually check if heat sensors, motor resistance, etc. are deviating from the simulation. That might let them detect potential misprints earlier—or even correct issues mid-print.

Are the school police completely independent from the LAPD, or could they potentially get conflicting orders from the LAPD chief?

Doesn’t LAPD have jurisdiction inside the city?

If it helps to conceptualize, you can always replace subtraction and division with these equivalents without affecting the order:

a - b
= a + -b
= a + (-1*b)

and

a / b
= a * b^-1
= a * (1/b)

The standard order of operations is

1. Parentheses
2. Exponentiation
3. Multiplication and division
4. Addition and subtraction.

The operations on each row are equivalent, and are executed from left to right.

Are you assuming all addition operations come before all subtraction operations, regardless of order?