I have a premonition that even if extraterrestrial life is discovered, and even if intelligent, hostile extraterrestrial life is somehow discovered, humans are not going to leave it alone.

After all, everyone involved can just say it's inevitable! If we don't do it, someone else will.

It’s the same, lame, self-fulfilling excuse that gets used for every bad thing someone does dressed up as a necessary thing.

You can see it right now with AI barreling forward. If we don’t develop it, someone else will, so we will ensuring it happens.

Of course we can never prove or disprove their claim of “someone else will” because they went ahead and did it.

You can't prove that walking through walls is impossible but you are an idiot if you keep trying. It hurts. That's called inductive reasoning.

We've never seen a case where some government in some country doesn't work as hard as possible to ensure that they have the biggest possible weapon. So you should believe someone is doing that. (Point in fact, we know of several governments currently trying to achieve nuclear, aerial, and AI superiority today.)

You can check out why scientists use this same philosophy in modern discovery and belief systems by catching up on the subject of "the philosophy of science." You'll find some 5,000 years of scientific debate and conclusions that roughly culminate in the above system of belief.

I’m well aware of why people do it, that doesn’t change my feelings about it or my opinion of it.

Developing the more heinous weapons before the other guy is the pragmatic thing to do, but it doesn’t give you the moral high ground.

> (Point in fact, we know of several governments currently trying to achieve nuclear, aerial, and AI superiority today.)

And why are they doing that? To get it first since others will do it anyway.

In the case of weapons development, it's not about whether someone will do it or not. It's about whether that is a good reason/excuse to then do it yourself

So, if someone else is going to develop something very harmful, is the best solution to develop something equally or more harmful? Or are there maybe other better solutions to this problem?

There might be, but we've never seen one. The culture with the lesser weapons is subjected to inferior quality of life when compared to the oppressing culture in all cases in history.

So, it might be worth entertaining whether you can convince everyone in your culture to be subservient?

Here again, historically that's never happened, so it's not worth believing. And that's just to check if there's a better solution. You might arrive at an even worse outcome.

Revolutions require even bigger weapons, despite sometimes having better long term outcomes. How many people have to die to test our theory?

Genghis Khan created centuries of peace horrifically with bigger guns. The Ottoman Turks similarly did. The US, now. And many more cases that we know of. All of these examples coincidentally allowed multiculturalism, and even more, they enforced it as a matter of law. Just as an interesting side point.

Life is terrible in all cases in history where it's not clear who has the biggest guns. That's always been the demarcation of the beginning of a new regime at the cost of millions of lives.

Revolutions are horrible. That's what we know for sure.

Huh? There have been lots of extremely nasty things which there has been implicit or explicit international agreement not to develop and/or deploy, and this has largely held. Weaponised nerve agents, most biological weapons, salted nuclear weapons (the doomsday machine from Doctor Strangelove, essentially), extremely large nuclear weapons...

Agreement? Never happened.

Sadam Hussein? IRGC? United States?

When?

It's never happened. The US itself doesn't abide by Geneva convention. Let alone other UN members.

You're living in pretense, I'm sorry to point out.

The world is run by terrorists. The only question is which terror group you want to retain power.

Personally, I like the multiculturalism that the US champions, and the Turks used to, and the Khan dynasty used to, and Persia used to. And only because it's the most likely recipe for the longest track record for peace.

Like it or not, history doesn't agree with your optimism as much as you and I wish it did.

I want to clarify my side point. I brought up multiculturalism because I think these societies find that to be a requirement for peace once they've established military superiority.

I want to clarify that because I could see someone taking that statement to mean that multiculturalism is inherently wrong, which is the opposite of what I'm saying.

No, but you might explore differently once you know there are other people everywhere and they might be stronger than you.

In the movies.

In reality the existential crisis will be used as a means for control, exploitation, and corruption ...

c.f. USA, Israel, UAE as spring-to-mind obvious examples for hints and tips as to how this plays out.

It's also interesting that your mind goes immediately to common enemy instead of some alternative.

This is a pretty obscure and in-the-weeds benchmark, but to me the models’ interpretation feels quite reasonable.

Apologies, I didn’t mean to imply it’s a benchmark, I just wanted to provide a reproducible example of where I see models make decisions that seem to be fine initially but might paint the software architecture into a challenging corner. I don’t expect models to read my mind, but I do see them produce a lot of verbose output, none of which is used to say “here’s a simple response to your ask, but have you also considered...”

It's obscure to use common functions from the standard library?

It’s just an odd way to judge whether an entire model “gets it” in my opinion. There are so many variables.

I think knowing which models use the standard library and which ones pull in dependencies as highly useful. As a dev I've always favored staying as close to the standard library as possible as it makes refactoring, understanding, and deletion much easier. Not everyone has the same preferences as you, it's nice to understand what matters for others too.

> When you change your mind, the old line is marked superseded, not deleted

To me, this seems like a design error. You're polluting context with false/outdated information (even if the LLM is instructed to ignore it). The biggest issue with the memory systems built into Claude et al. is that they're terrible at pruning old/conflicting information as the project evolves, so I'd hope a replacement would do something to improve that.

It's such a Claudism.

Everything is inundated with info about other things tried.

Comments and docs flooded with things found out in the process when you want something about the info you need to know now.

I've recently started trying out a little approach that half makes me cringe as I think of gastown, but sharing FWIW

- All decisions get logged to DECISIONS.md, sequentially

- Before writing a decision, read through past decisions to see if any conflicts

- If no conflicts, encode the decision into the CODE.md

- If any conflicts, ask a Tribunal of 3 agents to find a resolution—each of them should be prompted in slightly different ways

I've only done this for one pretty big project but so far it seems to be working well

Keeping history and using history are two different things though

You may want the old decision in the audit trail so you know why something changed without putting that old decision back into the model context every time

I think memory systems need a pretty strict separation between active memory and historical memory

From reading the readme, it looks like superseded decisions are not added to context.

> Next session, the relevant lines are added to Claude's context.

At least that's how I interpret it? If it is adding superseded decisions, that does seem bad.

Is it outdated? Its an avenue already visited, tried and abandoned, so valuable info regarding architectural decisions.

> Its an avenue already visited, tried and abandoned

I suppose if you could guarantee that the nondeterministic model can not only know all of those disparate pieces of information, but connect them together in that order, and arrive at a decision that it was abandoned because it was already visited and tried, every time.

This is an AI-generated ad for some SaaS thing.

But of course it helps.

True, after ingesting a stomach's worth of jet fuel the bird is powered for the rest of its life.

In bird culture, this is considered a dick move

Not really, it’s mainly a defense mechanism against hunters

> Shows the evidence behind every flag, so you can disagree with it

> Logs every Bluetooth device it hears, glasses or not

How much AGI is required before LLMs stop putting totally irrelevant technical details into marketing copy?

I also like:

> Nothing leaves your phone and there is no account.

This is good to know, but it's the exact type of thing Claude is obsessed with putting everywhere once you tell it NOT to do something. For example, if I tell it to commit but never push changes to GitHub, it can't help but leave "Changes committed; nothing was pushed." to EVERY SINGLE MESSAGE. The model just can't help but report back about everything it's NOT doing, and it drives me crazy.

I like those kinds of updates (as long as they are accurate of course)

What’s annoying is the stuff it decides to put into user facing copy, details that they’d never give a shit about, calling everything real, etc

I don't mind if models do that but I don't understand why a human doesn't take 2 minutes to proof and re-word/delete them

I feel as though it’s inherent in the models, as part of “alignment” or whatever. I’ve noticed whenever having an LLM help with data analysis, especially Claude, it just can’t help itself from putting caveats and guides and documentation inline or in footers.

And if you correct it, it will mention "a previous version said..."

None of the libraries in New York City are publicly owned. NYPL, BPL, and QPL are all private nonprofits that receive public funding. It’s an odd arrangement.

The online iD editor is powerful enough for even pretty complex/big edits in my experience, with good basemap options and intuitive controls. I don’t see any need to download JOSM.

Exactly my thoughts. Advising to download and use JOSM to a novice OSM mapper to make their first edit there (to add a single tag, of all things) instead of iD is nonsense to me, just totally unnecessary overkill. I don’t even understand how article author decided it’s a good idea, to be honest.

You can (and manufacturers do) release security updates for past major-version releases without backporting features.

At least for Android that's not a big deal as most of the system lives in independent applications.

I don’t see anything that odd about them. They’re young models of various ethnicities - basically the same people you’d see modeling the clothes at a store like Target. And they’re wearing/doing things that seem relatable to Apple’s core upper-middle-class customer base.