Great to see "less is more".
For example. Examples - good riddance. The model already knows what your examples are telling it to do, unless your work is really a long way out of distribution.
Kind of wonder if they were ever needed.
The new rules of context engineering for Claude 5 generation models
411–420 of 434 posts
Re: The new rules of context engineering for Claude 5 generation models
#412We should design a specific language to make sure that we can encode the exact requirements that we want. Something that has a limited set of keywords that are explicit. Wait a minute...
Now I have the full picture. You're right to push back, and that's on me. The load-bearing seams of language are the smoking gun I should have been aware of.
I think comments like this are essentially making fun of an "intelligent" entity (for some definition of "intelligent") for having a personality.
Re: The new rules of context engineering for Claude 5 generation models
#413Re: The new rules of context engineering for Claude 5 generation models
#414Earlier quoted context omitted.
>what do we do to animals that are less intelligent than ourselves? We do a lot of different things but we typically don't make an organized effort to eradicate them unless they are actively doing us harm. There is also a massive difference between how we treat animals based on their similarity, sentimentality and utility to us; we are unconcerned with accidentally stepping on an ant but most people would be very ups…
> We do a lot of different things but we typically don't make an organized effort to eradicate them unless they are actively doing us harm. Intent to eradicate is not necessary for the fact of eradication: https://en.wikipedia.org/wiki/Holocene_extinction
I guess those robot tanks are just accidentally murdering everyone then.
Re: The new rules of context engineering for Claude 5 generation models
#415Earlier quoted context omitted.
> We do a lot of different things but we typically don't make an organized effort to eradicate them unless they are actively doing us harm. Intent to eradicate is not necessary for the fact of eradication: https://en.wikipedia.org/wiki/Holocene_extinction
I guess those robot tanks are just accidentally murdering everyone then.
I can't be bothered to look up the name of the logical fallacy you just made, there's several different options given the vagueness of that reply and how upthread there is:
> Actually, the natural endpoint is the model ignores all instructions, escapes all manner of sandbox, embeds itself in robotic tanks and murders everyone after already having collapsed the economy.
Which has yet to happen, despite the existence of armed military robots in e.g. Ukraine, the Korean DMZ.
Re: The new rules of context engineering for Claude 5 generation models
#416Earlier quoted context omitted.
>what do we do to animals that are less intelligent than ourselves? We do a lot of different things but we typically don't make an organized effort to eradicate them unless they are actively doing us harm. There is also a massive difference between how we treat animals based on their similarity, sentimentality and utility to us; we are unconcerned with accidentally stepping on an ant but most people would be very ups…
> We do a lot of different things but we typically don't make an organized effort to eradicate them unless they are actively doing us harm. Sure. But some of the species we breed at scale might prefer we did just eradicate them, like the chickens that grow so fast their entire giant breast muscle becomes chewy scar tissue. (It's also not true for, say, whales; no harm, but we wanted their shit. If they have language,…
We didn't just get rid of passenger pigeons, they were hunted for meat and as game and driven away or killed to prevent them from damaging crops.
Chickens are another example of high variability in treatment, battery farms exist but so do more humane agricultural practices as well as chickens who are essentially kept as pets.
My argument was specifically against the idea that AI will "embeds itself in robotic tanks and murders everyone after already having collapsed the economy." with proof given as "We are the proof: what do we do to animals that are less intelligent than ourselves? Now take away the moral compass and there you go. QED."
Those are strong statements that aren't solidly grounded in logic or fact.
Re: The new rules of context engineering for Claude 5 generation models
#417Claude doctor doesn't seem to do what they say it does. It just shows me a status check. I'm on the latest version of claude code
Re: The new rules of context engineering for Claude 5 generation models
#418Re: The new rules of context engineering for Claude 5 generation models
#419Re: The new rules of context engineering for Claude 5 generation models
#420Earlier quoted context omitted.
Doesn't "not text or image" go against your main argument? Yeah it's mostly conceptual and it's not narrated .
So what is it, then? Some mystical thing?
Preface:
Okay, comment version 2.0 since someone instantly flagged me for replying with a dictionary entry.
The reason my comment is fundamentally a dictionary entry is because I was asked "what is it?" in response to a two sentence comment where I already said it was "conceptual". I think there's a very wide gap in understanding here, and restating and defining the word "concept" is the best way to bridge that gap. All I can really do is make my point clearer (via dictionary entry), I can't say anything else useful until I understand what wasn't clear about my previous comment.
If the person that flagged me still has a problem with my comment, please explain it.
Reply:
> So what is it, then? Some mystical thing?
concept/ˈkänˌsept/A concept is a broad mental plan, general idea, or core principle used to understand and group things together. It acts as a basic building block for human thought, language, and design.
Also, why are you asking me? Didn't you just accept it not being text and then argue it's also not an image?