[dead]
The new skill in AI is not prompting, it's context engineering
451–460 of 550 posts
Re: The new skill in AI is not prompting, it's context engineering
#452The central argument is that the importance of prompt "tricks"—essentially empirical workarounds for current LLM limitations will decline as the technology matures.
The truly valuable and future-proof skill is "context engineering". This focuses on providing the LLM with the information required to reason through the task at hand. Although current LLMs present a trade-off between the size of the context and the quality of the output, this is a constraint that we can expect to lessen with future advancements.
Re: The new skill in AI is not prompting, it's context engineering
#453Context “engineering” likely should involve knowing how the llm treats context size, say needle in hay stack performance, how context size affect hallucination rate, when to summerize context instead of entering the full thing.
Re: The new skill in AI is not prompting, it's context engineering
#454Just like the phasing out of prompt engineering, context engineering will phase out in around 6 months
Re: The new skill in AI is not prompting, it's context engineering
#455Earlier quoted context omitted.
I think making one would help you understand that they're not intelligent.
OK I, like the other commenter, also feel stupid to reply to zingers--but here goes. First of all, I think a lot of the issue here is this sense of baggage over this word intelligence--I guess because believing machines can be intelligent goes against this core belief that people have that humans are special. This isn't meant as a personal attack--I just think it clouds thinking. Intelligence of an agent is a spectru…
So according to your extremely broad definition of intelligence, also a casio calculator is intelligent?
Sure, if we define anything as intelligent, AI is intelligent.
Is this definition somehow helpful though?
Re: The new skill in AI is not prompting, it's context engineering
#456Earlier quoted context omitted.
Sometimes I wonder if LLM proponents even understand their own bullshit. It's all just tokens in the context window right? Aren't system prompts just tokens that stay appended to the front of a conversation? They're going to keep dressing this up six different ways to Sunday but it's always just going to be stochastic token prediction.
System prompts don't even have to be appended to the front of the conversation. For many models they are actually modeled using special custom tokens - so the token stream looks a bit like: translate to English An explanation of dogs: ... The models are then trained to (hopefully) treat the system prompt delimited tokens as more influential on how the rest of the input is treated.
> The models are then trained to (hopefully) treat the system prompt delimited tokens as more influential on how the rest of the input is treated.
I can't find any study that compares putting the same initial prompt in the system role versus in the user role. It is probably just position bias, i.e. the models can better follow the initial input, regardless of whether it is system prompt or user prompt.
Re: The new skill in AI is not prompting, it's context engineering
#457Earlier quoted context omitted.
It's magical thinking all the way down. Whether they call it now "prompt" or "context" engineering because it's the same tinkering to find something that "sticks" in non-deterministic space.
We used to define a specification. In other words; context. But that was like old man programming. As the laws of physics changed between 1970 and 2009.
Is this Haiku?
Re: The new skill in AI is not prompting, it's context engineering
#458Prompt engineering was just trying to fit all the context into one prompt - but actually there would often be a series of prompts both positive and negative so...
I get coining a new term and that can be useful in itself but I don't see a big conceptual jump here.
Re: The new skill in AI is not prompting, it's context engineering
#459Prompt engineering was just trying to fit all the context into one prompt - but actually there would often be a series of prompts both positive and negative so... I get coining a new term and that can be useful in itself but I don't see a big conceptual jump here.
It's not a big contextual jump. It's trying to solve for the problem where a lot of people think "prompt engineering" means "typing a prompt into a chatbot" - and the related problem that many people haven't yet realized you can (and should) dump documents, examples and other long-form content into an LLM to get good results.
Re: The new skill in AI is not prompting, it's context engineering
#460These discussions increasingly remind me of gamers discussing various strategies in WoW or similar. Purportedly working strategies found by trial and error and discussed in a language that is only intelligible to the in-group (because no one else is interested). We are entering a new era of gamification of programming, where the power users force their imaginary strategies on innocent people by selling them to the eq…
Having been a computer scientist and avid WoW player, I dislike this take. The best strategies always have a justification.