News for hackers. Stuff that matters.

A comment on The new skill in AI is not prompting, it's context engineering.

Re:The new skill in AI is not prompting, it's context engine... by niemandhier on Tuesday July 01, 2025 @07:45AM (#44432926)
LLM agents remind me of the great Nolan movie „Memento“.

The agents cannot change their internal state hence they change the encompassing system.

They do this by injecting information into it in such a way that the reaction that is triggered in them compensates for their immutability.

For this reason I call my agents „Sammy Jenkins“.

[ Parent | on HN ]
Re:The new skill in AI is not prompting, it's context engine... by StochasticLi on Tuesday July 01, 2025 @11:02AM (#44434596)
I think we can reasonably expect they will become non-stateless in the next few years.
[ Parent | on HN ]
Re:The new skill in AI is not prompting, it's context engine... by roflyear on Tuesday July 01, 2025 @11:25AM (#44434865)
Why?
[ Parent | on HN ]
Re:The new skill in AI is not prompting, it's context engine... by StochasticLi on Tuesday July 01, 2025 @02:08PM (#44436571)
That's where the research is going.
[ Parent | on HN ]
Re:The new skill in AI is not prompting, it's context engine... by tdaltonc on Wednesday July 02, 2025 @12:43PM (#44445858)
If agents are stateful a few years form now it will be because they accrete a layer of context engineering.
[ Parent | on HN ]