Hacker Newsnew | past | comments | ask | show | jobs | submit | colordrops's commentslogin

Some people don't care and build on top of Linux anyway. This lockdown will accelerate this. At some point a critical mass will eventually be reached, perhaps with the assistance of some corporate entity or organization of some sort that pushes it over the edge. Then there will be a real open competitor. Will take some time though.

Seems like a way to get non-citizen day laborers at super low rates without the liability.

Once all the context that a typical human engineer has to "build software" is available to the LLM, I'm not so sure that this statement will hold true.

But it's becoming increasingly clear that LLMs based on the transformer model will never be able to scale their context much further than the current frontier, due mainly to context rot. Taking advantage of greater context will require architectural breakthroughs.

Will it though? The human mind can hold less context at any one time than even a mediocre LLM. The problem isn't architecture. It's capturing context. Most of it is in a bunch of people's heads and encoded in the physical world. Once it's digitized and accessible through search, RAG, or whatever, the LLM will be able to use it effectively.

Human hold a lot of implicit context, I think far beyond any LLM. Context is not just what you consciously are thinking about in your head

Sure, but so do LLMs models. They have a huge subconscious (the model itself).

Recording every conversation a single person ever had, every book or text or site ever read, everything ever seen, is not a huge amount of data. Microsoft attempted this with a digital camera lanyard but they were too early.


Yeah, but the models are all based on explicit data. I'm saying humans have prior wiring that allows them to extract and keep context that LLMs do not have access to.

So the suggestion here is that RAG, tools, LLM memory, fine tuning, context management etc are not enough to take advantage of all this context? Is there any evidence that these things aren't on a trajectory to be optimized enough to do the job?

Does general intelligence require awareness though? I think you are talking about consciousness, not intelligence. Though to be frank consciousness and intelligence are not well defined terms either.

I predicted something similar a while back:

https://news.ycombinator.com/item?id=31092225


and it has existed for a while already

Could you share some examples?

I'm surprised to realize I'm familiar with most of the stack just from decades of Linux usage and no formal study of the stack.

Any recommendations for a particular weed puller?

I've never used a bad one, although I wouldn't class any of them as anything stellar. All of them have looked like a snake's twisted tongue.

Using them depends on the delicate combination and application of brute force and technique. If your technique and brute force is up to spec, a crowbar works as a makeshift weed puller.




It's kinda gimmicky but I found this thing on clearance at a local hardware store and it works fairly well (gets most weeds out without me having to bend over, which is nice): https://grampasweeder.com/collections/grampas-gardenware/pro...

It doesn't get everything but I can do more work on the tough ones when so many come right out.


My Fiskars Xact makes me look like a pensioner but it works really well: https://www.fiskars.com/en-gb/gardening/products/weeding-too...

Not that it gets to the bottom of my dandelions ofc, but it sure feels like it does!!


Unfortunately many people don't believe anything or even consider a possibility until some authority/expert figure makes a formal sounding statement about it.

Why wouldn't you turn on prompt caching? There must be a reason why it's a toggle rather than just being on for everything.

Writing to the cache is more expensive than a request with caching disabled. So it only makes economic sense to do it when you know you're going to use the cached results. See https://docs.claude.com/en/docs/build-with-claude/prompt-cac...

When you know the context is a one-and-done. Caching costs more than just running the prompt, but less than running the prompt twice.

How long until AOSP deviates so much from features in 3rd party ROMs that it becomes infeasible to for amateur developers to maintain them?


AOSP is used in many contexts like embedded devices where somehow enshackling it would screw up Google's self-interest in other market areas (like ensuring there is a wider population of Android developers).

But regardless, thirdparty ROMs will continue to exist regardless of how much effort it takes because the demand exists and will not merely dissipate.


Demand exists for a lot of things that don't exist. There's no technical reason that the internet couldn't have stayed mostly decentralized, with everyone hosting a node at their house, controlling their own data and feeds, with an experience not much different from current services and social media, and people would absolutely want this if the experience was reliable and had good UX. But so much effort has been shunted to centralized services that it never materialized. The demand is there, I've asked many people. Just one counterexample.

I'm working on a project myself but it's taking forever considering the large scope. Getting close to having it ready for technical individuals to try it out.

https://homefree.host


Consider applying for YC's Winter 2026 batch! Applications are open till Nov 10

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: