The analogy meanders a bit in getting to the point, but I generally enjoyed the journey of the read and also, I think this is starting to look like an increasingly important perspective.
Increasingly I've been asking the LLM to do things that are only seconds of work to do manually on my own. Earlier in the process of learning how to use these things I think that was a somewhat reasonable decision in the name of learning how these things work.
Now... It's maybe the case that getting a bit more practice in with keeping on top of how we do things the manual way ... might still have value? both for the token cost and perhaps for the value that gets unlocked through understanding your systems a little better?
I'm also a little torn since the leaders in the field of "tokens for everything" like Cherny and Karpathy are pretty vocal about not writing any code on their own anymore. While they also don't really pay for tokens the way the rest of us do (I'm out of pocket for mine) is there something to be said for joining that frontier wave of developers?
vLLM apparently already has an implementation of turboquant available - which is said to losslessly reduce the memory footprint required by 6x and improve inference speed by 8x.
From what I understand, the steps are:
1. launch vLLM
2. execute a vLLM configure command like "use kv-turboquant for model xyz"
3. that's it
I've got two kids under 8 years old, a full time job, and a developer-tools project that takes like 105% of my mental interests... so there's been a bit of a challenge finding the time to swap from ollama to vLLM in order to find out if that is true.
SO buyer beware :D - and also - if anyone tries it, please let me know if it is worth the time to try it!
this is pretty interesting - just from the screenshots it looks like an electron app?
have you or had you considered pursuing a vscode extension? (and by extension cursor?)
In my free time I'm also building out a set of agentic tools and I've found that there's so much you get "for free" by working in the space where developers already are.
Git integrations, text editing, highlighting, etc...
Now - I'm still in pre-alpha local development only so who knows what my adoption story will look like after I actually get it onto the vscode marketplace, but I dunno - just making converstaion I guess :D
Anyways, this really seems like an interesting project and I'll take a deeper look outside of office hours! Thanks for sharing.
oooh ... I wonder if it would be sufficient to just write a markdown gramar for TS?! I should ask my AI what it thinks.. I'm sure it'll tell me I'm absoloutely right and a very good and smart boy.
I first leanred about tree sitter a couple months back when I started looking at what was inside the NPM fodler for claude. It's a really cool library.
One of the things it made me think about is whether it made sense for using when editing large markdown files would it be more efficient to convert a document form markdown to DOM then back again for the purposes of editing a large markdown file via code agents? (or a json)
The theory being that agents are always asking me for pemission to use sed in bash to edit markdown files -- could tree-sitter do the same thing using its code-editing capabilities? And would that difference be materially impactful?
Could I lower the token cost of writing an extensive plan by choosing a format that allows me to use tree sitter?
I really haven't explored that much yet since I've been working on other things but it was more just one of those things that make you go hmmmmm... maybe someone else knows :)
this is really good. it aligns closely with what I've been writing on the same subject (but is way better :) )
nice share. thanks.
OH! I commented before I realized it was a product pitch ... literal seconds before she introduced ACE.
If I'm honest, the idea of a collaborative agentic coding interface is pretty novel and interesting. Still impressed with the article just the nature of my appreciation has changed.
> Feels like a fundamental bottleneck for production agent systems, so would love to compare how you're thinking about the latency vs accuracy tradeoff.
I'm really not focusing on latency right now. My short term goal is to prove the thesis that `ail` can improve same-model performance on SWEBench Pro vs. their own published results.
Can I run swebp with GLM-4.6 and get a score better than their published `68.20` https://www.swebench.com/?
The argument is that the latency right now just isn't the part we should worry about. If we're reducing the time to code something from ~6 weeks to 1 hour... then does it really matter tha we add an other 30 minutes of tool calls if we get it 100% right vs. 80% right?
so - my approach is still being built and I'm still very hand wavy around how it is going to come together, but effecively I'm building pipelines of prompts. Rather than running our LLM sequences as long running sessions where the entire context gets loaded on every turn (a recipe for rot), we unlock the ability to introduce a thinking layer at each step in between the process.
So before each turn is sent into the LLM we (potentially) run a local process to assemble a bespoke context of only what is required for that specific turn.
If a tool call is not going to be needed on the prompt, we don't include it in the system prompt on that round.
I'm still formalizing the spec at the moment and think I'm about six months to a year out before I have a full human ready UI running.
Essentially I'm trying to build an artificial neocortex and frontal lobe to provide a complete layer of Executive Function that operates on top of our agents - like Claude Code (or whatever else).
I'm basing the roadmap on the about 100 years of cognitive science. We've legitimately had names for all these failure modes (in humans) since the 1960's. We have observations of what we're witnessing in agents from 1848.
this is a pretty important piece and the research backs you up. Moving that context out of your system prompt dynamically is going to help reduce your lost in the middle effect. Context rots almost immediately. I've got a project that is being built to address this directly as well, but I'm still very early days.
Liu, N. F., Lin, K., Hewitt, J., Paranjape, A., Bevilacqua, M., Petroni, F., & Liang, P. (2024). Lost in the middle: How language models use long contexts. Transactions of the Association for Computational Linguistics, 12, 157–173. https://doi.org/10.1162/tacl_a_00638
Hey! This looks a lot like what I'm working on, from a slightly different angle. I think you're on the right track. In fact, cortex as a name is perfect since you're effectively building the executive function layer for search and selection. I also think rust is the right language to go with.
I'm going do a deeper read of your work in a bit. I'd love it if you took a look at my theory of artificial cognition The YAML of the Mind https://alexchesser.medium.com/the-yaml-of-the-mind-8a4f945a..., dropped in to the `ail` project and let me know what you think.
I just have to get the kids to school and I'll pop back into cortex later
Hey folks, I wrote this. If you're interested in the concepts or pressure testing the ideas a little deeper, please feel free to comment here or reach out directly.
I appreciate that it's pretty long feel free to point your LLMs at it
I'm writing the tool as proof of the spec. Still very much a pre-alpha phase, but I do have a working POC in that I can specify a series of prompts in my YAML language and execute the chain of commands in a local agent.
One of the "key steps" that I plan on designing is specifically an invocation interceptor. My underlying theory is that we would take whatever random series of prose that our human minds come up with and pass it through a prompt refinement engine:
> Clean up the following prompt in order to convert the user's intent
> into a structured prompt optimized for working with an LLM
> Be sure to follow appropriate modern standards based on current
> prompt engineering reasech. For example, limit the use of persona
> assignment in order to reduce hallucinations.
> If the user is asking for multiple actions, break the prompt
> into appropriate steps (**etc...)
That interceptor would then forward the well structured intent-parsed prompt to the LLM. I could really see a step where we say "take the crap I just said and
turn it into CodeSpeak"
What a fantastic tool. I'll definitely do a deep dive into this.
that's a bit of a meta discussion and it'd probably reveal some super interesting things about how tech culture have changed in the last ~15 years.
I've been on HN since 2010 (lost the password to my first account, alexc04) and I recall a time when it felt like every second article on the front-page was an bold directive pronouncement or something just aggressively certain of its own correctness.
Like "STOP USING BASH" or "JQUERY IS STUPID" - not in all caps of course but it created an unpleasant air and tone (IMO, again, this is like 16 years ago now so I may have memory degredation to some extent)
Things like donglegate got real traction here among the anti-woke crew. There have been times where the venn diagram of 4chan and hackernews felt like it had a lot more overlap. I've even bowed out of discussion for years at a time or developed an avoidance reaction to HN's toxic discussion culture.
IMO it has been a LOT better in more recent years, but I also don't dive as deep as I used to.
ANYWAYS - my point is I would be really interested to see a sentiment analysis of HN headlines over the years to try and map out cultural epochs of the community.
When has HN swayed more into the toxic and how has it swayed back and forth as a pendulum over time? (or even has it?)
I wonder what other people's perspective is of how the culture here has changed over time. I truly think it feels a lot more supportive than it used to.
> This isn’t a bubble inflating. It’s capital and intelligence relocating.
Is this GPT? this kind of not X but Y pattern is a real code smell for slop and can cause someone to immediately bounce out of your writing. The pattern can really feel hard hitting when you're reading what GPT put in front of you, but others are REALLY starting to reject it.
Even people recording short form video are doing it. They're reading out their chat-gpt-psychosis induced fever dreams using scripts written by chatGPT.
The 7-part tweets that build to a slop crescendo are doing my head in.
The solution might be some combination of:
1. leave the social media sites where the slop is irredeemable.
2. unfollow everyone, reset your algorithm.
3. be aggressive about who you add back in. Make sure they're humans having high quality discussions.
4. be aggressive about who you block. Lower the bar on blocking - one and done. No chances, no wait-and-see.
5. move to smaller communities of real humans.
None of this has worked for me yet. I'm still swimming in a vast sea of slopity slop slop. Dead internet theory appears to be playing out in front of us.
edit: On threads I've been trying to use their `dear algo` feature pretty aggressively but it doesn't work very well. I've asked it to remove some types of comments and it seems to just add more of them.
If I could have one of these cards in my own computer do you think it would be possible to replace claude code?
1. Assume It's running a better model, even a dedicated coding model. High scoring but obviously not opus 4.5
2. Instead of the standard send-receive paradigm we set up a pipeline of agents, each of whom parses the output of the previous.
At 17k/tps running locally, you could effectively spin up tasks like "you are an agent who adds semicolons to the end of the line in javascript", with some sort of dedicated software in the style of claude code you could load an array of 20 agents each with a role to play in improving outpus.
take user input and gather context from codebase
-> rewrite what you think the human asked you in the form of an LLM-optimized instructional prompt
-> examine the prompt for uncertainties and gaps in your understanding or ability to execute
-> <assume more steps as relevant>
-> execute the work
Could you effectively set up something that is configurable to the individual developer - a folder of system prompts that every request loops through?
Do you really need the best model if you can pass your responses through a medium tier model that engages in rapid self improvement 30 times in a row before your claude server has returned its first shot response?
According to google Tesla has shipped 8 million cars total since inception.
It is valued at 1.32 Trillion as of today. Which is roughly $165,000 per shipped vehicle.
Increasingly I've been asking the LLM to do things that are only seconds of work to do manually on my own. Earlier in the process of learning how to use these things I think that was a somewhat reasonable decision in the name of learning how these things work.
Now... It's maybe the case that getting a bit more practice in with keeping on top of how we do things the manual way ... might still have value? both for the token cost and perhaps for the value that gets unlocked through understanding your systems a little better?
I'm also a little torn since the leaders in the field of "tokens for everything" like Cherny and Karpathy are pretty vocal about not writing any code on their own anymore. While they also don't really pay for tokens the way the rest of us do (I'm out of pocket for mine) is there something to be said for joining that frontier wave of developers?