nowittyusername

10 ms·

91.
by nowittyusername 11mo ago
Glad to see evolution of proper context management. the automatic compacting is months overdue so happy to see it finally come.
92.
by nowittyusername 11mo ago
True innovation will come from open source for sure. As the developers don't have the same economic incentives to be "safe", "ethical" "profitable" or whatever. large corporations know this and fear this
93.
by nowittyusername 11mo ago
Yes, holy shit yes.... This has been my sentiment for a while now. I see lots of people everywhere especially on social media talking about the thing and no one actually did the thing or is even attempting to do the thing. Honestly thats w
94.
by nowittyusername 11mo ago
i feel ya. I started working on AI agents and because the whole thing was kinda new to me (had lots of experience on generative image side of things with finetuning etc..) I felt that there was a mountain of stuff to learn and no one to bo
95.
by nowittyusername 11mo ago
I was using stable diffusion 1.5 when it came out and had my first LSD trip shortly after that, prolly like a few months. Anyways, what struck me was how similar the closed eyes visuals were on LSD compared to the generated images from sta
96.
by nowittyusername 11mo ago
I recommend use low quantized models first. for example anywhere between q4 and q8 gguf models. Also dont need high context to fiddle around and learn the ins and outs. for example 4k context is more then enough to figure out what you need
97.
by nowittyusername 11mo ago
My man, we now have llms that are anywhere between 130 million to 1 trillion parameters available for us to run locally, I can guarantee there is a model for you there that even your toaster can run. I have a RTX 4090 but for most of my fid
98.
by nowittyusername 11mo ago
I agree with the sentiment but I also recommend you build a local only agent. Something that runs on llama.cpp or vllm, whatever... This way you can better grasp the more fundamental nature of what LLM's really are and how they work
99.
by nowittyusername 11mo ago
Reading this made me think of "Ferengi Rules of Acquisition" quite a lot. Maybe it was inspired by this...
100.
by nowittyusername 11mo ago
I am building my agent and hoard old LLM's like they are a precious commodity. Older models are less censored, more flavorful and don't have that RL slop factor. Of course the newer models have their place inside my agent but th
101.
by nowittyusername 1y ago
People will not get true benefits of these technologies until they own the hardware that the LLMS and its other various systems run on. Also until they are the ones to actually build the AI system or at the least had a community verify it t
102.
by nowittyusername 1y ago
Run for the hills!! Any advanced alien civilization that willingly contacts us should automatically be deemed an existential threat. Because they would be smart enough to know the societal damage they caused by this revelation and they did
103.
by nowittyusername 1y ago
My existing project is very similar to this with some other goodies. I agree with the author that focus on systems versus LLM's is the proper next move. Orchestrating systems that manage multiple different llms and other scripts toget
104.
by nowittyusername 1y ago
Metacognitive AI system. The focus here is on the various internal systems versus the LLM itself. Basically giving an AI agent the ability to do all the things it cant do from a one step turn interaction that you usually see from just a c
105.
by nowittyusername 1y ago
Humans are increasingly living in more and more chaotic environments where they perceive lack of agency (more and more accurately so) and so we crave stability where we have more control. All successful escapes from reality provide that sta
106.
by nowittyusername 1y ago
This is a very common sentiment I see everywhere and it really highlights how uneducated most people are about technology in general. Most folks seem to expect things to work magically and perform physics breaking feats and it honestly baff
107.
by nowittyusername 1y ago
I've been thinking to fully switching to Linux for a while now. But there was just not enough pressure for me to switch over, I think this might do it. I think the universe is done whispering in to my ear and is finally shouting from t
108.
by nowittyusername 1y ago
That's exactly what I am focusing on. IMO getting rid of the biggest pain points regarding script building but also benefitting from modern day AI systems. "Have your cake and eat it too." We know that scripts are far more re
109.
by nowittyusername 1y ago
Your post is making me think maybe there is quite a lot of lost knowledge out there somewhere that maybe has pertinence in modern day agentic AI system building use. I am currently experimenting in building my own AI system that uses LLM&#x
110.
by nowittyusername 1y ago
I must say, at least for me personally when I hear about such levels of incompetence it rings alarm bells in my head making me think that maybe intentional malice was involved. Like someone higher up had set up the whole thing to happen in
111.
by nowittyusername 1y ago
That's pretty nuts, they should make a mini documentary on how they made it and all the hurdles they had to overcome and all that jazz.
112.
by nowittyusername 1y ago
I agree, LLM's are capable of doing this right out of the box if you provide it grounding data like current time and a few other things in the system prompt. Its really odd that this is getting any attention.
113.
Sample Forge – Research tool for deterministic inference in LLM's(github.com)
1 points by nowittyusername 1y ago|1 comments
114.
by nowittyusername 1y ago
Hi folks, I made a research tools that allows you to perform deterministic inference on any local large language model. This way you can test any variable changes and see for yourself the affects those changes have on the output of the LLM&
115.
by nowittyusername 1y ago
your comment reminds me of another one i saw on reddit. someone said they found that using github diff as a way to manage context and reference chat history worked the best for their ai agent. i think he is on to something here.
116.
by nowittyusername 1y ago
I've used Claude Code for about 3 months now. Was a big fan until recent changes lobotomized it. So I switched over to codex about 2 weeks ago and loving it so far, way better experience. Today with the introduction of the new model,
117.
by nowittyusername 1y ago
This is a recurring theme with Google. Their models are phenomenal but the systems around them are so bad that it degrades the whole experience. Veo3 great model horrible website, and so on...
118.
by nowittyusername 1y ago
My recollection is different, its of a half horse half fish. Horse part facing to the left as if its standing on hind legs, horse is brown. Bottom half is blue fish tail.
119.
by nowittyusername 1y ago
Seinfeld episode if I ever seen one, something something Kramer with some side shenanigans...
120.
by nowittyusername 1y ago
The sampling process isn't random. If you sample with identical sampling parameters and identical values for said parameters, you will always get same results. You only start getting "non deterministic" behavior when you sta

More ›