nowittyusername

6 ms·

61.
by nowittyusername 9mo ago
This is perfect for me. I just started working on the voice related stuff for my agent framework and this will be of real use. Thanks.
62.
by nowittyusername 9mo ago
here you go I open sourced one of the projects https://youtu.be/EyE5BrUut2o
63.
by nowittyusername 9mo ago
Well why don't you look at it for yourself and tell me if this looks like a beginner project https://youtu.be/EyE5BrUut2o
64.
by nowittyusername 9mo ago
Most people have not fully grasped how LLM's work and how to properly utilize agentic coding solutions. That is the reason for issues when it comes to vibe coders having low quality code. But that is not the limitation of technology bu
65.
by nowittyusername 10mo ago
There is soo much marketing bs around these things it drives me nuts. and it doesn't help that the large labs and credible individuals like denis use these terms. "world models" are video generator with contextual memory but
66.
by nowittyusername 10mo ago
This is the most blatant buy the competition move if i've ever seen one....
67.
by nowittyusername 10mo ago
That phenomenon and others is what made it obvious that COT is not its "thinking". I think COT is a process by which the llm expands its processing boundary, in that it allows it to sample over a larger space of possibilities. So
68.
by nowittyusername 10mo ago
This brings back soo much nostalgia.. cool website.
69.
by nowittyusername 10mo ago
I've had this issues as well since codex models were introduced. i tried them but 5.1 regular on high thinking always worked better for me. I think its because its thinking is deeper and more nuanced it seemed to understand better wha
70.
by nowittyusername 10mo ago
Knowing Googles MO, its most likely not the model but their harness system that's the issue. God they are so bad at their UI and agentic coding harnesses...
71.
by nowittyusername 10mo ago
There doesn't need to be a correlation between some data structure and its effects for people to implement some sort of feature. There only need to be enough stupid people in powerful positions that believe in some sort of correlation
72.
by nowittyusername 10mo ago
let me just dig out a surprise Pikachu face from my pocket somewhere here ....
73.
by nowittyusername 10mo ago
This seems like one of those devices that seems like "meh" at a glance but grows on you once you used it. In fact just the Bluetooth button feature alone is warranted a second take let alone a mic embedded in to the ring with a cr
74.
by nowittyusername 10mo ago
You can change ANY sampling parameter once batch processing is off and you will keep the deterministic behavior. temperature, repetition penalty, etc.... I got to say I'm a bit disappointed in seeing this in hacker news, as I expect th
75.
by nowittyusername 10mo ago
I've been noticing that as well....
76.
by nowittyusername 10mo ago
My man did you even check my video, did you even try the app. This is not "bug related" nowhere did i say it was a bug. Batch processing is a FEATURE that is intentionally turned on in the inference engine for large scale provider
77.
by nowittyusername 10mo ago
determinism v nondeterminism is and has never been an issue. also all llms are 100% deterministic, what is non deterministic are the sampling parameters used by the inference engine. which by the way can be easily made 100% deterministic b
78.
by nowittyusername 10mo ago
Its not a magic technology, they can only represent data they were trained on. Naturally most represented data in their training data is NOT conversational. Consider that such data is very limited and who knows how it was labeled if at al
79.
by nowittyusername 10mo ago
The "miserable" existence ascribed by modern day humans to past human life is colored by their modern day psychological profile. If they were born and raised at that environment their psychological profile would be very different
80.
by nowittyusername 10mo ago
What makes us unhappy are the things that the modern world takes away from us. Sense of agency, sense of community, belonging, autonomy, recognition, and many other factors. The modern day human brain and mind is still lagging far behind ou
81.
by nowittyusername 10mo ago
Cute little game, I started in hopes of just trying it out and an hour something later finished the whole game.
82.
by nowittyusername 10mo ago
When humans domesticated animals and started tending to the fields is when IMO it all went down hill. That change brought in modern civilization with all its advantages but moreeso its disadvantages and maladaptive behaviors of the human mi
83.
by nowittyusername 10mo ago
You are correct that you cant change the content if its already biased. But you can catch it with your local llm and have that local llm take action from there. for one you wouldnt be instructing your local model to ask comparison questions
84.
by nowittyusername 10mo ago
You hit the nail on the head. Anyone who's been working intimately with LLM's comes to the same conclusion. the llm itself is only one small important part that is to be used in a more complicated and capable system. And that syst
85.
by nowittyusername 10mo ago
how much it cost me? well i been thinking about it for a long time now, probably 9 months. bought myself claude code and started working on some prototypes and other projects like building my own speech to text and other goodies like automa
86.
by nowittyusername 10mo ago
The local llm is the filter so yes you need one. and its not simpler to have the local llm do everything because the local llm has a lot of limitations like speed, intelligence and other issues. the smart thing to do is delegate all of the
87.
by nowittyusername 10mo ago
I realized that ads within context were going to be an issue a while ago so to combat this i started building my own solution for this which spiraled in to a local based agentic system with a different bigger goal then the simple original..
88.
by nowittyusername 11mo ago
I agree with the sentiment. things are moving so fast that waiting now is a legitimate strategy. though it is also easy to fall in to the trap of. well if we continue along these lines might as well wait 4-5 years and we get agi. which sti
89.
by nowittyusername 11mo ago
If you are building an agent, start from scratch and build your own framework. This will save you more headache and wasted time down the line. one of the issues when you use someone else framework is that you miss out on learning and under
90.
by nowittyusername 11mo ago
This thing rocks. i can imagine soo many uses for it. I really like the 3d pose estimation especially

More ›