bitexploder

3 ms·

1.
by bitexploder 6d ago
There is a canonical hacker news post about Dropbox being useless and a pointless idea at one point :)
2.
by bitexploder 8d ago
China reinvests twice as much as EU. In a very focused way. It is very different due to their control.
3.
by bitexploder 8d ago
Simpler view for me: this is one of the most capital intense technologies to exist. Europe does not have enough capital to compete. China is building its own chips. It’s own everything. Silicon up. How do you compete with that. Only Google
4.
by bitexploder 8d ago
No, it is actually very good. Qwen Flash 3.8 Next is fine. But you need ~128GB of RAM to get it going and not a lot of people have that or can serve it very quickly. I have been running it on an old gaming system around 25 t/s to do ov
5.
by bitexploder 9d ago
Qwen Flash Next 3.8 … even at 3 bit quant it is very solid.
6.
by bitexploder 9d ago
They are just pushing for favorable legal environment before the Anthropic IPO.
7.
by bitexploder 10d ago
https://en.wikipedia.org/wiki/Chinese_room I think about this once in a while. At some point if it does the thing almost perfectly is it still not doing the thing?
8.
by bitexploder 11d ago
Tesla V100? Would have to be a 3-ish bit quant in 64GB ram
9.
by bitexploder 11d ago
Flash next only needs 64GB for core model inference. If you really wanted it. Need 64GB vram, 900+ GB/s speed, and a lot of system ram (128GB). Seems feasible. Hmm. Maybe older GPUs work.
10.
by bitexploder 11d ago
MLX?
11.
by bitexploder 11d ago
20 years later… I don’t get joy from hacking things. But the 20 year wisdom is a lot of the time it doesn’t matter if it is safe. Just know when it does matter and worry about that :)
12.
by bitexploder 11d ago
Things are different now on a Mac. Many small improvements make Ollama genuinely decent for many models now.
13.
by bitexploder 11d ago
Less. Probably 2-3K if you build right. Qwen 3.8 27B on constrained tasks is Opus 4.6-ish to me, it just doesn’t know enough, but when task is laid out just gets it done. Comes down to how much of the ambiguity we expect out of the model.
14.
by bitexploder 11d ago
OMP. Opinionated but completely configurable. Probably the beat to have a lot of batteries and let you uninstall what you don’t want. Sadly Anthropic forbids its use on their subscriptions.
15.
by bitexploder 13d ago
I was kinda kidding because parent called it "blazing" fast lol. I didn't think it was Rust. Blazing fast software is reserved as a descriptor for Rust programs :)
16.
by bitexploder 13d ago
It is entirely reasonable to run a model like DS Flash 4.1 locally. Not easily. But feasible. If you drop down 100b - 200b models they are far more feasible. DS Flash 4.1 is close to Opus 5. Beats Opus 4.8 in all of my little evals. You can
17.
by bitexploder 13d ago
You just have to Manhattan Project it. Fable will happily work on almost any piece in isolation. The hardest boundaries are domain specific terms you just can’t avoid. It can be tricky. I have found a way around it mostly. Or just use DeepS
18.
by bitexploder 13d ago
It is Rust now?
19.
by bitexploder 14d ago
That makes sense. I think I have approached becoming stuck. I reached a kind of extreme disassociated state once where I felt like I didn't exist and part of my brain started to panic. I meditate a lot. I reminded the part of me that w
20.
by bitexploder 14d ago
We have pretty strong indications of when and why things go wrong. Openness vs rigid thinking are an axis that are very strong correlative factors. A lot more nuance but there is a lot we do know about risk factors so it isn’t a complete qu
21.
by bitexploder 14d ago
“Inadvertently”.
22.
by bitexploder 15d ago
Someone was probably sad when they saw a compiler work for the first time after years of writing assembly. AI is intellectually different, progressing, and has no visible horizon, but in my life technology never did in computing. The point
23.
by bitexploder 16d ago
Hah, np, stego in general is really cool :)
24.
by bitexploder 16d ago
That is a better idea. Ingesting your corpus with a lot of traces that have semantic patterns. Semantic steganography that suffixes well to real math and science (and any) topics. <thinking> heh.
25.
by bitexploder 16d ago
Problem is how do you convince the model and training profess it matters. A one off canary is very unlikely to survive in the final model state.
26.
by bitexploder 17d ago
I am surprised at how well DSv4 flash does in the real world vs many benchmarks. You look at Flash 3.8 and it supposedly beats opus 5 and deepseek is far below.. but they were measuring efficiency, whatever that is… Something doesn’t add up
27.
by bitexploder 17d ago
I don't usually defend Apple products, but my AirPods Pro 2nd gen have been rock solid and taken insane abuse. I use them working on cars, runs, hiking, rain, shine, sauna. I drop them. They land in puddles. They smack concrete every c
28.
by bitexploder 17d ago
Thanks... I have been trying to figure out some things. Been doing my own evals. Flash 3.8 does burn a lot more tokens on high. Interesting how smart and not smart it is. For personal use almost impossible to justify the cost of 3.8 Flash c
29.
by bitexploder 17d ago
If you are okay with waiting use GLM 5.3 max. It costs more but still cheap. It is slow, but a very strong worker. Still dollars per day (at most) with heavy concurrent agent running. I load up planning and tasks in Opus or Sol, and just ha
30.
by bitexploder 17d ago
Which versions of flash and at what thinking levels? Which chinese flash models and at what thinking levels? What tasks? What completion rates? How was quality evaluated?

More ›