syntaxing

8 ms·

31.
by syntaxing 1mo ago
Prefilling Kimi K3 with opus is a super interesting idea. That being said, I absolutely hate this website layout
32.
by syntaxing 1mo ago
Because distribution prices are spread throughout the users. This is the case for policies that allow the rich to be completely off grid which is not the case here in the US (most cap off around 1500W).
33.
by syntaxing 1mo ago
Simply not true. There’s a ton of induction and electric coil ranges with knobs. Just look at the Bosch line.
34.
by syntaxing 1mo ago
No way. I cook everyday and induction is vastly superior than gas. The only time fire is vastly superior is when I use a wok which I have a stove I use outdoors. Most induction stoves have temperature maintained too. So it can keep your oil
35.
by syntaxing 1mo ago
I’ve been using induction since I have had kids and I enjoy it so much more than fire. That being said, I rent so I can only use 120V ones and the options are so limited here. Most of the small 120V ones make an obnoxious hum.
36.
by syntaxing 1mo ago
I work in embedded space. Just because it’s in hardware doesn’t mean you can’t “protect” it. Most modern software (regardless if it’s hardware or not) can be cryptophically signed.
37.
by syntaxing 1mo ago
I don’t get why this is an issue? You can run Claude/OpenAI SOTA models through Amazon bedrock. These weights have to live somewhere to run on Bedrock.
38.
by syntaxing 1mo ago
SOTA American models are not. SOTA Chinese models are. From a physics aspect, closed source models cannot be too far from open source ones in terms of size. There’s only so much you can squeeze out a B100 style cluster even with fancy Dflas
39.
by syntaxing 1mo ago
Honestly, this is starting to make more and more sense. SOTA models are starting to converge to certain architecture and capabilities. I wouldn’t be surprised we end up with a base model ASIC + “fine tune” card where it’s a physical LoRA st
40.
by syntaxing 1mo ago
For personal stuff, I use it with AnythingLLM. It replaced any Google search for me. For coding, I run opencode though I have been debating switching to Pi. I would argue it’s at Sonnet 3 level.
41.
by syntaxing 1mo ago
I meant Qwen3.6. Unsloth supposedly has early preview of the model and the VRAM requirement is the same so most people expect similar model size and type.
42.
by syntaxing 1mo ago
I am so excited for Qwen 3.8 27B. It’s a shame how slow prefill (~3-400) is on a strix halo but it’s such a good model for agentic tasks.
43.
by syntaxing 1mo ago
This reminds me of Three body problem and how the scientist discovered the high strength wire was through quick physical experiments and use them as input to an AI model to determine if it works.
44.
by syntaxing 2mo ago
> View pricing in the Cloudflare dashboard ↗ Why… I wanted to see if it’s worth it to use cloudflare’s endpoint but I can’t even see the pricing
45.
by syntaxing 2mo ago
Interesting take on this https://aftermath.site/war-crime-n64-palmer-luckey-m64-modre...
46.
by syntaxing 2mo ago
I have a framework desktop, but depending on your need, DGX spark might be better. The prefill and NVFP4 is a significant advantage. But framework desktop is a better general computer. I expect to be able to use it for years to come. Where
47.
by syntaxing 2mo ago
I use it with a strix halo server. 35B runs stupidly fast. 27B is about 700 TPS prefill and 30 TPS token generation. Which interestedly is about what Kimi K3 gives me depending on provider.
48.
by syntaxing 2mo ago
I’m a mechanical design engineer (10 years) turned software engineer (7 years in autonomous vehicles). Is there a particular role you would recommend for someone with my experience?
49.
by syntaxing 2mo ago
As a renter myself, I get the author's pain. I ended up just installing this non destructive retrofit switch for most of my home automation + smart bulbs. [1] https://www.thirdreality.com/products/smart-switch-gen3
50.
by syntaxing 2mo ago
Same, I thought it was about some cool network tunneling that made the GPU look like native hardware on the Linux host.
51.
by syntaxing 2mo ago
> Open-weights models that don’t have dangerous capabilities are a public good. Who decides what is dangerous and what isn’t? Lawmakers usually have the say but Anthropic can easily bribe… I mean lobby them to favor your viewpoint.
52.
by syntaxing 2mo ago
Which quants? I get these speed (only 20-30 TPS) at Q4_K_M with MTP for 27B on my framework desktop. I draw sub 130W for the whole machine.
53.
by syntaxing 2mo ago
Fluid dynamics is my jam (most of my masters was studying it)! Even in the intro to fluid mechanics book, they spent a good half+ chapter why a flying wing is superior and how you can have it designed similar to movie theater seating. But m
54.
by syntaxing 2mo ago
I had a Meta recruiter talking to me a while back (maybe 5 years) and she ghosted me. Found her obituary online a couple weeks later. Recently, had another Meta recruiter ghost me but that’s because of the Meta layoff round robin. Probably
55.
by syntaxing 2mo ago
Bluetti has a small battery pack for now. I think the options for consumers will balloon once CATL mass produce sodium ion cells. Solix/anker, jackery, ecoflow, etc all use CATL cells already.
56.
by syntaxing 2mo ago
Probably operating temperature and safety.
57.
by syntaxing 2mo ago
I got a used speedmaster before the huge price increase in both new and used market that was triggered from “tariffs”. I love my watch and don’t regret getting it but I dread the costly servicing fee.
58.
by syntaxing 2mo ago
Interesting, you can’t use their SDK or you can’t load whatever you want on it?
59.
by syntaxing 2mo ago
Hoymiles just got UL3700 approved this week and sold by Bright Savers
60.
by syntaxing 2mo ago
Point 1 isn’t true anymore as of this week. The first UL3700 inverter from Hoymiles just got approved and is being sold by Bright Saver for consumers https://www.brightsaver.org/balcony-solar-kits

More ›