…The same Zig that ditched LLVM, to make their own compiler from scratch?

This is good. But also, this is sort of in character for Zig.

They’re pretty bad outside of English-Chinese actually.

Voice-to-voice is all relatively new, and it sucks if it’s not all integrated (eg feeding a voice model plain text so it loses the original tone, emotion, cadence and such).

And… honestly, the only models I can think of that’d be good at this are Chinese. Or Japanese finetunes of Chinese models. Amazon certainly has some stupid policy where they aren’t allowed to use them (even with zero security risk since they’re open weights).

Hostly, even a dirt cheap language model (with sound input) would tell you it’s garbage. It could itemize problematic parts of the sub.

But they didn’t use that because this isn’t machine learning. Its Tech Bro AI.

All true, yep.

Still, the clocking advantage is there. Stuff like the N100 also optimizes for lower costs, which means higher clocks on smaller silicon. That’s even more dramatic for repurposed laptop hardware, which is much more heavily optimized for its idle state.

I am well into Avatar obsession. It’s lived rent free in my brain too long.

Watch it till you get bored. And do not, in any circumstances, watch the last season.

Isn’t that in the same universe as Eureka?

I am down for Eurekaverse.

This is interesting, because “add ads” usually means margins are slim, and the product is in a race to the bottom.

If ChatGPT was the transcendent, priceless, premium service they are hyping it as… why would it need ads?

Same with auto overclocking mobos.

My ASRock sets VSoC to a silly high coltage with EXPO. Set that back down (and fiddle with some other settings/disable the IGP if you can), and it does help a ton.

…But I think AMD’s MCM chips just do idle hotter. My older 4800HS uses dramatically less, even with the IGP on.

Yeah.

In general, ‘big’ CPUs have an advantage because they can run at much, much lower clockspeeds than atoms, yet still be way faster. There are a few exceptions, like Ryzen 3000+ (excluding APUs), which idle notoriously hot thanks to the multi-die setup.

Eh, older RAM doesn’t use much. If it runs close to stock voltage, maybe just set it at stock voltage and bump the speed down a notch, then you get a nice task energy gain from the performance boost.

Depends.

Toss the GPU/wifi, disable audio, throttle the processor a ton, and set the OS to power saving, and old PCs can be shockingly efficient.

Yeah. Corporate sponsorship for the purposes of “get us the hell out of this monopoly” is kinda how open source has worked so far.

Sometimes its fine. See: Linux

The base Mac Mini is not super powerful. Physically, the silicon comparable to AMD Strix Point, which you’d find in any AMD laptop.

I am not trying to rag on Apple here: their stuff is fine. It’s ridiculously power efficient. It’d be beyond excellent for a handheld like the Steam Deck, or a VR headset.

…But a plug in gaming console? That’s more ‘M4 Pro’ silicon. And what they charge for that speaks for itself.

My interpretation is noise sensitivity. Especially when thinking.

M chips are super expensive. They’re optimized for low clockspeed/idle efficiency and pay through the nose for cutting edge processes, whereas most gaming hardware is optimized for pure speed/$, with the smallest die area and cheapest memory possible, at the expense of power efficiency.

And honestly the CPU/GPU divide over traces is more economical. “Unified memory” programming isn’t strictly needed for games at the moment.

And, practically, Apple demands very high margins. I just can’t see them pricing a console aggressively.

This is what basically anyone in the ML research/dev industry will tell you, heh.

Sigh.

I swear, Epic would probably have a decent reputation (and storefront) if Sweeney would just shut his trap and delete all his social media.

Interested. Also a little interested in the MC server, per your website.

Never played Starbound; how many players are typically on a server concurrently? Or at least y’alls server?

Sometimes. As a tool, not an outsourced human, oracle, or some transcendent companion con artists like Altman are trying to sell.

See how grounded this interview is, from a company with a model trained on peanuts compared to ChatGPT, and that takes even less to run:

…In 2025, with the launch of Manus and Claude Code, we realized that coding and agentic functions are more useful. They contribute more economically and significantly improve people’s efficiency. We are no longer putting simple chat at the top of our priorities. Instead, we are exploring more on the coding side and the agent side. We observe the trend and do many experiments on it.

https://www.chinatalk.media/p/the-zai-playbook

They talk about how the next release will be very small/lightweight, and more task focused. How important gaining efficiency through architecture (not scaling up) is now. They even touch on how their own models are starting to be useful utilities in their workflows, and specifically not miraculous worker replacements.