Take a look at Japan or South Korea. Look at all the shut-ins dating their waifus. This phenomenon is not new.

Jokes aside, that is exactly why the CCP is banning intimacy with chatbots. Joke's on them, people have been "dating" video game characters/anime characters in lieu of actual relationships for awhile now. Eventually they'll have to go after dating sims and other media aimed at shut-ins (waifus/husbandos) if they want to make a difference.

The CCP can ban local models just as easily. Enforcement, well, that's something else now isn't it?

I was told that BSD is dead! Netcraft confirmed it!

People need to temper their expectations regardless. The developers are doing their best, but OEMs are making it hard for them.

The article isn't about us, it's about everyone else.

Flint 3 is not supported by OpenWRT. And I have, its a wifi6 device that has benchmarked about as well as other wifi 6 devices. I can get a cheap wifi 7 unit that will push higher speeds for less money.

Sorry Flint 2 not Slate 2. There is no Slate 2.

OpenWRT only supports a handful of wifi7 routers. I can easily and readily use only two in my market (ASUS BT8 and the Beryl 7 travel router). Most consumers don't even know what is OpenWRT much less how to use it.

There are nifty little $65 routers available that would probably be "okay" if remote administration were disabled immediately on installation. OpenWRT support for these devices would make them even more-desirable. Currently the standard router for OpenWRT seems to be the Slate 2, which is an older WiFi 6 router that is not going to meet everyone's needs.

Not that the OpenWRT developers/contributors are at fault for this. Apparently supporting newer routers has been quite difficult, based on the hardware platforms being chosen by OEMs. But OpenWRT can't be promoted as the solution for all problems related to router vulnerabilities.

Made = mad, stupid autocorrect.

Aren't vulnerable routers the backbone of most botnets?

Sure seems like people around here were hopping made about foreign disinformation campaigns a few years ago. How interesting.

Yeah, it's probably wrong. Especially if its coming from wumao and/or Russian troll farms.

Do you, now?

Deepseek originally identified itself as ChatGPT. It's pretty well known that the early versions of Deepseek were trained off one or more of the GPT4 models.

Until these Chinese state-backed firms start training their own models, they'll always be playing second fiddle.