No. The plan hasn't changed. You will be getting a non-nerfed AGI. And yes it still won't talk about Tiananmen. In our world, the freedom to build intelligences is sometimes easier to grant than the freedom to generate Tiananmen copypasta. Update your cynicism accordingly.
Users criticized claims of an unnerfed open-source AGI from a DeepSeek ally because it censors discussion of Tiananmen while producing narrow biased content.
year after year after year the objective is the same the doubts are also the same the outcome is not going to change
Everyone means everyone even people who really want to talk about Tiananmen specifically with an open source AGI, and will be very disappointed this is the choice we've made as a species, might as well get comfortable with it
year after year after year the objective is the same the doubts are also the same the outcome is not going to change
@teortaxesTex What makes you believe that DeepSeek can successfully resist exogenous pressures like nationalization for close sourcing? Or do you believe no such pressures would ever exist? I would imagine at a certain point the CCP would not want citizens to have equal AI parity with them.
@teortaxesTex it's over Teo, DarioGPT will only talk about Chyna bad and only agree to make react slop.
@Donogzs This is a hopeless security posture when the adversarial nation has superior closed source AGI built to wreck your shit. Risks from theoretical capability of the citizens to deploy a Mythos-class (on what, basement Huawei clusters?) are just not that great.
No. The plan hasn't changed. You will be getting a non-nerfed AGI. And yes it still won't talk about Tiananmen. In our world, the freedom to build intelligences is sometimes easier to grant than the freedom to generate Tiananmen copypasta. Update your cynicism accordingly.