You can already run a large LLM (like sonnet 3.5) locally on CPU with 128GB of ram which is <300 USD, but can be offset by swap space. Obviously, response speed is going to be slower, but I can't imagine people will pay much more than 20 USD for waiting 30-60 seconds longer for a response.
And obviously consumer hardware is already being more optimized for running models locally.
But most of the time, we as engineers don't pick the winners. Some C-Suite executive or middle manager, who isn't very technically inclined, picks the winners, and we as engineers are forced to make it work.
As I don't think a engineer has ever had the chance to choose a company's CRM, the CRM with better marketing would always win over the CRM with better engineering.
Content is being banned because the governments don't think the citizens collectively can handle that content. This is the government thinking for the people. I wouldn't have an issue if it was really argued and debated, but I don't really see that happening. It is mostly just the government unilaterally deciding that said content is bad and banning it with no really possibility for debate. It isn't too hard to imagine a situation where content would help citizens but harm the government, and who is really deciding what to ban in such a case.
Countries are seeing how much banning content helps governments in places like China, and want that easy "solution" to their problems instead of doing the difficult thing of properly educating citizens and cultivating free and open discourse.
I guess my analogy is bad because it is debatable if fast food, in itself, is bad for you. My issue with your analogy is that the use of violence is abused by governments so much more than it is every used correctly (if that is even possible) and that all citizens would benefit from governments condemning violence. The reason governments don't condemn violence isn't because their use of violence helps citizens, it is because their use of violence helps themselves. And the same thing goes for banning content. So I guess your analogy is actually pretty good,
just counterproductive to your point.
There are two things wrong with this argument. One, it implies that there isn't a better way to deal with an such a malicious organization. And two, it doesn't acknowledge how such a ban creates an obvious opening for abuse.
Holding up the classic Western ideals of Democracy and Freedom is hard because it is much deeper than simply giving people the freedom to access all information so they can form their own opinions. It also requires that these people are educated and trained to be competent critical thinkers and be able to intelligently form their own opinions. It holds its citizens and the government to high standards and will collapse if these standards aren't met. Accordingly, better education and trust in citizens is the better solution, not banning.
As for the obvious opening for abuse, it doesn't have to be said that every system will eventually be maximally exploited, and creating this opening for exploitation will eventually be exploited as well. It is just a matter of time...
Another poor but useful analogy is fast food. Banning bad media is like consuming fast food. It is quick and easy, and "satisfies" the goal within some basic parameters, but it really does more harm than good in the long term.
And obviously consumer hardware is already being more optimized for running models locally.