Followers is a somewhat meaningless metric. Instead look at something like likes or views. The most relevant posts will spread via the algorithm and get the most views.
>it's just generating a plausible add-on to the document
A plausible document that follows the alignment that was done during the training process along with all of the other training where a LLM understanding its actions allows it to perform better on other tasks that it trained on for post training.
I find it more likely that the author got more fluent from his job, friends, and other every day things you run into by living in Japan. Spending every day reading, writing, talking to, and listening to one's coworkers and then after work also talking more with those coworkers or friends would be much more time than a single magic event per week.
They dictate the capabilities that their device is offered and how the device is designed. It is up to the consumer to decide if that is worth the price of the device.
To be blunt it doesn't matter if you have a choice or not - this sort of behavior shouldn't be permitted either way. It's an appliance that at this point serves an essential function in society so user hostile behavior ought to be strictly prohibited.
The guiding principle should continue to be that manufacturers and retailers don't get to control the second hand market or dictate what users do with the things they purchase. Digital controls used to thwart the owner's freedom should be outlawed.
Apple is, because of vendor-lock in. Once you're sufficiently dependent on Apple's ecosystem it becomes painful to switch to a competitor because it requires switching to a different smartphone which then locks you out of most of Apple's ecosystem.
You are correct and I don't get OP's point. Don't want apple rules, don't use apple products. They are the business, they can do what they want with it, right?
We have consistently made exceptions to this rule in situations with limited choices. We would not abide by the electric company dictating a range of things, even of you have the option to run your own generator.
The truth is there are two reasonable platforms, as long as that is the case we should apply scrutiny.
I'd go even farther than that. The US should adopt an equivalent of the second amendment regarding end user control over personal electronics and it should bind not only the government but also private enterprise. We are increasingly dependent on these devices to go about our day to day lives and they have not only been used against us for mass surveillance but are also quickly gaining the ability to exhibit intelligence and act autonomously.
There are standards for interoperability and user-friendliness with all kinds of devices, and we should expect the same from modern devices.
It would have been pretty peculiar and unacceptable if your telephone in the 80s couldn't call your neighbour because the telephone company just decided to not make them interoperable, why shouldn't it be the same here?
I always assumed it was priced outrageously to have a big enough margin to start fulfilling the preorders and refund requests from the original kickstarter. The device does not sell very many units so it won't benefit from bulk pricing.
Pretty much all the routers that are targeted by the ban would be OpenWrt derivatives, AFAICT. It’s basically the Android of routers, except without the Google resources.
Google Wifi Is one of the main lines that aren’t based on OpenWrt.
Just in case anyone else was wondering, it seems that some early products (running AirOS) were modified versions of OpenWRT, but later software/hardware is not.
So yes, this comment is correct, but it threw me since I wasn’t following the company back then and I hadn’t heard of that history before.
As another commenter pointed out these models are being trained how to save and read context into files so denying them to use such an ability that they have just makes your claim tautological.
Ok but it's already known that you shouldn't initialize your network parameters to a single constant and instead initialize the parameters with random numbers.
Both you and the comment above are correct; initializing with iid elements ensures that correlations are not disastrous for training, but strong correlations are baked into the weights during training, so pretty much anything could potentially happen.
reply