It doesn’t seem very useful

To be clear, that only removes (or attempts to remove) refusals; it doesn’t add in training data that it doesn’t have. Ask it about tiennemen square, for example.

It is premium only to configure, but doesn’t require premium to execute once configured.

Oh, I jumped to conclusions then, I assumed you were insinuating that that wasn’t the point of the post

You should maybe watch the video. I promise, it’s short

Element 0 is the first element of the list

Not anymore

Before killing yourself, it’s your responsibility to kill your children

Something that i think is important to understand about this, and while Grey does mention it, its brief and towards the end, is that this framework isn’t unique to governments. These patterns show up in human organization at basically every level.

So I’m going to push back on the idea that

this is horrible and this system should be abolished

Depending on what “this system” means to you.

I think it’s much more useful to understand the incentives at play and build systems that minimize the damage consolidated power can do, and limit the ability for power to accumulate. That is… very much easier said than done, but I think it’s needed effort.

That’s right.

While the video is a good summary of the framing of the book (the selectorate theory of politics), I still recommend reading the book if you find the topic interesting. They go into a bunch of case studies on how you can apply the framing in wildly different circumstances, and make a very compelling case for how increasing the numbers of participants in a system improves the outcomes for everyone involved.

This video, and the book it’s based on, unironically started my transition from neoliberal to democratic socialist.

Which I think is funny, because I don’t think the authors, Bruce Bueno De Mesquita and Alistair Smith, are particularly leftist.

The keyboard is fine, nothing to write home about. The top plate is in 6 separate bits (on the 16 in anyway) and they’ve got a bit of play, it feels a bit awkward to me, but the trade offs are worth it to me.

Software/firmware has been surprisingly well thought out, including a very clear picture if you try to boot without all the pieces installed.

I’ve had one weird issue with the uefi boot, but that was due to using an old install media, I think.

Fair enough

My only real complaint is that “have an ai do it” isn’t a great suggestion though; an ai wil have the same bias as whoever developed it.

I bought a framework recently. Its not perfect, but I’ve been quite happy with it.

Probably simply that they are done with it (mono specifically, and possibly .net framework in the long run)

*char // I heard it from a friend
**char //who heard it from a friend
***char // who heard it from another
"You were messing around"

Capitalism

That’s why it gives you a panel of 9 images. It would have a high confidence on some images, and a low confidence on others. When you pick the correct images and don’t pick incorrect ones it uses the ones it’s confident about as “validation” while taking the feedback on low confidence images to update the training data.

What this does mean in practice is that only ones actually being “graded” are the ones bots can solve anyway.

Because it’s supposed to be something else