I was curious: Why have it, instead of just using memcpy_and_pad?
AI's answer (paraphrased) was
* Avoid possible bugs from manually write sizeof(dest)
* Enforces the __nonstring Attribute
* signals: "I am converting an actual C-string into a fixed-width legacy memory field." vs copy binary data & pad it.
Interesting to learn about the __nonstring attribute:
If the alternative is "burn more tokens on finding issues than the attackers do", formal verification starts to look comparatively feasible cost. Think of it as setting an upper bound on cost, vs just burning more and more tokens.
AI assistants would reduce effort of verification too.
One difference with Mullvad is VPN traffic is ephemeral. Here, a VPS has a persistent disk attached, that could contain identifying information (if it is necessary to do useful work).
Personally I'm more optimistic about Servo - because originating at Mozilla, I imagine more web browser experience and expertise went into its architecture, and also because Rust.
I think the "burning the boats" action is releasing very low API pricing. Once consumers see the lower price they would balk if it was raised back up enough to provide some margin on inference. So the article is saying OpenAI won't have a way to switch back to profit from inferences.
It was initially opt in for me, then they made it mandatory.
(Sure, I could pay by check but consumer banking technology/US in the US already feels like is is lagging a decade behind other countries without voluntarily going further back. Paying by check every month would be quite inconvenient.)
I'd already decided to avoid bilt as much as possible, but reading this thread prompted me to try going a little further.
> Request to Know... The specific pieces of Personal Information we collected about you.
> You have the right to opt-out from having your Personal Information and Sensitive Personal Information sold to third parties. You also have the right to opt-out from having your Personal Information and Sensitive Personal Information shared with third parties for purposes of cross-contextual advertising
This looks like a technical decision on where to focus investment.
Using words like "bias" and "descrimination" imply there is some wrongdoing.
From a quick read (maybe I'm missing something) this seems to me to be coming from "You didn't fund the project I care about" rather than any wrongdoing.
I've also done a lot of rock climbing. As other replies said, you get comfortable with the gear and the height and loose the fear.
When Covid broke my regular climbing habit. One interesting part of going back to it after a year or so of not climbing - I felt some fear of heights again.
It was mild fear, and didn't have me to hanging on extra tight or something. I remember the odd sensation of observing this fear in myself and thinking something along the lines "Oh that's weird, fear of heights - I don't remember the last time I had this."
Like a muscle, if you start training to be comfortable again it seems to come back quicker if you had trained it well previously. I guess fears have homeostasis too.
> Unlike tokenization, BLT has no fixed
vocabulary for patches.
iiuc this means: the vocabulary of patches is not known prior to training.
I guess once training has established a vocabulary of patches, that same fixed vocabulary is used for inference (if this is not true I don't see how it could work).
Imagine if they added something like this to UE5 licensing:
If your game has not been updated in N years...
1) Internet Archive can distribute it for free
2) Let people distribute modified versions that does not need license key or whatever copy protection.
Harder but extra cool: To get a UE royalty discount, put source code in escrow set to release it if game not updated in N years.
> He slammed the door in my face, loud and sharp, like an acoustic lemon.
I think its not really about the sound, it is about the response you might make after a big hit of lemon flavor: puckering up, wincing, pulling your neck back - I can definitely imagine responding to a distasteful and loud sound in the same way.
Also, I felt this implies the proper-historically-correct-techniques are much better than the wrong-modern-impression-techniques, and would win in combat.
Makes sense, but it would be cool evidence to see simulated combat to test this.
I was curious: Why have it, instead of just using memcpy_and_pad?
AI's answer (paraphrased) was * Avoid possible bugs from manually write sizeof(dest) * Enforces the __nonstring Attribute * signals: "I am converting an actual C-string into a fixed-width legacy memory field." vs copy binary data & pad it.
Interesting to learn about the __nonstring attribute:
https://github.com/torvalds/linux/blob/1a3746ccbb0a97bed3c06... https://github.com/search?q=repo%3Atorvalds%2Flinux+__nonstr...