What if we trained a smaller KV cache for our documents offline? Using a test-time training recipe we call self-study, we find that this can reduce cache memory on avg 39x (enabling 26x higher tok/s and lower TTFT) while maintaining quality. These smaller KV caches, which we call cartridges, can be trained once and reused for different user requests!
From what i understand, the FTC reviewed the Instagram acquisition twice before giving their blessing. I wish we all had this kind of power to remake decisions depending on how they turn out. Would be pretty awesome.
@ Aaron - Thanks, as always, for the thoughtful note but I disagree with you on this. Yes, there is risk of over-dilution if you raise boatloads too much out of the gate (including the very real and potentially fatal risk of being undisciplined on spending). But at the end of the day, if a founder figures out the business in that first 12 month window and is still sitting on another 12-24 months of cash, she has so much more power than if she is 4 months to cash out. She can take capital now but exactly on her terms. She can decide to keep pushing and raise in 12-18 months when the business will be in even better shape (and valuations higher, etc). Time becomes her friend, not her enemy. In my opinion, one would be crazy not to take more capital when it's this insanely cheap (valuations for 2 people with a powerpoint have never been higher). As we've seen with so many great companies, p/m often takes more than 8 months. Take more than you need and buy time. It will come out in the wash. I know this may sound self serving given my VC Cloak but having raised my first venture round as a founder in 1998, it comes from a place of experience (and pain). Thanks for listening.
I think we're mostly on the same page here. I'm less concerned with companies that raise 18-24 months of runway than the ones who are coming out of seed rounds with 36 months or more. I think balance is critical, and I'm hoping that founders find more of that balance vs. the recent pattern I've observed of founders taking every available dollar.
Almost always however there is something special about the founders e.g. they previously exited a successful startup or are simply a rockstar who all the VCs follow on Twitter.
It's not recent but Canva got funding with just a Powerpoint. And I can't imagine Justin Kan needing an MVP for Atrium in order to get into YC.
Yes: why? Or better: what. As in: what are you trying to tell us? That private messages are publicly available information, anyway? Or friends list, when set to private?
(Although two of those mentions are in the context of “consent decree”, which arguably doesn’t make it look any better)
A taste:
“This is just giving third parties permission to harvest data without you being informed of it or giving consent to it,” said David Vladeck, who formerly ran the F.T.C.’s consumer protection bureau. “I don’t understand how this unconsented-to data harvesting can at all be justified under the consent decree.”
No where in this article do they say users gave consent to have their information shared which they were required to do before such services could be implemented. At best, that's bad reporting and misleading at worst it's an agenda.
I really don’t think you are doing your employer any favors by your transparently obtuse attempts to misrepresent the actual content of the accusations.
i don't work for FB anymore. The content of the accusation leave out two critical pieces. 1. Users gave consent to share private information. 2. Only publicly available information was shared these partners. Those are pretty important omissions in my opinion.
The article clearly says users did not consent. And even if they did, that would presumably clash with your other assertion, namely that “only publicly available information was shared”.
So I stand by my accusation that you are pretending to be stupid, because no person working at Facebook would really be unable to find the accusation of private data being shared in an article titled “Facebook gave Spotify and Netflix access to users’ private messages”
You seem to be asserting that you have knowledge of the situation, enough to dispute the article. And you're an ex-employee of FB. That's most interesting. What else can you tell us? Be detailed.
I am sorry if i offend.. that is not the intention, this is an important question - is the naivety of the staff part of the problem?
maybe i am wrong but user blaming seems so very basic to me, and completely ignores how you are responsible for the things you create and the damage they do - and the false confidence or foolish acts they lead people into.
I would like to believe that on average we are ready to just not accept user blaming for privacy and security issues. if they are not qualified to wield the tools then we should not have given them out. as in: "blaming the user is an immature response" seems like a statement it should be okay to say.
I mean, the articles don't suggest it was shared publicly :-) They suggest it was shared with specific companies - but that it was shared nevertheless. That sounds like a serious violation of users' privacy.
If by "publicly available information" you mean the EULA, this can in practice not really be expected to be read and understood by the average user. A wall of text with vague remarks is par for such things. Add to this a constantly moving set of settings that change meaning over time (who can read this post now again?).
you don't need to give consent for publicly available information. if we are friends FB and you like a song on spotify and share that on fb publicly, why shouldn't I be able to see that when i'm on FB or Spotify? It makes my experience better on both. That's what instant personalization was. Again, the times failed to talk about that.
