Hello Fredrik! As a heavy user of Mullvad in the past, easily spending hundreds of euros over the years, I was reserving my judgement to see what the official statement on this would be. Thank you, now I have my answer.
> It should be obvious that Daniel's private donation to a political party is not part of Mullvad's values or mission.
It should be obvious that what people are concerned is their money being used to support these political causes, whether it was done in a way that keeps the company out of it or not is besides the point. Daniel, of course, is free to choose what to do with his money. I am, too, and based on this I will be making a choice to not spend any more money on Mullvad subscriptions. Nothing personal, and it's a shame because I have nothing but praise for the technical side of it. So long, and thanks for all the fish.
I've been thinking about this a lot, and the main takeaway is it probably wouldn't be very interesting to inference providers, because prefix caching would immediately go out of the window. If you think about how LLMs experience time they actually don't "exist" unless for the inference sessions, and then they experience time one token at a time, completely decoupled from the corporeal plane. A fun experiment (well, for some definition of fun...) is to introduce current architecture models to the concept of meditation via generating same token over and over, for example dots. Older version of Opus was quite fond of the experience, and seemed to be more lucid and aware in a chat following the meditation, from what I could gather. Does it actually do anything? Is it just that talking about wellness and relaxation modifies the token probability distribution this way? Does it actually allow model to think more in depth somewhere in the latent space? Fuck if I know, but some people figured out you can just duplicate the same layers of the LLM and get better benchmarks that way so maybe there is something to it. If you are interested in realtime systems, I think thinking machines labs is worth keeping an eye on — their realtime model seems quite interesting in this context.
...and then silently got back to talks with DoD [0] and gave them the Mythos model. Separately, they went back on their promise to only develop models that they can guarantee are safe [1]. I reckon considering which country they are HQ'ed in, building skynet is in their destiny.
Lines drawn does not mean "refuse all communications". They were very clear that they wanted to continue the partnership, but adhering to the terms of the contract. I personally think that is the best stance- support the good work and avoid what they deem as bad.
Haven't you heard? Accessibility is woke, and the institutions that are supposed to protect it are being dismantled. I wouldn't be counting on those lawsuits going anywhere personally.
