🚨𝗕𝗥𝗘𝗔𝗞𝗜𝗡𝗚: Major Premier League clubs, including Arsenal and Liverpool, are reportedly unhappy with the punishment handed to Manchester City and are pushing for further sanctions.
Despite the 10-point Premier League deduction, two-year ban on registering first-team players and £300 million compensation order, the clubs are reportedly demanding even harsher consequences.
Relegation and the stripping of titles are among the additional punishments they want imposed on Manchester City.
I’m still stuck on this article. The writer says when she was 22 years old she met a 37-year-old famous man, and then went over to his house. When she got there, she sat on the edge of his bed. He made no move on her, didn’t force a sexual exchange, nothing. Then she left his house the next day and that was it.
20 years later she recounts the story from the POV that this man was likely a creep, and that the reason he didn’t sexually advance on her is because (at age 22) she was too old for him. With no evidence, she accuses him of being a pedophile.
Never mind that she went there willingly, and he treated her with respect. She still found every possible way to turn herself into a victim. She took a theoretical and made it into fear porn for her readers. This story is worthless, it did not need to be told, and it arguably harms real victims.
@mitsuhiko@igormaka@kvalekseev That brings me to a wishlist I have from anyone who is acting as a proxy: Tell me how this model from this provider performs with some benchmarks. I have seen the same model hosted perform very differently with different providers.
@mitsuhiko@igormaka@kvalekseev Most important control is control on routing for ZDR. I also want to have controls on where a model I would prefer to route. Lot of times I am not neccessarily looking at the cheapest, but a better reliable provider.
@mitsuhiko@igormaka@kvalekseev I would totally see myself using radius if these two requests are in place. But as it stands - without clarity on ZDR and model provenance - it is hard to ignore the cheaper options available.
Lumo 2.0 utilizes powerful underlying models, with zero-access encryption. No conversation logs. No training on your chats.
Keep your mathematical discoveries safe with private AI.
@opencode For OpenCode Go, recently you introduced "Enable Models Hosted in China" as an option. It would be great if you have an option to choose providers who have ZDR and also another option for providers who have use data for model training. Would be useful for both Zen/Go
@MeryemArik9 I still the same standard pricing for real time in your pricing page. Is that a mistake or the price you mentioned applies only to async calls ?