@LefterisJP friends in Milan are setting alarm for 3am to open windows since it's too hot before that time, similar in Paris) I think Berlin is relatively o.k..
@slobkebap Claude basically tells not to) BPC-157 and TB-500 have essentially zero controlled human trials. The dose math on the box ("10 clicks × 1.7mg GHK-Cu, 0.33mg BPC-157, 0.33mg TB-500") is internally inconsistent(?) with the stated 70mg/10mg/10mg/50mg totals.
Chinese government used to send pandas as a sign of goodwill. Frontier models should be America's pandas - sent to friendly countries. Freedom dust, selectively distributed.
@emilykmay nah, women are "pornbrained" by evolution, emotional connection increases the chances of bringing the offspring up together. but I agree that dropping the standards across the board might help.
@garrytan What is the real advantage of making it agentic that justifies massive token cost? i.e. if I'm reluctant to share load of personal data yet interested in surfacing/connecting ideas from the book I just need to feed it to a model and prompt accordingly
@cryptopunk7213 she has to route the questions from the kid she prefers to a better model and the less favorite to a quantized once to save on credits. but honestly this is bullshit.
@teortaxesTex Show me on this matreshka where a European bureaucrat hurt you. Because the euromisery is not in numbers, certainly not when compared to us/china/russia populations. Co-dependance of eur and us and even china and eur is stronger then rage-baity posters are willing to admit.
@thebadstats Generally super low level from both sides, and an amazing level of disrespect - Konstantin saying things like “are you a moron or pretending to be one” or “stop talking”. Wtf?
@thebadstats I was surprised of how weak a debater/speaker Destiny (admittedly I have never before seen him talking) - he repeatedly failed to expose bad faith or misinterpretation of Kisin, such talks do us (Destiny side) really no good))
@thedankoe does one bring existing subscriptions to model providers or api access? the problem is that if I already subscribe to 2-3 providers I kinda don't want to poll same models for more $$