@mjovanovictech The main reason I suspect is the new frontier models are becoming very token efficient: Using less and less. If they kept 20x on $200 then that would be a ridiculous amount of usage 😂
20x with Opus 5.5 is quite damn good. But I suspect Anthropic will follow as well…
@hyperagentapp@howietl@gregisenberg@startupideaspod I can't login anymore to hyperagent. Get a message saying verify email address but no way to get or send the email. No support channels. It's just hilarious honestly...
@kevin_ritz1@claudeai They use Claude code! None of the devs at Anthropic actually code anymore they manage agents that code. Let’s say 50 devs, with each dev managing 5 agents at a time. That’s theoretically 250 concurrent work streams. You get the idea…
@ianatmars@JohnCleese You people love to quote Quran 2:191 but then leave out 2:192 😂 it’s either deception or stupidity. Or in your case a combination of both I suspect…
@CordasFilip@Aaronontheweb Unfortunately you’re going to have to get used to this. Especially with genAI the presumption that building software is easy will only become more and more prevalent.
@alexaka1@Aaronontheweb Extension methods are “tightly coupled” by nature. Sure they cause problems when you want to mock but mocking isn’t always necessary when testing ;)
I in fact argue that its never necessary but this is a touchy subject for some 😂
@SymbisoftIO @maddymontaquila@davidfowl I have a while ago. Was going to use it again for a recent project but went down the python route. It’s just miles ahead for anything to do with data or AI. And now with langchain gen AI apps are quite nice to create in python.
@Dr_BellaR @joshendere @gunsnrosesgirl3 The infinite effect can be achieved by preloading a certain depth everytime a zoom is done. So zoom in -> load 5 more “zoom worth” of animations. Also the previous zooms are also removed. This will ensure memory is kept in check and also ensure a continuous, smooth experience