@sysedit nice :blobaww:
Failed project, making it open source one day if I find time so everyone can see the failure
Abro directo!!!
-Estreno cΓ‘mara!
-Jugaremos a los juegos de la Todogodotober 2025
-Multitasking!!! (a.k.a. Tareas)
π https://twitch.tv/rafalagoon
Ven si quieres ver y probar juegos, si quieres que te echemos una mano con tus proyectos o simplemente a charlar.
@gavi yep, squid = calamar, also there is a musician called calamaro which is totally unrelated but kinda funny
@gavi ohh very distinguished squidward, I forgot it was called like that not calamardo like in Spanish :blobpats:
@gavi is there a reason or story behind the calamardo with wig picture? I like it
Happy Monday from this anxious pizza eater πΆ π
https://www.thisiscolossal.com/2025/10/alison-friend-dog-only-knows-paintings/
Interesting although the post confuses open source with open weight which is quite different, right now the only real open source SOTA AI that I know is ApertusAI, no deepseek no llama, those are open weights, trained on closed doors with dubious data and the only a weak version is released for free to give them good marketing and to deny more market share to they're competitors and attarct possible customers.
Hoy me he levantado con esto en la cabeza
https://youtu.be/JmvCQdELarw?si=Jo-l-19ttrPNfIhJ
@javascript omg que es esta genialidad π π€―
Augmented virtuality
Managing extreme AI risks amid rapid progress
https://arxiv.org/pdf/2310.17688
#agi #stopagi #ai #stopagi #pauseai #aisafety #artificialintelligence #GeoffreyHinton
@gnaddrig well, there are many AI researchers that are very concerned about the impact AI could have in the near future, if you can do a developer job or a researcher job with AI, then does it matter if it thinks or not? You could have an army of extremely intelligent agents that doesn't sleep and that do work orders of magnitude faster than humans, to the servicie of a bunch of people.
@gnaddrig what we have as consumers is already insanely powerful, can you imagine what is not public? You don't need ASI or even AGI to have a very dangerous weapon. AGI would mean having a tool smarter than humanity, do you think that tool will not have it's own goals? Do you think they are spending half a trillion dollars on data centers to make chatbots for customer support? I'd love to know what is your opinion on this, and if you have tried state of the art tools like claude code