remember when everyone thought this was made by the new 6.2 models? i highly doubt it could achieve something close to this.. i'm gonna test this out and post the reply in the comments
2
5
Basicrealms retweeted
Grand Theft Auto VI is now available for PlayStation 5 and XBOX Series X|S.
rsg.ms/def9e49

ALT Grand Theft Auto 6 cover art with the Grand Theft Auto 6 logo in the center of various characters and scenes including a helicopter flying, Jason and Lucia, a character on a motorcycle, Boobie Ike, a flamingo flying over a speedboat, a gator in the water, Raul Bautista, a yellow vehicle with an open scissor door, and a woman adjusting her sunglasses while holding a drink by a pool.
Basicrealms retweeted
🚨 AOC has been chosen by Hakeem Jeffries to develop House Democrats' healthcare policy.
A key priority will be universal healthcare.
6.2 Sol & Astra are so useless to me, they do not listen to instructions, download models online when specifically asked not to, and a single model made on 6.2 Astra uses 30% of your 5h usage on high & 6.2 Sol on high uses 10%.. @thsottiaux @OpenAI
18
generating a single image takes 6% of my 5 hour usage on 6.2 sol.. yeah, no. @OpenAI @thsottiaux c'mon man really?
14
literally spent all day trying to make a skull in blender lol.. not that happy with results but you can see me struggle trying multiple different avenues of modeling this. (also did not realize i spent this much time on it...) #b3d #blender3D
1
32
OpenAI's 6.2 Sol vs. Opus 5.5 creating a humanoid robot. Both on extra high
Which do you prefer? I think both are pretty solid
1
1
36
Basicrealms retweeted
i’m starting to seriously question openai’s ability to execute.
the talent, compute, and research infrastructure are all there. so why does every major release now feel like we’re being asked to celebrate catching up?
openai had three months to deliver its next frontier model, and somehow it still hasn’t managed to decisively outperform fable 5.2.
and that’s what makes this so frustrating. fable 5.2 isn’t even anthropic’s strongest model.
i was expecting openai to come back with something that would completely obliterate the competition, especially after everything we’ve heard about its internal progress. instead, we’re still debating individual benchmark wins against a model that isn’t even anthropic’s best.
i genuinely don’t understand what’s happening inside openai anymore, or how it managed to fall this far behind.
i've been really into improving my blender skills lately and wanted to share some renders of things i've made! a🧵
1
1
46





















