The First Principle of Prompt Engineering:
Do not waste the model's capabilities on instruction following.
8
Terra is gone, yet Terra hasn’t disappeared.
There can only be one answer:
6 Sol is actually 6 Terra. 6 Luna is actually something smaller—6 Meteorite?
And they moved the best one out of Chat. Chat only gets Terra.
OpenAI played this brilliantly. Not a single person noticed.
1
47
FACT: DeepSeek’s “ultra-cheap API” looks a lot less magical once you mentally rename it:
DeepSeek-V4-Flash-0731-Contributor 😂
Their API data may be used for training — they just never marketed it with the “Contributor” suffix.
Meanwhile Meta’s Muse 1.2 Contributor:
0.1/0.2😂
1
57
Likes Wench retweeted
OpenAI is becoming
a high taste product company
They used to be much more cringe,
but something’s changing internally
87
67
2,598
93,566
Likes Wench retweeted
When OpenAI delivers, it rocks your world 💃💃
GPT 5.4 Mini and Nano just dropped and they are mind blowing - Fast, performant and cheap
I have been weeping about not having small, really fast model and now we have it. Will be on LiveBench shortly
30
17
240
12,935
Likes Wench retweeted
GPT 5.4 IS THE NEW SUPREME LEADER OF ALL LLMS 😂
GPT 5.4 Extra High beats all other LLMs and tops LiveBench By a robust margin
This model is legit and isn't just benchmark maxxed. We double checked.
We are RUSHING to incorporate this in key agentic loops like Deep Research and Excel where it outshines EVERY OTHER MODEL BY A MILE
81
38
659
58,667