OpenAI CEO Sam Altman: The Model Finally Excels in Design
Sam Altman stated that seeing OpenAI's models finally excel in design still makes him feel like his "brain is a bit fried." Previously, he criticized some individuals for only opening hardcore questions to those deemed "worthy," otherwise quietly downgrading access rights.
Source: Public Information
ABAB AI Insight
As the CEO of OpenAI, Sam Altman has previously shared multiple advancements in model capabilities. His remarks on design capabilities reflect breakthroughs in models like o1 and GPT-5 in the creative field, similar to the iterative path of multimodal capabilities. The discussion on access fairness continues the controversy over OpenAI's user stratification, with Altman emphasizing transparency and value orientation, motivated by balancing resource allocation and community feedback. Current AI models are in an accelerated phase of transitioning from text cognition to multimodal design creativity. Essentially, this represents a technological substitution: breakthroughs in design fields by models replace some human creative labor, with the mechanism being that training data and architecture optimization allow AI to move from "understanding" to "generating aesthetics," accelerating content and product design efficiency. ABAB News · Cognitive Laws
Breakthroughs in model design capabilities lower the creative threshold again.
Access fairness is the cornerstone of trust; downgrading operations harm reputation.
Hard questions need to be open; AI progress relies on collective feedback.