Latest about GPT-4
OpenAI’s GPT-5: Smart upgrade or power-hungry mistake?
AI A new report suggests GPT-5 is a power guzzler compared to its predecessors, consuming 40 watt-hours of electricity to generate a medium-length response of about 1,000 tokens.
OpenAI Resurrects GPT-4o — But It’s No Longer Free
AI OpenAI has restored GPT-4o for ChatGPT following backlash — but the warmer, more personable model is now paywalled.
ChatGPT becomes a "coding beast" with the new GPT-4.1 update
AI OpenAI is bringing GPT-4.1 and GPT-4.1 mini to ChatGPT, and the new AI models excel in web development and coding tasks compared to OpenAI o3 & o4-mini.
OpenAI can develop GPT-4 with only 5 people now
AI OpenAI CEO Sam Altman recently indicated that the company is no longer compute-constrained, further claiming that the company can comfortably foster the development of advanced AI models.
NVIDIA just debuted its new open-source advanced AI model with state-of-the-art capabilities to take on OpenAI's 'magical' GPT-4o
OPEN-SOURCE AI NVIDIA recently unveiled powerful open-source multimodal large language models (LLMs) dubbed NVLM 1.0 with the flagship model, NVLM-D-72B featuring up to 72 billion parameters.
OpenAI's "collapse" fears reportedly date back to Chief Scientist Ilya Sutskever's abrupt departure — beyond recent bankruptcy reports with projections of $5 billion in losses within the next 12 months
COLLAPSED OPENAI High-level executives at OpenAI feared the AI firm would collapse following Chief Scientist Ilya Sutskever's abrupt departure.
OpenAI CEO Sam Altman admitted GPT-4 "kind of sucks," and now we learn the AI model consumes up to 3 water bottles to generate a mere 100 words
COOLING WATER According to new research by the University of California, OpenAI's GPT-4 AI model consumes up to 3 water bottles to generate 100 words.
Sam Altman's "GPT-4 is the dumbest model any of you will ever have to use" promise might reportedly come to fruition with a hint of 'Strawberry' in two weeks
STRAWBERRY OpenAI might be gearing up to release its much-anticipated 'reasoning' AI model, Strawberry, in two weeks.
