Spelling Whizz

Exchange

Tax

Cars

German

OpenAI launches GPT-5, claims ‘PhD-level’ expertise


OpenAI launches GPT-5, claims ‘PhD-level’ expertise

ISLAMABAD: OpenAI has launched its most advanced AI model to date, GPT-5, calling it “smarter, faster, and more useful” than its predecessors, with capabilities reportedly reaching PhD-level expertise across multiple domains.

Unveiled by CEO Sam Altman ahead of its Thursday release, GPT-5 is being positioned as a unified AI system that combines fast responses for everyday queries with deep reasoning for complex problems.

It automatically chooses which mode to operate in – standard or “GPT-5 thinking” – based on the user’s prompt and task complexity.

This is the kind of intelligence that would have been unimaginable just a few years ago, Altman remarked, calling GPT-5 a new chapter in the evolution of artificial intelligence.

GPT-5 shines in writing, health, coding

OpenAI says the model offers major upgrades in real-world performance, significantly reducing hallucinations, improving instruction-following, and resisting sycophantic replies.

Slain AAC leader’s family in G-B facing threats as police fail to arrest culprits

GPT-5 particularly shines in writing, health, and coding – a space where it now rivals professional developers in generating and debugging complex applications and websites.

Plus users get enhanced access to GPT-5, while Pro subscribers can use GPT-5 Pro, a version with deeper reasoning capabilities.

A scaled-down version handles overflow when usage caps are met, with plans to merge all versions into a single, seamless experience in the future.

The launch further intensifies the AI race among tech giants, as firms like Google, Meta, and Anthropic push their own advanced models.

GPT-5’s arrival also follows OpenAI’s increasing efforts to position ChatGPT as a serious tool for professionals, especially in healthcare and software development.

Maryam Nawaz celebrates new start for out-of-school children

OpenAI has yet to disclose when GPT-5 will become the default model for API and enterprise customers, but early testers suggest its capabilities mark a major leap from GPT-4.

You May Also Like