header-langage
简体中文
繁體中文
English
Tiếng Việt
한국어
日本語
ภาษาไทย
Türkçe
Scan to Download the APP
Partnered With:

OpenAI hit by an epic triple whammy: GPT-6.2 flops, Google's Neon goes viral, and Claude proves the Riemann Hypothesis.

Beating AI News Flash: AI blogger Binkg posted that, according to a friend working at OpenAI, internal morale at the company is low and it may be going through its toughest week since inception.


On November 17, OpenAI released GPT-6.2 Astra and Sol, but drew sharp criticism from multiple early testers. Well-known developer Theo said bluntly that although Astra is very smart, it is difficult to use in practical development. He once let the model run continuously for 11 days in an attempt to optimize a Rust compiler, but it did not submit any effective improvements and at one point slowed the program's running speed by 56%.


Theo also complained that Astra often fails to grasp the key points, falls into ineffective loops when encountering complex tasks, and after completing work uses a large pile of obscure jargon to explain simple problems. OpenAI product lead Tibo responded that early testing used an old version of the model and that extensive modifications had been made before the official release. After retesting, Theo acknowledged some improvement but was still dissatisfied.


Yet Google launched Gemini 4 Neon the next day. The new model's natural conversational style and programming capabilities won praise from many developers. Binkg believes that Google has already overtaken OpenAI this week.


Anthropic, meanwhile, stole the spotlight in mathematical research. On November 20, researcher Levent Alpöge announced collaboration with Claude to prove the Riemann Hypothesis, which has troubled the mathematics community for more than 160 years, and published about 17 million lines of Lean formal proof code. The next day, the Lean and Mathlib communities issued a joint statement confirming that, after verification by multiple independent tools, the proof is valid.


OpenAI's financial situation is also not easy. The company's third-quarter operating loss reached $10.9 billion. The latest funding round was originally planned to raise up to $30 billion, with a target valuation of $1.5 trillion. Now the valuation has dropped to $1.25 trillion, and only the first $11 billion of financing has been completed.


Binkg finally asserted that OpenAI no longer has any capability in which it can firmly remain the industry's number one, and that it "may really be finished."

举报 Correction/Report
Correction/Report
Submit
Add Library
Visible to myself only
Public
Save
Choose Library
Add Library
Cancel
Finish