Friday, September 12, 2025

Latest

New Study Reminds Us That ChatGPT Does Not Really *Understand* What You Want It To Do

Amid chatter about ChatGPT’s reportedly degrading performance, a new study found that recent open-sourced large language models (LLMs), including OpenAI’s GPT-3, perform “surprisingly better” on datasets released before the training data creation date than on datasets released after. 

The University of California, Santa Cruz paper by Changmao Li and Jeffrey Flanigan suggests that it isn’t that ChatGPT’s performance is degrading because new tasks are different from what the models are trained on, but that we forget that these models, especially the groundbreaking GPT-3, performed astoundingly well because they were trained with massive amounts of data, with a vast amount of examples of what is asked of them, and not particularly because they understand the tasks per se. 

As writing teacher and AI in education specialist Anna Mills puts it, it’s like “it has studied advance copies of lots of tests,” however, “when you give it new tests (tasks with no examples in its training data), it performs worse.”

The paper emphasizes that LLMs use a retrieval-based approach that mimics intelligence, as tech entrepreneur Chomba Bupe points out.

OpenAI may be having trouble catching up. Claims of getting “lazy” (which many have equated to “degrading”) have been plaguing OpenAI’s paid model, GPT-4, in recent weeks. 

The company, through its X account, explains that training a chat model “is not a clean industrial process,” and that it’s “less like updating a website with a new feature and more an artisanal multi-person effort to plan, create, and evaluate a new chat model with new behavior!”


Information for this story was found via X, and the sources and companies mentioned. The author has no securities or affiliations related to the organizations discussed. Not a recommendation to buy or sell. Always do additional research and consult a professional before purchasing a security. The author holds no licenses.

Video Articles

Avino Q2 Earnings: Steady As She Goes

MEG Energy: Cenovus Comes In With $7.9 BILLION Offer

Orla Mining Q2 Earnings: Musselwhite Changes Everything

Recommended

Northern Superior Expands Philibert With 350 Metre Step Out Testing 1.10 g/t Gold Over 25.5 Metres

Goliath Resources Hits 18.58 g/t Gold Over 5.00 Metres At Surebet

Related News

OpenAI Signs $11.9 Billion Deal with CoreWeave

OpenAI struck a five-year, $11.9 billion agreement with cloud provider CoreWeave, taking a direct $350...

Wednesday, March 12, 2025, 04:22:00 PM

Did CryptoGPT Raise $10 Million For A ChatGPT Knockoff?

CryptoGPT, a zero-knowledge (ZK) layer 2 blockchain, has raised $10 million in funding, capitalizing on...

Tuesday, April 11, 2023, 12:12:00 PM

OpenAI’s Soaring Revenue Shadowed by Billion-Dollar Losses, NYT Report Reveals

Artificial intelligence giant OpenAI is experiencing rapid revenue growth but facing significant financial challenges, according...

Monday, September 30, 2024, 10:56:00 AM

Microsoft Made Preparations to Absorb OpenAI’s Departing Staff — Except They Didn’t Leave

Microsoft (Nasdaq: MSFT) was getting ready to receive OpenAI employees who wanted to follow CEO...

Monday, November 27, 2023, 07:37:22 AM

China Arrests Man Over Release of ChatGPT-Generated Fake News

Today in that’s-probably-not-how-you-use-artificial-intelligence, a man in China has been arrested for reportedly using ChatGPT to...

Thursday, May 11, 2023, 06:14:00 AM