OpenAI Pursues New Way To Fight AI “Hallucinations”

It’s been just revealed that OpenAI is pursuing a new way to fight AI hallucinations. Check out the latest reports below.

On Wednesday, OpenAI revealed its latest initiative to combat AI “hallucinations” through a more advanced approach to training AI models.

This study is being conducted during a critical period of discussion on the issue of misinformation originating from AI systems. This is especially relevant given the surge in generative AI technology and the upcoming 2024 U.S. presidential election.

Last year, ChatGPT, the chatbot created by OpenAI using GPT-3 and GPT-4, contributed significantly to the growth of generative AI. In just two months, it attracted over 100 million monthly users, setting a record as the fastest-growing app.

To date, Microsoft has invested more than $13 billion in OpenAI, and the startup’s value has reached roughly $29 billion.

Sometimes, AI models such as OpenAI’s ChatGPT or Google’s Bard may create false information and present it as facts, leading to what is known as AI hallucinations.

Here’s an instance to consider: Google’s promotional video for Bard in February featured a chatbot that made a false statement about the James Webb Space Telescope. Additionally, ChatGPT mentioned “bogus” cases in a recent filing in a New York federal court, which could result in sanctions for the involved attorneys.

It seems that Microsoft and Twitter might have a beef. Check out what happened and what made Elon Musk to throw a threat at Twitter.

Here are the tweet that caught our attention earlier today:

They trained illegally using Twitter data. Lawsuit time.

— Elon Musk (@elonmusk) April 19, 2023

The notes continue and reveal the following:
“Starting on April 25, 2023, Smart Campaigns with Multi-platform will no longer support Twitter.
You’ll be unable to:
– Access your Twitter account through our social management tool
– Create and manage drafts or Tweets
– View past Tweets and engagement
– Schedule Tweets”

Elon states that this is related to illegally using Twitter Data. https://t.co/NZ2lUGwgo2

— T(w)itter Daily News  (@TitterDaily) April 19, 2023

You can check out more data about the issue in Microsoft’s notes.