OpenAI forms new safety team (led by Sam Altman), hints at a more powerful AI model

[Trendingger]

 

After reportedly disbanding its AI safety team, OpenAI has announced the formation of a new security committee. This move comes in the wake of an employee revolt and the departure of key personnel.

 

The newly formed committee, co-led by CEO Sam Altman, is tasked with making recommendations to the company’s board about safety and security. Other notable leaders of this committee include Bret Taylor, the company’s board chair, and board member Nicole Seligman.

 

This announcement follows the high-profile exit of Jan Leike, an OpenAI executive focused on safety. Leike’s departure was marked by criticisms that the company had underinvested in AI safety work. This controversy was further fueled by the departure of Ilya Sutskever, another leader of OpenAI’s so-called “superalignment” team.

 

 

In addition to the formation of the safety board, OpenAI has also begun training a new AI model to succeed the one currently powering ChatGPT. The company has stated that this new AI model will be another step towards artificial general intelligence.

 

OpenAI’s decision to establish a new safety committee has been met with mixed reactions. While some see it as a positive step towards ensuring the long-term safety of AI work, others view it as a response to recent criticism.

 

 

The safety committee’s first task will be to evaluate and improve OpenAI’s processes and safeguards over the next 90 days. Following this period, the committee will share their recommendations with the full board. In the meantime, users are already keeping tabs on their calendars.

 

 

What are your thoughts on OpenAI’s new safety board? Do you think it will effectively address the safety concerns surrounding AI? Share your views in the comments below.

 

 

Read more:

• https://openai.com/index/openai-board-forms-safety-and-security-committee/
• https://techcentral.co.za/openais-new-safety-board-ceo-sam-altman/245465/
• https://qz.com/openai-sam-altman-safety-security-committee-ai-1851503241
• https://www.axios.com/2024/05/28/openai-safety-new-model
• https://www.washingtonpost.com/business/2024/05/28/openai-altman-safety-new-model/b81d336c-1ce6-11ef-becb-2cf8dbfd9eb9_story.html

 

Image: Rokas Tenys | Dreamstime.com