The AI Breakdown: Daily Artificial Intelligence News and Discussions
White House Secures Responsible AI Commitment from OpenAI, Meta, Google, Anthropic and Others
Fri Jul 21 2023Description
Sergey Brin has returned to Google and is actively working on AI. He is involved in technical discussions about Google's forthcoming AI model Gemini and convenes weekly discussions of new AI research with employees. Brin's intervention in personal matters such as hiring sought-after researchers is notable. Google's Gemini project aims to build a general-purpose AI that can rival GPT-4, with a potential release by the end of the year. Senator Bob Casey introduced two bills related to AI in the workplace, one preventing employers from using entirely automated systems for employment decisions and another creating a task force to study workplace surveillance tools. The White House has been engaging with AI companies on voluntary commitments to ensure safe and transparent development of AI technology. Advancements in AI development include Runway Gen 2's ability to generate videos from starting images and ChatGPT's new feature called Custom Instructions. Companies commit to publicly reporting their AI systems' capabilities, limitations, and appropriate use, as well as prioritizing research on societal risks such as harmful bias and discrimination. Concerns about biases in training data leading to biased AI systems persist. The effectiveness of industry commitments compared to legislation is a subject of debate, but there has been a positive response from some analysts and industry leaders regarding the White House's approach. Microsoft and other companies have expressed support for the commitments, although more needs to be done in terms of technical safety and societal oversight of AI development.
Insights
Sergey Brin's Return to Google
Sergey Brin, co-founder of Google, has returned to the company and is actively working on AI. He is involved in technical discussions about Google's forthcoming AI model Gemini and convenes weekly discussions of new AI research with employees.
AI in the Workplace and Government Initiatives
Senator Bob Casey introduced two bills related to AI in the workplace: the No Robot Bosses Act and the Exploitative Workplace Surveillance and Technology's Task Force Act. The White House has been engaging with leading AI companies on voluntary commitments to ensure safe, secure, and transparent development of AI technology.
Advancements and Concerns in AI Development
Recent advancements include Runway Gen 2's ability to generate videos from starting images and ChatGPT's new feature called Custom Instructions. Companies commit to publicly reporting their AI systems' capabilities, limitations, and appropriate use, as well as prioritizing research on societal risks such as harmful bias and discrimination. Concerns about biases in training data leading to biased AI systems persist.
Chapters
- Sergey Brin's Return to Google and AI Work
- AI in the Workplace and Government Initiatives
- Advancements and Concerns in AI Development
Sergey Brin's Return to Google and AI Work
00:01 - 06:24
- Sergey Brin, co-founder of Google, has returned to the company and is actively working on AI
- Brin is involved in technical discussions about Google's forthcoming AI model Gemini
- He convenes weekly discussions of new AI research with employees
- Brin has intervened in personal matters such as hiring sought-after researchers
- Google's Gemini project aims to build a general-purpose AI that can rival GPT-4
- Gemini could be released by the end of the year
AI in the Workplace and Government Initiatives
06:15 - 13:12
- Senator Bob Casey introduced two bills related to AI in the workplace
- The No Robot Bosses Act would prevent employers from using entirely automated systems for employment decisions
- The Exploitative Workplace Surveillance and Technology's Task Force Act would create a task force to study workplace surveillance tools
- The White House has been engaging with artificial intelligence (AI) and collaborating with leading AI companies on voluntary commitments.
- The goal is to ensure safe, secure, and transparent development of AI technology.
- Companies involved in these commitments include Amazon, Google, OpenAI, and Tropic.
- Commitments include internal and external security testing of AI systems before release, sharing information on managing AI risks, investing in cybersecurity safeguards for proprietary model weights, facilitating third-party discovery of vulnerabilities in live AI systems, developing mechanisms to identify AI-generated content, publicly reporting capabilities and limitations of AI systems including security risks and societal risks like fairness and bias, prioritizing research on societal risks posed by AI systems such as harmful bias and discrimination.
Advancements and Concerns in AI Development
12:44 - 18:02
- Runway Gen 2 can generate videos from starting images
- ChatGPT rolled out a new feature called Custom Instructions
- Companies commit to publicly reporting their AI systems capabilities, limitations, and areas of appropriate and inappropriate use.
- Commitment to prioritize research on societal risks of AI systems, including harmful bias, discrimination, and privacy protection.
- Concerns about biases in training data leading to biased AI systems.
- Commitment to develop and deploy advanced AI systems to address society's challenges.
- Consultation with various countries on similar voluntary commitments.
- Debate over the effectiveness of industry commitments compared to legislation.
- Positive response from some analysts and industry leaders regarding the White House's approach.
- Statements from Microsoft and other companies expressing support for the commitments.
- Acknowledgment that more needs to be done in terms of technical safety and societal oversight of AI development.