OpenAI's superalignment team, co-led by @ilyasut
, has revealed its first research, exploring promising pathways to weak-to-strong model alignment (aka ways for puny humans to persuade ridonkulously smart AIs to obey them):
POLICY
-
OpenAI’s Superalignment Team Reveals Weak-to-Strong Model Alignment Research
By
–
-
AI Chatbots’ Political Persuasiveness: Stanford Research Findings
By
–
Politics are changing with chatbots that “talk” in human-like ways: A team from @StanfordPACS
’s Polarization and Social Change Lab and Stanford HAI explored the boundaries of AI’s political persuasiveness. -
Data Lineage in AI Models: Production Governance Requirements
By
–
Lineage basically means tracing data that went into a model/prompt, another part of a system coming together as we start to move into a prod phase. And AI is just a story about data moving around. It isn’t just about making sure an app breaks or whiffs governance requirements.
-
China tightens venture capital regulations amid cybersecurity concerns
By
–
#China venture capital, private equity funds rattled by proposed changes! https://
reuters.com/business/finan
ce/china-venture-capital-private-equity-funds-rattled-by-proposed-changes-2023-12-13/
… #RCE #ZeroTrust #ZeroDay #cybercrime #hacker #privacy #APT #bot #CISO #DDoS #hacking #phishing #CyberAttack #cybersecurity #Security #infosec #AppSec #CyberSec #databreach #Hacked -
AI Regulation Impact on Open Innovation: Policy Discussion
By
–
Yesterday, our Head of Public Policy Ben Brooks (
@opensauceAI
), joined Amanda Brock of @openuk_uk and Pete Cihon of @Github at @linuxfoundation 's AI dev to discuss the impact of AI regulation on open innovation in AI. In this informative conversation, they explore how AI -
Public Sector AI Must Be Inclusive and Effective
By
–
Public sector AI must be inclusive and effective for everyone. Experts, including @stanfordhai co-director @drfeifei and senior fellow Daniel E. Ho, weigh in on the conversation via @USATODAY
: -
Open Foundation Models: Policy Brief on Governance and Risks
By
–
New issue brief: Open foundation models have garnered much attention from policymakers around the world. Our latest brief highlights the benefits of open foundation models and calls for greater focus on their marginal risks. https://
hai.stanford.edu/issue-brief-co
nsiderations-governing-open-foundation-models
… -
Training Data Ethics: Copyright and LLM Model Development Concerns
By
–
This is a great point beyond the direct news implications…What does this deal mean for how quotes and other info in various stories are used to train OpenAI's models? Will people be worried their words and ideas might be misused by LLMs?
-
Tesla Recalls Autopilot Software in 2 Million Vehicles
By
–
Tesla Recalls Autopilot Software in 2 Million Vehicles:
-
San Francisco delays algebra education compared to early teaching methods
By
–
I taught basic algebra to my son when he was 5 and he learned it fine. Insane that in San Francisco they don't teach it until 9th grade.