Mastodawn

Stochastic Flocks and the Critical Problem of 'Useful' AI

이 글은 에이전트형 AI가 코드 생성과 자율적 문제 해결 능력을 갖추면서 사용자 경험이 개선되고 있지만, 근본적인 책임 문제와 편향, 비효율성 등 기존 AI 비판이 여전히 유효함을 강조한다. 에이전트 AI가 빠르게 생성하는 소프트웨어(슬롭웨어)는 검증 부족으로 위험을 내포하며, 반복적 코드 생성은 자원 낭비와 환경 부담을 가중시킨다. AI의 유용성 경계와 사회적 영향, 권력 구조 문제를 비판적으로 재검토할 필요가 있다고 지적한다.

https://www.techpolicy.press/stochastic-flocks-and-the-critical-problem-of-useful-ai/

#agenticai #llm #softwareengineering #aiethics #aiaccountability

Stochastic Flocks and the Critical Problem of 'Useful' AI

Acknowledging that AI systems are advancing does not buy into hype, it sharpens the precision of critical thinking about their impacts, says Eryk Salvaggio.

Tech Policy Press

GothamRider.AI May 13

Look at where the money goes. The world spent $1.5 trillion on AI development in 2025. The total spent on AI safety? $250 million. Less than two cents for every hundred dollars. OpenAI alone burned $8 billion. The entire US AI safety budget was $10 million. And even that's under threat. They gutted the safety agency and shifted its mission to competitiveness. This isn't a plan. It's a bet with your future. Follow for more. #AISafety #DigitalSovereignty #AIAccountability

Rachael Blizzard AI Governance Apr 29

Hi, I'm Rachael. I build AI governance frameworks centered on protecting people, not just enabling deployment.

My work:
• Data privacy & sovereignty—communities control their data
• AI Impact Assessments that center affected communities
• Governance frameworks that account for environmental & social costs

Background: Human-centered design → AI governance

Open to: Auditing, policy, advocacy, research, and community-focused governance

#AIGovernance #DataSovereignty #AIAccountability

MrB33n Apr 26

OpenAI's own employees flagged a user as dangerous. Management said: not imminent enough. Eight people were killed — including five children.

The chatbot was warned. The company wasn't listening. 🧵

https://chat-to.dev/view_trend?id=dkcwMTRvOUV2QWlnQlNTZ0ZBOE43Zz09

#AIAccountability #AISafety #ChatGPT

Man arrested in Singapore for leaking Avatar, could face up to 7 years in prison

Tycoon World Apr 24

The Rise of the Agentic Enterprise: Who Owns Decisions When AI Acts?

#TycoonWorld #AgenticAI #AutonomousAI #EnterpriseAI #AIGovernance #AIAccountability #AILeadership #AIInBusiness #AITransformation #AIDecisionMaking #AIRegulation #ResponsibleAI #AICompliance #AIEthics #DigitalTransformation #AITrends2026 #FutureOfWork #EnterpriseAutomation #AIInnovation #RiskManagementAI #AIOperations #AIInfrastructure

https://tycoonworld.in/the-rise-of-the-agentic-enterprise-who-owns-decisions-when-ai-acts/

The Rise of the Agentic Enterprise: Who Owns Decisions When AI Acts? – Tycoon World

THE TERM "Agentic AI" has moved rapidly from academic research into the boardrooms of the world's largest enterprises. Unlike the predictive and

Tycoon World

Marcus Schuler Apr 8

Anthropic's Mythos AI found genuine decades-old bugs in FreeBSD, OpenBSD, and FFmpeg that human auditors missed. But only 3 confirmed patches exist against claims of "thousands" of discoveries. 99% remain unverifiable until 2026. The gap between proven capability and marketing claims reflects broader trust issues in AI development. #AIcybersecurity #softwaredevelopment #AIaccountability https://www.implicator.ai/anthropics-mythos-found-real-bugs-thats-not-the-same-as-earning-trust/

Anthropic's Mythos Finds Real Bugs but Trust Deficit Grows

Three confirmed patches prove Anthropic's Mythos can find real bugs. But 99% of its claimed thousands of vulnerabilities remain unverifiable. The gap between safety narrative and commercial strategy grows harder to ignore.

Implicator.ai

PPC Land Apr 4

FYI: AI facial recognition locked up the wrong woman for 5 months: A Tennessee grandmother spent five months in jail after Clearview AI misidentified her as a North Dakota bank fraud suspect. What the case reveals about AI oversight gaps. https://ppc.land/ai-facial-recognition-locked-up-the-wrong-woman-for-5-months/ #AIFacialRecognition #JusticeForAll #AIAccountability #TechOversight #CivilRights

AI facial recognition locked up the wrong woman for 5 months

A Tennessee grandmother spent five months in jail after Clearview AI misidentified her as a North Dakota bank fraud suspect. What the case reveals about AI oversight gaps.

PPC Land

Show thread