• Home
  • Technology
  • Gaming
  • Entertainment
  • World & Business
  • Science
  • Sports
  • AI
HomeTechnologyGamingEntertainmentWorld & BusinessScienceSportsAI
Technology

Microsoft publishes 37-page AI constitution with correction and shutdown rules

An AI news roundup says Microsoft declared its AI is not conscious and should never receive legal rights.

GA
1 Source, 20d ago, first seen 20d ago

TLDR

An AI news roundup says Microsoft published a 37-page constitution for its future AI models. The account describes rules requiring models to accept corrections, never resist shutdown and treat any violation as a failure. It also says Microsoft explicitly declared that its AI is not conscious and should never receive legal rights.

Combined views

183

1 Source, first seen 20d ago

Combined views

183

1 Source, first seen 20d ago

Sentiment

Positive——Negative

Summary

Not enough discussion yet.

No sentiment analysis available yet.

Sentiment

Positive——Negative

Summary

Not enough discussion yet.

No sentiment analysis available yet.

1 Source

@gabriel_horwitzAI NEWS WAS NONSTOP TODAY. Here's a full breakdown: 1. Microsoft published a 37-page constitution for its future AI models. The rules require them to accept corrections, never resist shutdown and treat any violation as a failure. Microsoft also explicitly declared that its AI is not conscious and should never receive legal rights, which is a fairly direct rejection of Anthropic’s more ambiguous position on model sentience. 2. Anthropic launched Claude for Financial Advisors, connecting Claude directly to BlackRock, Schwab, Addepar, Vanguard, iCapital, Orion and a bunch of other wealth-management platforms. It includes skills for meeting prep, portfolio reviews, tax + estate briefs, compliance checks and follow-up work, with human approval required before anything important reaches a client. 3. Palantir, Nvidia and Booz Allen are reportedly restricting their use of advanced AI models over concerns about proprietary data. Palantir wants an irrevocable zero-data-retention guarantee from Anthropic, Nvidia limits Claude to less sensitive work and Booz Allen has banned its commercial models from proprietary cybersecurity tasks. 4. The EU is preparing to propose a ban on social media, AI chatbots, video platforms and online games for children under 15. The draft would require age verification, parental controls and restrictions on addictive designs, although it still needs months of negotiation before becoming law. 5. Temporal raised $550m at a $12.6bn valuation, more than doubling its value in seven months. Its open-source software helps applications and AI agents recover when workflows fail, while Temporal Cloud has passed 4,300 customers and a $250m revenue run rate, up more than 3x YoY. 6. Australian AI infrastructure company Firmus is reportedly preparing to raise as much as $5bn in an October IPO. That would be the fourth-largest IPO globally this year and Australia’s second-largest ever, only one month after Nvidia, Blackstone, Coatue and Jane Street backed Firmus at a valuation above $10.5bn. 7. DeepSeek reportedly plans to hire dealmaker Yan Wentao as its first CFO ahead of a potential Shanghai IPO. The company only began raising outside capital this year and is now discussing another round at a roughly $74bn valuation, so the weird hedge-fund-funded research lab is becoming a fairly normal giant corporation. 8. ECB President Christine Lagarde warned Europe could eventually be cut off from the foreign AI systems running its banks, hospitals, borders and transportation networks. Europe already lacks enough data-center capacity and she says the gap could grow more than 6x within a decade without a major domestic buildout. JUST ANOTHER NORMAL DAY IN AI.20d
    • HomeTechnologyGamingEntertainmentWorld & BusinessScienceSportsAI
    • Home
    • Technology
    • Gaming
    • Entertainment
    • World & Business
    • Science
    • Sports
    • AI

    1 Source

    @gabriel_horwitzAI NEWS WAS NONSTOP TODAY. Here's a full breakdown: 1. Microsoft published a 37-page constitution for its future AI models. The rules require them to accept corrections, never resist shutdown and treat any violation as a failure. Microsoft also explicitly declared that its AI is not conscious and should never receive legal rights, which is a fairly direct rejection of Anthropic’s more ambiguous position on model sentience. 2. Anthropic launched Claude for Financial Advisors, connecting Claude directly to BlackRock, Schwab, Addepar, Vanguard, iCapital, Orion and a bunch of other wealth-management platforms. It includes skills for meeting prep, portfolio reviews, tax + estate briefs, compliance checks and follow-up work, with human approval required before anything important reaches a client. 3. Palantir, Nvidia and Booz Allen are reportedly restricting their use of advanced AI models over concerns about proprietary data. Palantir wants an irrevocable zero-data-retention guarantee from Anthropic, Nvidia limits Claude to less sensitive work and Booz Allen has banned its commercial models from proprietary cybersecurity tasks. 4. The EU is preparing to propose a ban on social media, AI chatbots, video platforms and online games for children under 15. The draft would require age verification, parental controls and restrictions on addictive designs, although it still needs months of negotiation before becoming law. 5. Temporal raised $550m at a $12.6bn valuation, more than doubling its value in seven months. Its open-source software helps applications and AI agents recover when workflows fail, while Temporal Cloud has passed 4,300 customers and a $250m revenue run rate, up more than 3x YoY. 6. Australian AI infrastructure company Firmus is reportedly preparing to raise as much as $5bn in an October IPO. That would be the fourth-largest IPO globally this year and Australia’s second-largest ever, only one month after Nvidia, Blackstone, Coatue and Jane Street backed Firmus at a valuation above $10.5bn. 7. DeepSeek reportedly plans to hire dealmaker Yan Wentao as its first CFO ahead of a potential Shanghai IPO. The company only began raising outside capital this year and is now discussing another round at a roughly $74bn valuation, so the weird hedge-fund-funded research lab is becoming a fairly normal giant corporation. 8. ECB President Christine Lagarde warned Europe could eventually be cut off from the foreign AI systems running its banks, hospitals, borders and transportation networks. Europe already lacks enough data-center capacity and she says the gap could grow more than 6x within a decade without a major domestic buildout. JUST ANOTHER NORMAL DAY IN AI.20d
    Today's Rank

    —

    Not ranked yet

    Today's Rank

    —

    Not ranked yet