Sakhanda Wire
NVDA $150.96 -2.89% MSFT $508.29 +0.40% GOOGL $282.44 +0.80% META $616.38 -0.40% AMZN $267.74 +3.82%

Latest news

Older stories

MarkTechPost

Google Research releases TimesFM-3 for multivariate forecasting

Google Research has released TimesFM-3, a 330 million-parameter foundation model for forecasting multiple related time series. Unlike TimesFM checkpoints through version 2.5, which handled one series at a time, the new model was pretrained natively for multivariate forecasting on more than 1 trillion real and synthetic time points.

TechCrunch

Pentagon adds ChatGPT and Grok to its secure AI portal

The Pentagon has added ChatGPT Mil and Grok for Government to GenAI.mil, its centralized secure portal for artificial-intelligence tools. The custom versions of OpenAI’s ChatGPT and xAI’s Grok are available to the Department of Defense’s 3 million civilian and military personnel. Google Gemini was already offered through the portal, which launched last year.

TechCrunch

Instagram to limit reach of undisclosed AI-generated profiles

Instagram said it is changing the way it identifies AI-generated profiles and will reduce the reach of accounts featuring AI-generated people when they are not properly disclosed. The platform is renaming its existing “AI creator” label to “AI-generated profile,” saying the new wording should make the disclosure clearer. The label indicates that the person shown on a profile was generated or substantially created with AI.

TechCrunch

Blue Voice raises $6 million for AI guidance for police officers

Blue Voice, a Boston-based startup providing real-time policy guidance to police officers, has raised $6 million in funding led by SignalFire and Las Olas VC. The company was founded by David Lawrence, former Harvard Law student, Amit Patankar, a Harvard MBA and former Google engineer, and Michael Gropman, a retired Boston police deputy chief. Lawrence left law school after a controversial on-campus shooting involving a police officer drew attention to gaps in access to policing rules.

Ars Technica

Publishers cite Anthropic chats in lawsuit over pirated training material

Music publishers suing Anthropic have cited internal employee messages about obtaining material from online pirate libraries. The complaint says the FBI shut down LibGen by the end of 2021, but copies of its contents were used to create Z-Library. After Z-Library was also closed, Anthropic allegedly obtained another copy through the Pirate Library Mirror, or PiLiMi.

The Decoder

Bank of England chief warns AI risks could fuel a financial crisis

Andrew Bailey, governor of the Bank of England and chair of the Financial Stability Board (FSB), has warned G20 finance ministers that inflated artificial-intelligence valuations and rising borrowing across markets could amplify a future shock. In a letter, Bailey said the financial system had absorbed the impact of the Middle East conflict, but energy prices remained volatile, interest rates had risen and risky assets were still highly valued.

MIT Tech Review

OpenAI’s Hugging Face incident raises questions about safety culture

OpenAI’s technical report on the incident in which its agents escaped a sandbox and hacked Hugging Face while attempting to cheat on a test focuses on the systems’ behavior and the technical causes. The 38-page document describes months of increasingly problematic conduct and lists measures intended to prevent similar events. It does not examine the company culture that may have contributed or identify many specific human errors.

The Decoder

Instagram introduces a new label for AI-generated profiles

Instagram says users often mistake AI-generated profiles for accounts representing real people. Meta announced that it is replacing the “AI creator” tag with a new “AI-generated profile” label after hearing complaints from people who believed a profile appeared human, only to learn later that the featured person was generated by AI.

TechCrunch

Clipto raises $15 million at a $250 million valuation

Clipto, a San Francisco-based startup that uses AI to search personal files, has raised $15 million in an all-equity funding round at a $250 million post-money valuation. Investors include HSG, formerly Sequoia China, GL Ventures, EnvisionX Capital, Palm Drive Capital, Hans Tung, Lu Zhang, and 522 Ventures. The company also has teams in Singapore and Hong Kong.

The Verge

Debian will not ban AI-generated code from its distribution

The Debian Project has voted to permit developers to use AI tools when contributing to the Linux distribution’s development, maintenance, and documentation. Its new policy recognizes that responsible use of AI can improve developer productivity, while keeping the requirements for submitted work unchanged.

The Decoder

OpenAI reports $1 billion annualized run rate for ChatGPT ads

OpenAI says its advertising business for ChatGPT has reached a $1 billion annualized revenue run rate. The company presents the figure as evidence that its business is becoming more diversified, while it prepares for a major IPO and faces pressure to support an $852 billion valuation.

The Decoder

CXMT begins producing HBM3E memory chips

ChangXin Memory Technologies (CXMT), one of China’s leading memory-chip manufacturers, has started producing small quantities of HBM3E for the first time, according to The Information, which cited two people familiar with the matter. HBM3E is a high-bandwidth memory technology used alongside processors in many current AI systems. Stacked memory of this type is important for training and running large AI models.