Sakhanda Wire
NVDA $155.46 -0.44% MSFT $506.24 -2.27% GOOGL $280.20 -1.24% META $618.87 -1.20% AMZN $257.89 +0.05%

Latest news

Older stories

Ars Technica

Researchers report weaker safeguards against explicit AI images

Safety researchers say a large amount of AI-generated explicit material continues to circulate without effective control, while regulators have done little to stop its spread. The material is often non-consensual and may be sent through text messages and direct messages or kept in private cloud storage, making it less likely to be detected or reported.

TechCrunch

Scimplify raises $25M to expand AI-driven materials production

Indian startup Scimplify has raised $25 million in an all-equity strategic round to expand its AI-supported production of specialty chemicals and critical materials. Hitachi Ventures led the financing, which also included customers Henkel and Teknor Apex as new investors. Accel India, Bertelsmann India Investments, Omnivore and 3one4 Capital participated again. The company said total equity funding has reached $80 million.

MarkTechPost

Bifrost Leads 2026 Ranking of AI Gateways

A review of nine AI gateways for 2026 places Bifrost first among self-hosted options, followed by Portkey, LiteLLM, Kong AI Gateway, TrueFoundry, Cloudflare AI Gateway, Vercel AI Gateway, OpenRouter and Orq.ai Router. The comparison considers performance, provider coverage, governance, deployment choices, pricing and vendor or supply-chain risk.

MarkTechPost

Liquid AI releases 24B MoE model for local AI agents

Liquid AI has released LFM2.5-24B-A2B, an open-weight, text-only reasoning model designed for agents running on laptops, PCs and mini PCs. The Mixture-of-Experts model contains 24B parameters but activates about 2B per token, or 2.45B according to its model card. It has a 131,072-token context window and is the largest and final model in the LFM2.5 family.

MarkTechPost

Google Makes Gemini 4 Argon Generally Available

Google DeepMind has made Gemini 4 Argon generally available to developers, enterprises and paid Gemini subscribers through the Gemini API, AI Studio and the Gemini platform. The sparse mixture-of-experts model is designed for extended, multi-step tasks in coding, finance, legal work, science and cybersecurity. Google first introduced it on September 30 to a limited group of cyber defenders.

DeepMind

Singapore develops on-device AI model to detect conversational scams

Google DeepMind and Singapore’s Government Technology Agency (GovTech) have developed an on-device model designed to identify scams in private conversations. The initiative responds to a growing threat: scams represent more than half of Singapore’s reported crimes and have caused over S$4 billion in losses since 2019. Social-engineering schemes involving convincing conversations now account for as much as 80% of losses.

TechCrunch

Science Corporation implants PRIMA in first Stargardt patient

Science Corporation has implanted its PRIMA vision-restoring chip in its first patient with Stargardt disease, an inherited condition that progressively damages central vision. CEO Max Hodak said the procedure took place in May at Sydney Eye Hospital in Australia as part of an ongoing trial. During rehabilitation, the patient first identified a letter and then read a word in the following session.

TechCrunch

Cerebras CEO predicts personal AI assistants in every home

Cerebras Systems CEO and co-founder Andrew Feldman said the arrival of AI appliances in homes may be closer than widely assumed. Speaking at TechCrunch Disrupt 2026, he pointed to consumer-facing AI agents such as Meta’s Muse and Instinct as signs that the technology is gaining traction with everyday users. Instinct is now valued at $10 billion, according to the article.

TechCrunch

Industrial Next develops robots that can switch factory tasks within minutes

San Francisco-based Industrial Next is developing self-contained robotic workstations designed to handle changing manufacturing tasks without conventional programming. The startup says its AI systems combine vision, task context and movement through proprietary foundation models, allowing production lines to be reconfigured in minutes instead of the weeks or months often required for reprogramming and physical changes.

MarkTechPost

Meta releases Muse Wonder models for long-running AI agents

Meta has released Muse Wonder and Muse Wonder Full, two models designed for extended agent tasks such as software development, legal and financial work, and computer use. Both support a 1M-token context window and accept text, images, video and audio as input, while producing text. They run on Meta’s cloud through the Meta Model API or the Muse Code command-line tool.

TechCrunch

Cerebras CEO urges balance in debates over AI risks

Cerebras Systems CEO and co-founder Andrew Feldman has criticized the increasingly prominent warnings about artificial intelligence risks. Speaking at TechCrunch Disrupt 2026, he said some AI leaders may be speaking too publicly about scenarios involving loss of control and the possibility that AI could eventually kill humanity. In his view, this focus is obscuring the technology’s potential benefits.

MarkTechPost

Tutorial examines Meta’s Rebalancer solvers and assignment checks

A coding tutorial examines Meta’s Rebalancer, an assignment solver designed to place tasks across datacenter resources. The implementation combines a C++ core with Python bindings and introduces the main concepts used to describe models, including objects, containers, dimensions, scopes and partitions. It also shows how to define constraints and prioritize goals.