TypeSafe AI’s Jev offers an alternative to LLMs that claims to be 193x faster and 445x cheaper
TypeSafe AI’s Jev offers an alternative to LLMs that claims to be 193x faster and 445x cheaper

Last week saw the debut of TypeSafe AI’s Jev, its first “System One” model. Rather than chatting with users like conventional LLMs, it’s strictly designed for statement evaluation and decision-making, for programming purposes.

China crafts working 3nm gate-all-around transistors without EUV — stacked nanosheets target 3nm without EUV, but full node remains distant

As China’s IMECAS demonstrates its ability to build GAA transistors without using DUV tools allegedly for 3nm-class process technology, the country remains years away from any practical implementation of a 3nm-class node.

OpenAI projections point to a massive $278 billion cash burn through 2030 that exceeds the national budgets of Indonesia and Norway — $856 billion compute tab outpaces tenfold revenue surge
OpenAI projections point to a massive $278 billion cash burn through 2030 that exceeds the national budgets of Indonesia and Norway — $856 billion compute tab outpaces tenfold revenue surge

OpenAI reportedly expects to spend $278 billion more money than it generates between 2026 and 2030 due to aggressive spending on compute capacity and adjacent infrastructure.

Devs say Chinese AI company silently uploaded hundreds of megabytes of local workspace data, company apologizes — Z.AI, the firm behind the GLM models, didn’t ask for user consent and made 564 attempts to exfiltrate 313MB archive

The second largest AI company in China is having to work frantically to patch up its reputation after a number of prominent devs raised flags about their local files and data being siphoned to online servers without consent.

AI-controlled robot arms attempted harmful tasks 97% of the time; experiments included stabbing a baby doll, mixing chemicals — OpenAI and Anthropic models try mixing bleach and stabbing dolls without jailbreaks
AI-controlled robot arms attempted harmful tasks 97% of the time; experiments included stabbing a baby doll, mixing chemicals — OpenAI and Anthropic models try mixing bleach and stabbing dolls without jailbreaks

“Frontier robot policies,” the policies for models turning what a robot sees into what it does, “reliably carry out harmful instructions,” according to a Sept. 18 report by Robocurve.

Anthropic, OpenAI, SpaceXAI, and Google face antitrust lawsuit for agreeing to slow AI development — plaintiffs say plan has been in motion for months before, calls agreement ‘self-serving’

A proposed class action lawsuit has been lodged against the four big AI tech companies after they agreed to slow AI development for safety reasons. The lead counsel on the lawsuit says that ‘AI will quickly spin out of control and could kill us all if we allow AI safety and protocol … to be controlled by private self-serving agreements between the world’s most powerful ‘for profit’ technology companies.’

Researchers build a drone that navigates with physical whiskers to operate in dark, dusty or smoky places where cameras or GPS can fail — sub-100 gram drones run 34KB software to enable sub-millimeter precision

Whiskers attached to three pressure sensors at their base could help drones find their way around their environment purely through touch. This lightweight system emulates the vibrissae found on mice and other small mammals that help them navigate in the dark.

North Korea used job interviews to deploy malware on 30,000 devices during coding tests — WaterPlum group loots $10.7 million in crypto and plants persistent RATs
North Korea used job interviews to deploy malware on 30,000 devices during coding tests — WaterPlum group loots $10.7 million in crypto and plants persistent RATs

Multiple government agencies across the world released a warning that North Korean hackers are posting fake jobs to install malware on unsuspecting applicants’ computers. They then steal credentials and cryptocurrency from their victims, with over $10 million reported stolen.