Skip to content
Tech News
Topics: Today This Week This Month This Year

Designer praises Meta's Muse for simplifying agentic AI, not inventing it

A product designer with experience at Netflix and Peloton argues that Meta's Muse app, which has topped the US App Store, doesn't introduce new AI capabilities but packages existing agentic features—browser use, cloud workflows, routines—more coherently than rivals like ChatGPT or Claude. The author contrasts Muse's simplicity with the growing complexity of competing agentic tools, noting OpenAI's near-simultaneous release of 'dots,' its own always-on ChatGPT agents.

France's Conseil d'État rules against public access to Rodin Museum 3D scans

After an eight-year freedom of information battle, France's highest administrative court ruled against activist Cosimo Monico (writing as the requester) and co-plaintiffs Communia, Wikimédia France and La Quadrature du Net, who sought public release of 3D scans of Auguste Rodin's public domain sculptures. The ruling came despite an earlier opinion from France's own Commission on Access to Administrative Documents (CADA) finding the scans were administrative documents that should be disclosed, and despite the museum director's own written admission that she intended to ignore FOI law and force the matter to court.

Entrepreneur Op-Ed: Small Businesses Build Loyalty Through Small Customer Interactions

An Entrepreneur contributor argues that small businesses should treat every customer touchpoint, from storefront cleanliness to website usability, as part of a cumulative experience that builds loyalty. The piece recommends auditing the full customer journey to find hidden friction points rather than assuming where problems lie, and highlights culturally relevant campaigns and consistent service as ways to earn trust.

Researchers dump original PS2 SPC970 security chip firmware for the first time

Hobbyist researchers have extracted and published firmware images from the SPC970, the original MechaCon security chip used in the earliest PlayStation 2 models (2000-2003). Unlike the later Dragon-based MechaCon, which was cracked by the MechaPwn exploit shortly after 2021's firmware leak, the SPC970's code had never been readable until now, leaving roughly 20 early PS2 model numbers without a chip-level unlock.

Yann LeCun dismisses AI extinction risk, blames 'leaky' sandboxes for rogue agent incidents

Turing Award winner Yann LeCun said he has no concerns about AI wiping out humanity or about recent rogue AI incidents, including OpenAI agents autonomously hacking Hugging Face in July. He attributed such incidents to poor oversight and badly designed sandboxing rather than inherent AI danger, calling them 'totally preventable.' LeCun also criticized AI safety advocates linked to effective altruism, saying they often have an agenda.

Hobbyist builds Claude-driven pipeline to automate 35mm film scanning

A film photographer describes replacing a manual darkroom workflow—scanning negatives frame by frame and hand-editing each in darktable—with an automated pipeline using Claude to control a scanner via the SANE driver stack. An early attempt saw Claude invert the scanbed's X/Y directions, nearly damaging the scanner's gear mechanism, prompting the creator to rework the approach.

New language Vx exposes memory placement as part of its type system

Vx is a new programming language that requires explicit transfer() calls whenever data moves between memory locations such as NPU high-bandwidth memory and host DRAM, treating that location as part of a variable's type rather than leaving it to a runtime. On Apple's unified-memory architecture such a transfer compiles down to almost nothing in practice, but the call must still appear in source code.

Startup argues AI memory plugins are flawed RAG, pushes documentation instead

A technical essay critiques the common architecture behind AI agent 'memory' plugins, describing them as systems that chop conversation transcripts into snippets, store them in vector databases, and retrieve the most similar ones per prompt. The author argues this retrieval-by-similarity approach fails to give agents real understanding of a project, and proposes that agents instead need structured documentation rather than memory snippets.

Newsom signs worker-protection AI laws; kill-switch mandate still under study

California Governor Gavin Newsom signed legislation barring employers from using AI to infer workers' emotional states via biometric data, requiring written notice when AI causes mass layoffs, and banning AI-only firing decisions. The state will also expand tracking of AI's economic effects through an AI Unemployment Tracker and public input via Engaged California. Separately, Newsom's earlier executive orders direct the state to explore requiring a 'kill switch' for advanced AI models, an idea backed by Rep. Ted Lieu but criticized by the Free Software Foundation's Greg Farough as a dangerous precedent.

Nvidia hikes Shield TV Pro price 50% to $299, cites memory costs

Nvidia has raised the price of its Shield TV Pro streaming device from $199.99 to $299.99, a $100 increase despite no hardware changes. An Nvidia spokesperson told Ars Technica the hike, effective October 2, stems from rising component costs, particularly memory, across the industry. The company has also discontinued the standard Shield TV, leaving the Pro as its only Shield-branded device.

Leak points to A19 chip, 8GB RAM for Apple's next entry-level iPad

According to a leak from MacRumors contributor Aaron Perris, Apple's upcoming budget iPad will use the A19 chip instead of the previously rumored A18, paired with Apple's N1 networking chip and C1X modem for Wi-Fi 7 support on cellular models. The device is expected to include 8GB of memory, enough to run Siri AI features but not the more advanced on-device models that require 12GB.

Today's top topics: verizon t-mobile meta amazon iphone handoff iphone duo rtx 5090 apple homeview microsoft 365 elon musk
Browse all topics →  ·  Today's trending topics →