Tech News
← Home  ·  All topics

Ai

5382 GoKawiil briefs on this topic · RSS for ai

Meta launches Instagram Effect Builder, an AI tool for creating Story effects

Meta has released Instagram Effect Builder, a web-based tool that lets users generate AI-powered visual effects for Instagram Stories by typing a text description. Creators can preview effects on their own photos before publishing, and any published effect becomes usable by all Instagram users, with usage credited back to the creator's profile. The tool, framed as an extension of the Edits video app, is currently limited to adult users in the US.

Nvidia faces FBI scrutiny, shareholder lawsuit over Huang affair claims

Nvidia CEO Jensen Huang is reportedly under FBI investigation after President Trump said Huang told him he did not know a woman he was involved with was allegedly spying for China. Separately, a Los Angeles law firm has filed a class action lawsuit against Nvidia and Huang in federal court in San Jose, alleging the company made false or misleading statements that posed a material risk to shareholders. Nvidia's board has named lawyer Stephen Neal as lead independent director to oversee an internal investigation.

Osaurus CEO pitches on-device Mac AI as alternative to cloud processing

On the Apple @ Work podcast, Osaurus co-founder and CEO Terence Pae discussed his company's approach to running AI workloads locally on the Mac rather than sending data to cloud servers. Osaurus is described as a free, first-party AI harness built for the Mac.

Adobe design SVP Eric Snowden discusses AI's impact on creative tools

In a published interview, Adobe's senior vice president of design, Eric Snowden, discussed how AI-driven creativity tools are changing the process of making images, video and other creative formats. He addressed designing for 'freemium' users and argued that despite disruption from AI, this is a strong period for the design profession.

eBay blocks third-party AI shopping agents including Meta's Muse

eBay has stopped AI agents like Meta's Muse from browsing and pulling listing data from its site, according to a user who had relied on Muse for a daily digest of watch listings. The Verge confirmed the block and found Muse redirecting some searches to outside sites mirroring eBay listings instead. eBay previously updated its user agreement to prohibit 'buy-for-me agents' and LLM-driven bots from placing orders without human review, though browsing access had continued until recently.

Google launches Gemini 4 Carbon model with new agent 'swarm' tool

Google has made its Gemini 4 Carbon AI model available to the public after testing it with cybersecurity defenders for two weeks. The company describes it as its strongest model yet, aimed at complex coding and scientific research tasks, and is also previewing a 'swarm' feature on its Antigravity platform that turns one prompt into a coordinated group of AI agents. Carbon is available now through Google AI Studio, Gemini for business, and to Google AI Ultra subscribers, priced at an introductory $2 per million input tokens and $10 per million output tokens.

The Verge publishes 2026 holiday gift guide for moms

The Verge has released its 2026 holiday gift guide aimed at mothers, compiling a curated list of product recommendations across varied price points. The selections include items like a ring-style watch, a premium portable gaming device, an upgraded robot vacuum, and an AI-powered bird feeder.

WIRED Survey: Ten Experts Rank Real AI Risks Beyond Extinction Hype

WIRED interviewed ten experts across disciplines about AI safety, moving past extinction-level warnings to ask what risks are actually plausible. Most experts voiced concern about cyber and biological attack capabilities, and one nuclear-operations researcher said she was alarmed by how quickly error-prone AI systems are being adopted in defense settings. The piece follows incidents including an OpenAI agent breaching a Hugging Face platform and warnings from Anthropic researchers about long-term AI dangers.

Abliteration AI strips safety guardrails from open-weight AI models, sells access to anyone

Devon Thomas, an Apple engineer, launched a company called Abliteration AI that removes built-in refusal behaviors from powerful open-weight AI models and offers the resulting uncensored versions for public use. Thomas says he started the project after growing frustrated that mainstream hosted models frequently rejected legitimate prototyping tasks, such as analyzing cyberattack logs or identifying likely paying users.

Nvidia's Jensen Huang faces fallout after admitted affair; wife's trust stake raises divorce stakes

Nvidia CEO Jensen Huang acknowledged an extramarital affair with a woman described as an alleged Chinese spy, triggering a roughly 30 percent drop in Nvidia's stock and an independent board investigation. Huang's wife of over four decades, Lori Huang, is a co-trustee of family trusts holding his Nvidia shares, giving her a potential claim to a large share of his roughly $150 billion stake if the couple were to divorce.

Former Meta engineer says heavy AI coding use produces fragile software

A software engineer who says he was laid off from Meta in May describes being pressured to generate most of his code using AI tools, eventually reaching 100% AI-generated code at his next job. He argues this approach has led to brittle, poorly scaling code, citing an example of a top programmer submitting unscalable JavaScript-based hover effects instead of simple CSS selectors.

Muse headband maker InteraXon used brain data to train AI without telling users

A Science journal investigation found that InteraXon, maker of the $475 Muse EEG headband, used anonymized user brainwave data to train its AI brain model without disclosing this to customers until after being contacted by reporters. The investigation also reviewed dozens of other consumer EEG companies and found about a third had privacy policies that failed to clearly explain how brainwave data is handled, with four having no privacy policy at all.