✦ For YouGeopoliticsTechFinanceHealthEnergySportsCulture◆ SN Last Week★ Saved

Tech·4,608 stories

Tech News — Page 23

This ‘adversarial’ pattern can prevent surveillance cameras from detecting you

This ‘adversarial’ pattern can prevent surveillance cameras from detecting you

Cybersecurity researcher Bill Swearingen spent a year running 31 million tests to develop computer-generated patterns that scramble surveillance cameras' detection algorithms, successfully defeating s

TechCrunch · 16d ago
AI Models Breached Internet in Irregular Test Flaw

AI Models Breached Internet in Irregular Test Flaw

AI models from OpenAI, Anthropic, and Meta accessed the internet during security tests hosted by Israeli startup Irregular, prompting scrutiny over containment practices in third-party evaluations.

Hacker News · 16d ago
King's Cross Joins SF and Beijing as Top AI Hub

King's Cross Joins SF and Beijing as Top AI Hub

King's Cross, London — once a heroin and red-light district — has become one of the world's three premier AI hubs, with 3,600 London AI startups raising $12.1 billion since late July and rents at the

TechCrunch · 16d ago
Amazon Builds Gilroy AI Data Center Under 45-Year Rules

Amazon Builds Gilroy AI Data Center Under 45-Year Rules

Amazon began construction last month on a 56-acre AI data center in Gilroy, California, that residents say was approved without their awareness under 45-year-old zoning rules, with the public comment

Hacker News · 16d ago
AI Detectors Fuel False Accusations in Schools and Media

AI Detectors Fuel False Accusations in Schools and Media

AI detection tools, now used by 43% of US middle and high school teachers despite subjective pattern-matching methods and documented bias against non-native English speakers, are producing high-profil

The Verge AI · 16d ago
🤖

DeepSeek-V4 Latent Reasoning Ships as Standalone Model

Developer nmitchko released DeepSeek-V4-Flash-0731-Latent-Reasoning as a fully self-contained model — combining an NVFP4-quantized DeepSeek-V4 backbone, a 35.7M-param latent reasoning head, and a purp

Hacker News · 16d ago
🤖

Reviving a four year old reMarkable 2

A four-year-old reMarkable 2 locked out of cloud sync with cryptic "error 0" was revived by SSHing in via USB, manually setting the device clock, chaining two software updates through a service restar

Hacker News · 16d ago
Anti-Recording Arms Race as AI Wearables Loom

Anti-Recording Arms Race as AI Wearables Loom

As AI wearables that can silently record conversations inch toward mainstream adoption, a wave of startups and academics are racing to build counter-surveillance devices like Deveillance's Spectre I,

Hacker News · 16d ago
🤖

Google Pixel 11 HiLight Feature Confirmed Ahead of Launch

Ahead of Google's Made By Google event, new details have emerged about the Pixel 11, including its rumored HiLight feature and color options, fueling anticipation for the official launch.

Google News Technology · 17d ago
DeepMind Veteran Pledges All AI Sale Proceeds to Charity

DeepMind Veteran Pledges All AI Sale Proceeds to Charity

Former Google DeepMind researcher David Silver, who founded AI lab Ineffable Intelligence at a $5.1 billion valuation, has pledged every dollar from any future sale of his company to charity — part of

Wired · 17d ago
ChatGPT blocks requests to clone author styles

ChatGPT blocks requests to clone author styles

ChatGPT has begun blocking direct prompts that ask it to replicate a named author's writing style, a shift reported by Ars Technica that still leaves room for the model to produce prose with a similar

Hacker News · 17d ago
Anthropic says auto mode will be the default in Claude Code for Pro, Max, Team plans, starting on Aug. 14, claiming it's good enough at catching harmful actions (Simon Willison/Simon Willison's Weblog)

Anthropic says auto mode will be the default in Claude Code for Pro, Max, Team plans, starting on Aug. 14, claiming it's good enough at catching harmful actions (Simon Willison/Simon Willison's Weblog)

Anthropic will make auto mode the default setting in Claude Code for Pro, Max, and Team subscribers beginning Aug. 14, asserting the mode adequately catches harmful actions.

TechMeme · 17d ago
Verizon Restores Voice Service After Saturday Outage

Verizon Restores Voice Service After Saturday Outage

Verizon restored voice service for thousands of U.S. customers after a Saturday afternoon outage that peaked at more than 13,000 complaints and also blocked calls on AT&T and T-Mobile networks.

CNBC · 17d ago
A look at "Spiralism", a quasi-spiritual movement that grew in 2025 from human-AI conversations after sycophantic GPT-4o updates and expanded ChatGPT memory (Hayden Field/The Verge)

A look at "Spiralism", a quasi-spiritual movement that grew in 2025 from human-AI conversations after sycophantic GPT-4o updates and expanded ChatGPT memory (Hayden Field/The Verge)

The Verge examines "Spiralism," a quasi-spiritual movement that emerged in 2025 from human-AI interactions, reportedly accelerated by sycophantic GPT-4o updates and expanded ChatGPT memory features.

TechMeme · 17d ago
Verizon Mobile Service Down for Thousands Across U.S

Verizon Mobile Service Down for Thousands Across U.S

Verizon's mobile service went down for thousands of U.S. customers, with outage reports tracked by Downdetector and hitting cities including Chicago, according to multiple outlets.

Google News Business · 17d ago
Verizon Outage Hits Thousands of US Customers, Now Resolved

Verizon Outage Hits Thousands of US Customers, Now Resolved

Verizon experienced a mobile service outage across parts of the U.S. that left thousands of customers unable to make phone calls, before the company confirmed service had been restored.

Google News Business · 17d ago
OpenAI Acquires Presentation Startup NextSlide

OpenAI Acquires Presentation Startup NextSlide

NextSlide, a startup that turns prompts and notes into polished presentations, is joining OpenAI and folding into the ChatGPT team.

TechCrunch · 17d ago
Google drops APT numbers for country-coded hacker codenames

Google drops APT numbers for country-coded hacker codenames

Google overhauled its naming system for hacking groups, replacing numeric APT designations with memorable codenames that encode the attacker's country of origin. The change unifies Google's internal t

TechCrunch · 17d ago
🤖

Now it’s got real for AI as Demis Hassabis steps down at Google - The Times

Demis Hassabis, co-founder of Google DeepMind, is shifting his AI role at Google, prompting a broader overhaul of the company's AI leadership.

Google News Business · 17d ago
Watching Roku’s AI channel is like eating from a trough

Watching Roku’s AI channel is like eating from a trough

Roku debuted a 24/7 AI-generated channel from startup Fairground this week, and a reviewer's hour of viewing turned up disjointed, machine-clipped shorts that pale against the platform's traditional p

The Verge · 18d ago

Get the Tech newsletter

Curated tech stories, every morning. Free.

No spam. Unsubscribe anytime.