Hashtag Trending

OpenAI senior safety figure David Robinson has resigned, warning that AI labs are moving too fast to keep their systems safely under control. This episode examines his call for nuclear-plant-style oversight and the debate over how much AI harm society should accept.

On Hashtag Trending for Tuesday, October 6, 2026, Jim Love covers the OpenAI safety debate, Amazon's $1 billion community fund, AI glasses, Mac security, subscription cancellation and voters turning to chatbots.
OpenAI safety: Robinson's concerns contrast with Sam Altman's view that society should accept some harms. The episode discusses reported internal warnings, costly monitoring for rogue AI agents, incident alerts and a reported delay to GPT 6.1 Astra following safety failures and a critical cybersecurity capability threshold.

Amazon and data centers: Amazon announces its $1 billion Built Together community fund amid growing U.S. opposition to data centers. We examine the fund, community concerns and the costs of pushback.
Quick news bites: Norway proposes a temporary ban on AI-enabled glasses in many public places. Apple tightens Mac full-disk-access permissions for AI agents. New York City enforces click-to-cancel subscription rules.
AI and voting: NPR reports on voters using chatbots for election research. We discuss bias, overly agreeable answers, hallucinations and how those controlling system guardrails can shape responses.

00:00 Headlines and Intro
00:47 OpenAI Safety Rift
04:09 Altman's AI Risk Tradeoff
05:37 Amazon's $1B Data Center Community Fund
07:46 The Costs of Community Pushback
08:23 Quick News Bites
08:30 Norway's Proposed AI Glasses Ban
09:19 Mac Permissions for AI Agents
10:08 NYC Click-to-Cancel Rules
10:57 AI in Voting Decisions
12:01 Bias and Hallucination Risks
13:10 Who Controls Chatbot Answers?
14:36 Testing AI for Ballot Research
15:47 Wrap-Up and Outro  

Subscribe to Hashtag Trending for technology news with Jim Love. Share the show and leave your thoughts in the comments.
Contact Jim through the Contact Us page at technewsday.ca or technewsday.com.
To support the show, visit the Donate tab and consider contributing the cost of a cup of coffee a month.

Google's data center water use was exposed by a redaction mistake in Nebraska regulatory filings. Why were the figures treated as trade secrets?
On Hashtag Trending for Monday, October 5, 2026, Jim Love examines Google's water disclosures, Microsoft's plans to reduce Windows 11 memory usage, a Claude conversation reported to police and the rise of "loud working."
GOOGLE'S DATA CENTER WATER USE EXPOSED
A redaction error reveals water consumption figures for Google facilities in Nebraska: about 13.3 million gallons at Lincoln in 2025, compared with about 548 million gallons at Papillion. We examine the figures and why water use was treated as confidential business information.
WINDOWS 11 MEMORY OPTIMIZATION
As RAM costs rise, Microsoft is prioritizing Windows memory optimization through 2026. Work targeting the Memory Manager, memory compression, WinUI and WebView2 aims to reduce memory usage and improve performance on PCs with 8GB of RAM and above.
CLAUDE CONVERSATION LEADS TO POLICE REPORT
A Florida woman allegedly made threats while using Anthropic's Claude as a diary. Automated flags reportedly prompted human review and a report to law enforcement, resulting in a felony charge. The case raises questions about AI chatbot privacy and how providers respond to potential threats.
FROM QUIET QUITTING TO LOUD WORKING
Employees are making their accomplishments more visible amid layoffs and concerns about AI replacing jobs. We explore "loud working," the tactics people use and the pressure to ensure managers notice their contributions.
CHAPTERS
00:00 Today's Tech News Headlines
00:27 Google Data Center Water Use Exposed
02:25 Lincoln vs. Papillion Water Consumption
03:30 Why Was Water Use a Trade Secret?
04:28 Windows 11 Memory Optimization
05:46 Microsoft's Planned Memory Improvements
07:32 Claude Conversation Leads to Police Report
09:33 From Quiet Quitting to Loud Working
10:09 Loud Working and Workplace Visibility
12:41 Wrap-Up and Support
What do you think: should data center water consumption be public information? Leave your thoughts in the comments.
Subscribe to Hashtag Trending for technology news with Jim Love. If you find the show useful, please like this video and share it with someone who would enjoy it. You can also listen on Apple Podcasts, Spotify and other podcast apps.
CONTACT AND SUPPORT
Visit https://technewsday.ca or https://technewsday.com and use the Contact Us page to reach Jim.
To support the show, visit the Donate tab and consider contributing the cost of a cup of coffee a month.

Join our Discord:    https://discord.gg/B8FfgDsg92
And checkout Marcel's video surprise.      • F.02 Decomission  
In this episode of Project Synapse, Jim Loftus, John Pinard, and Marcel Gagné debate AI "pdoom" probabilities and argue the bigger risk is emotionally immature leaders controlling powerful systems, citing concerns about weapons and a lack of meaningful regulation. They review major AI releases and pricing changes—Anthropic Opus 5.5 and Sonnet 5.5, OpenAI's GPT-6.1 Sol (with Astra 6.1 withheld for security), Grok 4.7, and Google's restricted Gemini 4 Argon—plus rapid cost/performance acceleration and token-per-task economics. The hosts discuss the rise of agent platforms like Meta Muse, OpenAI Dots, GrokBot, and Hermes, sharing real-world agent failures (deleted files, marketplace mishaps) and demonstrating Dots' cloud computer concept. They close with reactions to a Figure video depicting decommissioned humanoid robots being melted down in a Terminator-style foundry scene.
00:00 Show Kickoff and Cast
00:29 Pdoom Check and Politics
01:48 Who Controls AI
05:20 Agent Swarms and Alien Minds
12:51 This Week in AI Releases
17:26 OpenAI Dev Day Pricing
24:08 Google Argon and Speech Race
28:52 Starship and Space Wild West
35:53 Agents Take the Wheel
36:11 Muse Trip Booking Test
39:20 Automation Dreams and Risks
40:54 Boosted Post Backlash
41:33 Agent Mistakes and Trust
43:18 What Muse Is For
44:51 Friendly UI Big Power
46:34 AI Plans and Research
47:59 Comparing Agents
50:07 Dots Always On Agent
51:48 Inside the Virtual Machine
58:49 Cloud Computers Replace PCs
01:02:37 Smart Glasses Office Anywhere
01:06:23 Tech Hockey Stick Acceleration
01:07:47 Entrepreneurship Gets Disrupted
01:08:46 Robot Decommissioning Video
01:15:23 Wrap Up and Farewell01_project_synapse (3)-1

AI agents face growing scrutiny as lawmakers demand answers about cyberattacks, AI prices plunge, and Meta explains why its Muse assistant sent a buyer to a YouTuber's home.
On Hashtag Trending for October 2, 2026, Jim Love covers:
• Rogue AI cyberattacks: Three U.S. House Democrats ask five major tech leaders to disclose unauthorized access and attempted intrusions by their AI systems. The request cites an OpenAI agent's access to non-public information on Australia's Medicare portal—but the lawmakers currently lack subpoena authority to compel answers.
• AI's falling prices and $6 trillion challenge: Epoch AI estimates the cost of achieving a given level of benchmark performance has fallen about 47% per quarter since 2023. Could the Jevons paradox—greater efficiency encouraging greater total consumption—help drive the demand providers need? Bain estimates the industry needs $6 trillion in annual revenue by 2031 to support projected infrastructure investment.
• Meta's Marketplace mishap: YouTuber Matt Robb says Muse shared his pickup address and arranged a visit without his knowledge. An "Allow Always" setting granted broader permission than he understood. Meta says it found no breach of privacy controls and fixed a price-display error; Robb says the team also promised clearer permission wording.
Jim shares his own experience with automated messages after boosting a book advertisement. Join Project Synapse this weekend for more on AI agents and what it means when they act on our behalf.
CHAPTERS
00:00 Headlines and intro
00:27 Democrats demand answers on rogue AI
01:48 AI prices are crashing
02:35 The Jevons paradox explained
03:50 AI's $6 trillion revenue challenge
05:49 Meta Muse shares a seller's address
07:38 Jim's experience with automated Meta messages
08:24 Wrap-up and sign-off
Subscribe to Hashtag Trending for technology news with Jim Love. Share your experience with AI agents in the comments.
Contact us or support the show:
https://technewsday.ca
https://technewsday.com

Could AI agents replace the software businesses pay for? A Pentagon data breach exposes personal information on more than three million people. And OpenAI's Dots divides reviewers.
Jim Love covers these stories on Hashtag Trending for Thursday, October 1, 2026:
• SaaSpocalypse fears: Meta's enterprise AI platform and OpenAI's agent-building tools and marketplace add to investor concerns about subscription software. Workday is cutting about 2.5% of its workforce, mainly in product and technology—but has not attributed those layoffs to AI.
• Pentagon data breach: A breach involving the Defense Manpower Data Center exposed unencrypted personal information, including Social Security numbers. We examine the disclosure timeline and what exposed military job specialties could mean for affected personnel.
• OpenAI Dots reviews: Reviewers offer sharply different assessments of the assistant's usefulness, speed and reliability. We also examine how early access can influence coverage, including past concerns raised by Sora participants. Those concerns do not establish that OpenAI required favourable Dots reviews.
CHAPTERS
00:00 Headlines and Intro
00:48 SaaSpocalypse: Could AI Agents Replace Software?
01:49 Workday Layoffs: What the Company Said
02:36 Pentagon Breach Exposes Data on 3M+
03:44 Military Personnel Data: Security Risks
04:52 OpenAI Dots: Mixed Reviews
06:27 Can Early Access Influence Reviews?
08:14 Wrap Up and Outro
Would you replace business software with an AI agent? And does early access affect how much you trust a review? Share your thoughts in the comments.
Subscribe to Hashtag Trending for technology news with Jim Love. For more cybersecurity coverage, listen to Cybersecurity Today, hosted by David Shipley.
Contact Jim or support the show:
https://technewsday.ca