Hashtag Trending

Apple Caps Employee AI Spend, Google Weighs Space Data Centers, and Big Tech Sues California
David Shipley guest hosts Hashtag Trending for Nov. 9, 2026, covering major tech headlines: Apple is reportedly cutting daily AI spending limits for some employees (from $300 to $200, with possible lower caps) as other companies also rein in AI budgets, emphasizing cheaper models and on-device AI.
Google is testing AI computing in orbit with a small TPU-equipped satellite but says space data centers could cost up to 100x more than Earth-based facilities, with major hurdles like heat, radiation, orbital crowding, and launch constraints.
Police still have made no arrests in an $11.4 million NVIDIA hardware highway theft as similar heists surge.
OpenAI sought $30B but has raised $11B so far amid heavy losses and massive compute spending.
Google, Meta, and TikTok are suing California to block AB 1709, a law limiting addictive social media features for kids under 16.
00:00 Today's Tech Headlines
00:31 Meet the Host
00:47 Apple Caps AI Spending
01:48 How Token Billing Works
02:29 AI Budget Cuts Spread
03:45 Is the AI Party Over
04:11 Data Centers in Space
04:30 Google's Orbital TPU Test
05:07 Why Space Data Centers Fail
07:01 Nvidia Highway Heist Update
08:14 Why AI Hardware Theft Pays
09:12 OpenAI Funding Falls Short
09:53 OpenAI Burn Rate Reality
11:26 Big Tech Sues California
13:21 Host Take on Addictive Apps
14:23 Wrap Up and Where to Listen

Project Synapse: David Colombo on Tesla Hacking, AI-Driven Zero-Days, and the Next Two Years of Cybersecurity
Project Synapse opens with a humorous AI-generated song about tech CEOs and introduces guest David Colombo, who at 19 disclosed TeslaMate flaws that enabled remote control of Teslas in 13 countries, later advising organizations like GSK and KnowBe4 and founding Alaris Security. Colombo recounts growing up in rural Germany, learning to code, finding early vulnerabilities, and explains why TeslaMate and ecosystem tokens were attractive targets, noting weak scoping and suspicious-request detection. The discussion shifts to AI's rapid advancement in offensive security, democratizing zero-day discovery and shrinking patch windows, with debate on unauthorized "patching" versus change-management realities and the need for safer, faster patching methods. Colombo predicts the next two years will be stressful as defenders catch up, calling for greater support for security efforts. Hosts also discuss model scaling, synthetic data, emergent self-correction, misinformation risks, "compute campuses," and evolving AI pricing and image tools.
00:00 Cold Open AI Song
01:43 Show Intro Guest Setup
03:13 David Colombo Origin Story
06:45 TeslaMate Hack Breakdown
10:56 AI Supercharges Attacks
14:24 Should We Auto Patch
19:39 Why AI Deployments Insecure
21:14 Two Year Security Forecast
24:50 Post Interview Security Talk
28:20 New Models Data Rumors
30:04 Synthetic Data Model Scaling
31:52 Regulation Conspiracy Chat
33:31 Build Infrastructure First
35:00 Astra Finds Model Errors
36:13 Recursive Self Improvement Debate
40:34 Bad Data and Election Misinformation
43:09 Sycophancy and Prompting Risks
49:34 Tech CEO Satire Video
52:31 Compute Campuses Rebrand Jokes
55:08 ChatGPT Pricing Tiers Explained
57:47 Image Tools and AI Slop
01:00:54 AI Economics and Public Goods
01:06:28 Companion Bots and Sexbot Buzz
01:08:56 Wrap Up and Where to Comment

Jim Love's Hashtag Trending for Friday, November 6, 2026 covers volatile tech markets as investors interpret Democratic election wins amid concerns over AI spending, data center restrictions, and looming congressional scrutiny, including attention on Elon Musk, while noting financial pressures such as Oracle's near-junk credit status and upcoming fundraising moves by Anthropic and OpenAI. The episode also reports backlash against Google's school account changes that force schools to accept AI features in Search or disable Search entirely, drawing criticism from NYC Mayor Zoran Mandami and Lina Khan. Reuters findings are cited on suspected surveillance of Starlink users via targeted ads linked to DiscoverTech, alongside Citizen Lab research showing spyware delivery through "adfection." Finally, studies on AI companion apps describe benefits and distress when updates or shutdowns disrupt attachments, recommending user consultation and clear reminders that AI lacks real emotions.
00:00 Headlines and Intro
00:29 Tech Stocks After Election
01:33 Congress Scrutiny and Musk
03:08 AI Bubble and IPO Watch
05:03 Google AI in Schools
07:03 Ads Used for Surveillance
09:04 AI Companions and Attachment
10:49 Weekend Show and Wrap Up

Link to Digital Governance Council  https://dgc-cgn.org/
On Thursday, November 5, the show covers US election-night results affecting data center development, with communities passing dozens of measures to pause, restrict, require resident approval for, or outright ban projects, driven by concerns over electricity bills, water use, noise, and local control. Related AI and infrastructure stocks fell, including Oracle, CoreWeave, and NVIDIA, highlighting that community acceptance is a key constraint alongside chips, financing, and power. Despite a broader market rally, Tesla and SpaceX declined, and Trump Media fell about a quarter. The episode also discusses OccupySaaS, a laid-off-engineer movement using AI agents plus human review to build open source alternatives to SaaS giants, starting with Auth0 and Salesforce, though both projects remain early. Finally, Canada's Digital Governance Council releases a resolution urging stronger sovereign control over critical data and computing, Canadian-controlled digital payments infrastructure with democratic oversight, and greater influence over digital-economy rules and standards.
00:00 Headlines and Setup
00:32 Voters Reject Data Centers
01:56 AI Stocks Feel the Heat
02:22 Tesla SpaceX Trump Media Slide
03:08 OccupySaaS Open Source Rebuild
04:09 Reality Check on DIY SaaS
04:40 Canada Pushes Digital Sovereignty
05:34 Three Priorities for Infrastructure
06:33 Wrap Up and How to Support

AI Chatbots Botch Midterm Ballots, Intel's Chip Slump Deepens, Cisco Bets on Unlimited AI Refactor
On hashtag trending for Wednesday, Nov. 4, 2026, Jim Love covers testing by The Washington Post showing AI chatbots generated inaccurate voter guides ahead of the US midterms, including inventing races, misidentifying candidates, and frequently failing to correctly identify congressional districts, with some errors traced to AI-generated voter-info websites; OpenAI updated ChatGPT's Democracy Works integration after the tests and Amazon paused Alexa's voter guide. The episode also reports steep chip-stock declines led by Intel, with Yahoo estimating chip and memory stocks lost about $3.7 trillion since Oct. 6, and notes AMD's upcoming earnings as a signal for industry pressure. Cisco is reportedly giving a two-person team unlimited AI tokens to refactor 98% of legacy code in two years. Wells Fargo will require customers to revoke AI agent access and keep proof. New Brunswick introduces legislation targeting unfair digital practices affecting consumers under 16.
00:00 Headlines Rundown
00:43 Chatbots Botch Ballots
02:51 Chip Stocks Slide
04:37 Cisco Unlimited Tokens
06:33 Wells Fargo Agent Proof
07:21 Protecting Kids Online
09:19 Wrap Up and Support