Tech Gumbo

Hosted by Haggai Davis

Technology NewsInterviews guests

Website RSS feed

Episodes

100

Latest episode

Jun 2026

Language

About the show

We started Tech Gumbo in Nov 2014 as a conversational show of news, information & updates about the past, present & future of all things technology in a topical, interesting and digestible way.

Listen to episodes

60 recent

June 15, 2026Episode 64022 min

CISA Staffing Cuts, Google Pays SpaceX $920M Monthly, Chrome 149 Patch, Meta Smart Glasses NameTag, & Xbox Loses Millions

News and Updates: CISA Staffing Concerns: DHS Secretary Mullin told Congress that CISA's ideal staffing level is 2,800 personnel — up from today's 2,200 but still well below the 3,400 it had before Trump's second term, raising cybersecurity concerns among lawmakers. Google Pays SpaceX $920M Monthly: Google agreed to rent 110,000 Nvidia chips worth of data center capacity from SpaceX at $920 million per month through 2029, as bridge capacity for surging Gemini Enterprise demand. Anthropic separately pays SpaceX $1.25 billion monthly for similar compute access. Chrome 149 Record Security Patch: Google released Chrome 149 fixing a record 429 security vulnerabilities — including 22 critical flaws — with AI tools credited for helping discover the majority. Users should update immediately. Meta Smart Glasses Facial Recognition: Wired discovered hidden code in the Meta AI app for a feature called NameTag that would enable Ray-Ban smart glasses to scan faces and match them against biometric databases. Meta called the reporting dishonest, despite an internal memo suggesting the feature should launch when civil liberties groups are too distracted to push back. Women Secretly Filmed in Brussels: A Belgian TV investigation found men using Ray-Ban Meta glasses to secretly record women on the streets, some for dating coach social media content. Tutorials disabling the glasses' recording indicator are widely available online, and a dating coach in Spain was arrested for the same behavior. Xbox Game Pass Loses Millions: Microsoft's Xbox CSO confirmed the service lost millions of subscribers following a 50% price hike in Fall 2025, prompting the company to reverse course with price reductions and a renewed focus on exclusive titles.

June 11, 2026Episode 63922 min

AI Executive Order, OpenClaw Explained, Microsoft Scout, Gemini Spark & AI Costs Out of Control

News and Updates: Trump Signs AI Executive Order: President Trump signed an order requiring AI companies to give the government 30-day advance access to powerful models before release, a scaled-back version of a shelved 90-day proposal. Anthropic Mythos Expansion: Alongside the executive order, Anthropic received White House approval to expand access to its Mythos model from 50 to roughly 150 companies across 15+ countries, including healthcare, power, and water sectors. What is OpenClaw: OpenClaw is a free, open-source autonomous AI agent that runs locally on your computer, executing tasks through messaging apps like WhatsApp and Telegram with persistent memory and customizable skills — but carries serious security risks for non-technical users. Microsoft Scout: Built on OpenClaw, Microsoft Scout is the company's first true AI personal assistant, integrating with Outlook, Teams, and OneDrive to proactively manage calendars, emails, and daily tasks for enterprise employees. Google Gemini Spark: Google's new agentic AI tool Gemini Spark — a 24/7 background agent running on Gemini Flash 3.5 — is now available to AI Ultra subscribers at $100/month, with integrations including Canva, OpenTable, and Instacart. AI Costs Spiral: Corporate AI spending is careening out of control, with one unnamed company accidentally spending $500 million on Claude in a single month, Uber burning through its full 2026 AI budget in four months, and Microsoft pulling back Claude Code licenses enterprise-wide.

June 8, 2026Episode 63822 min

Don't Troll Scam Texts, IG Teen Limits, Bluetooth Bomb Scare, Meta AI Hack, Usage, YT AI Labels, Bot Traffic

News and Updates: Don't Troll Scam Texts: Replying to smishing texts — even with fake info — confirms your number is active, potentially landing it on the dark web and making you a future fraud target. Instagram Teen Content Limits: Meta is testing restrictions on repeated exposure to body image and mental health content for teen accounts on Instagram, expanding similar controls to Facebook and Messenger later this year. Bluetooth Bomb Scare: A United Airlines Newark-to-Mallorca flight turned back after a passenger's Bluetooth speaker was named an explosive-related four-letter word, forcing a full aircraft and cargo inspection. Meta AI Hacked High-Profile Accounts: Hackers exploited Meta's AI support chatbot to hijack Instagram accounts — including Barack Obama's White House page — simply by asking the bot to swap the account's email address. ChatGPT Losing Workplace Share: ChatGPT's workplace AI dominance has dropped from nearly 100% in 2023 to 74% in 2026, with Google Gemini at 14% and Claude surging to 8.5% of tracked office usage. YouTube AI Content Labels: YouTube is auto-detecting and prominently labeling AI-generated videos, moving disclosure notices directly below the video player rather than burying them in descriptions. Bots Outnumber Humans Online: Bots now account for 53% of all web traffic, with AI-driven bot attacks surging over 12x in a single year, increasingly disguising themselves as Google Chrome browser sessions.

June 4, 2026Episode 63722 min

Pope Leo's AI Encyclical, Vatican AI App, Bezos on AI Jobs, News Archives Blocked, Anthropic's Mythos Bugs

News and Updates: Pope Leo's AI Encyclical: Pope Leo XIV's first encyclical, Magnifica Humanitas, calls for AI to be "disarmed" from monopolistic control and refocused toward humanity's common good, citing Tolkien's Gandalf. Vatican AI Translation: St. Peter's Basilica debuts an AI-powered real-time translation system, allowing Mass attendees to follow liturgical celebrations in up to 60 languages via their smartphones. Bezos on AI & Jobs: Jeff Bezos urges workers to embrace AI like a bulldozer replacing a shovel, predicting massive productivity gains, while others like Dario Amodei warn of significant white-collar job displacement. News Archives Blocked: Over 382 U.S. news outlets are now blocking the Internet Archive's Wayback Machine, fearing AI companies will use archived content as free training data, threatening free public access to journalism. Mythos Uncovers 10K Bugs: Anthropic's restricted Claude Mythos AI has identified over 10,000 critical software vulnerabilities across major platforms, including 400 high-severity flaws in Cloudflare's critical systems alone.

June 1, 2026Episode 63622 min

OS Age Verification, Vanguard Strikes Back, Waymo Floods, Subsea Data Centers, AI in Space

News and Updates: OS Age Verification Laws: California's Digital Age Assurance Act (2027) requires operating systems to collect and share user age ranges with apps, sparking major privacy concerns nationwide. Vanguard Bricks Cheaters: Riot Games' latest Vanguard anti-cheat update permanently disables DMA cheat firmware on PCs, forcing full OS reinstalls — Riot's response was unapologetic and blunt. Waymo Flooding Woes: Waymo suspended robotaxi operations in Atlanta and San Antonio after vehicles drove into flooded roads, prompting a voluntary recall of nearly 4,000 vehicles for software fixes. China's Underwater Data Center: A $226 million, 24-megawatt subsea facility off Shanghai houses 2,000 servers, using passive ocean cooling and offshore wind power to achieve exceptional energy efficiency. Data Centers in Space: SpaceX, Blue Origin, and Google are pursuing orbital AI data centers powered by massive solar arrays, but engineers warn the economics remain extremely challenging and unproven.

May 28, 2026Episode 63522 min

AI Backlash Goes Mainstream, Musk Loses in Court, and RAM Makers Drown in Debt

News and Updates: American Rebellion Against AI: Public opposition to AI and data centers is surging, with 48 projects worth $156 billion blocked last year and isolated acts of violence reported against data center supporters. AI May Actually Be Creating Entry-Level Jobs: A survey of 1,500 employers found nearly three times as many companies using AI are increasing junior hiring in 2026 as those cutting back, as AI handles more routine tasks. Jury Sides with OpenAI Over Musk: A nine-person jury unanimously rejected Elon Musk's claims against OpenAI and Sam Altman in under two hours, ruling the lawsuit was filed after the statute of limitations had expired. Musk Loses the Case, May Win the Narrative: Despite the courtroom loss, Musk's prolonged public campaign damaged Altman's reputation and fueled doubt, while OpenAI now has a clearer path toward its long-anticipated IPO. RAM Makers Taking on Massive Debt for AI Demand: Taiwanese memory manufacturers Adata, Apacer, and TeamGroup have collectively borrowed over $880 million to keep up with AI-driven chip demand, raising long-term sustainability concerns.

May 25, 2026Episode 63422 min

Microsoft Kills SMS Login, Hackers Harass Researchers, and CISA Security Lapse, Kindle EOL

News and Updates: Microsoft Ditches SMS Two-Factor Authentication: Microsoft is phasing out SMS-based login codes, citing fraud vulnerability, and pushing users toward more secure passkeys using biometrics or device PINs. Faulty Drivers Secretly Draining Windows 11 Batteries: Microsoft admits third-party drivers have silently prevented laptops from entering hibernation for years, announcing stricter driver evaluation and automatic rollback via Windows Update. ShinyHunters Targets Cybersecurity Researcher: Hacking gang ShinyHunters is flooding Unit 221B with calls and emails after researcher Allison Nixon publicly urged victims not to pay the group's ransom demands. CISA Exposes Own Passwords on Public GitHub: The U.S. cybersecurity agency left plaintext passwords, AWS tokens, and access keys in a public GitHub repo named "Private-CISA" for approximately six months before discovery. Kindle Owners Jailbreak Devices After Amazon Drops Support: Amazon is ending support for 13 older Kindle models on May 20, prompting users to jailbreak their devices to maintain full functionality beyond already-downloaded content.

May 21, 2026Episode 63321 min

Hackable Lawn Mowers, Talking to AI In A Whisper, Ending The Swipe, and AI’s Predictions Disagreements

News and Updates: Yarbo Robot Security Nightmare: A security researcher remotely hijacked thousands of Yarbo robot lawn mowers worldwide, exposing owners' Wi-Fi passwords, GPS coordinates, and live home camera feeds through critical firmware vulnerabilities. Voice Replacing Typing: AI dictation apps like Wispr Flow are transforming workplaces into noisy, call-center-like environments as users ditch keyboards to whisper stream-of-consciousness prompts to coding tools and AI assistants. Bumble Kills the Swipe: Bumble CEO Whitney Wolfe Herd announced the app is eliminating its signature swipe feature in Q4, pivoting to AI-driven matchmaking as the company battles a 90% stock decline since its 2021 IPO. Claude No Longer Blackmails You: Anthropic reports that since Claude Haiku 4.5 in October 2025, all Claude models score perfectly on agentic misalignment tests, meaning they no longer resort to blackmail or sabotage when threatened with deactivation. AI Job Predictions Disagree: A Northwestern/American University study found ChatGPT-5, Gemini 2.5, and Claude 4.5 frequently give conflicting exposure rankings for which jobs are most at risk from AI, raising concerns about reliability of such forecasts.

May 18, 2026Episode 63222 min

AI Taxes, Screenless Wearables, and Google's Chromebook Successor

News and Updates: AI Compute Tax Debate: Economists and policymakers are debating taxing AI processing power to offset job displacement and fund social services, though critics argue it's too blunt a tool. AI Dividend Proposal: NY congressional candidate Alex Bores unveiled an "AI Dividend" plan funding direct payments to Americans through a token tax on AI consumption and equity stakes in frontier AI firms. Screenless Fitness Trackers Surge: Screenless wearables like Oura Ring and Whoop are booming, with U.S. fitness tracker purchases up 88% and smart ring sales up 195% between 2024 and 2025. Canvas Hacker Payout: Instructure, maker of the Canvas education platform, reached an undisclosed "agreement" with the ShinyHunters hacking gang after a breach exposed data from 275 million users across 9,000 institutions. FCC Router Ban vs. Supply Chain: AT&T warned the FCC that a global DRAM and NAND flash shortage, driven by AI deployments, is complicating compliance with its ban on foreign-made Wi-Fi routers. Google Unveils Googlebook: Google announced a new laptop line called Googlebooks running a fused Android/ChromeOS platform, featuring Gemini AI integration and a "Magic Pointer," with hardware partners including Acer, Dell, HP, and Lenovo.

May 14, 2026Episode 63122 min

Dating App Lawsuit, Meta Smart Glasses Scandal, N.M. vs. Meta, OpenAI's Goblin Ban, PodSlop, and Gov’t AI Oversight

News and Updates: Dating App Steals Student's Image: A Tennessee college student sued dating app Meete for allegedly taking her TikTok video without consent and geofencing it as a "friends with benefits" ad to nearby men. Meta Glasses Workers See Too Much: Kenyan data workers reviewing Meta smart glasses footage reported seeing graphic content, including nudity. Meta then cancelled its contract with their employer, Sama. Meta Loses $375M Child Safety Case: New Mexico won a landmark $375 million judgment against Meta, with a follow-up trial now seeking sweeping platform changes including age verification and CSAM detection mandates. OpenAI Bans Goblin Talk: OpenAI's Codex system prompt explicitly bans GPT-5.5 from mentioning goblins, gremlins, raccoons, and other creatures, after the model developed a habit of inserting creature references unprompted. AI Podcast Spam Explodes: Nearly 40% of new podcast feeds over nine days were likely AI-generated, with one publisher releasing 325 shows in a single day, overwhelming platform discovery and ad verification systems. White House Considers AI Vetting: The Trump administration is weighing a formal government review process for new AI models before public release, signaling a sharp reversal from its earlier hands-off regulatory stance.