Refresh Computers Tech Talk

05-17-26 Social Media Is Engineered To Be Addictive To You and Your Child

Send us Fan Mail [https://www.buzzsprout.com/2165515/fan_mail/new] Record profits used to mean stability. Now it can mean something else: a company realizes AI can do more with fewer people, and the layoffs start anyway. We talk through the real-world signals, starting with Cisco cutting thousands of jobs while reporting strong results, plus other examples like GM trimming IT roles. The takeaway isn’t panic, it’s clarity: AI job displacement is already happening, and the smartest move is to understand where the tech is headed before it reshapes your department around you. We also get practical for anyone shopping this summer. New computer prices are still elevated, so we break down the specs that actually matter for a fast Windows 11 experience in 2026: 16GB RAM as the sweet spot, a solid state drive as non-negotiable, and CPUs new enough to keep up with modern browsers and AI features. We explain why bloatware makes many brand-new PCs feel slower than they should, and why a properly refurbished laptop or desktop can be a better value without the junk. Then we shift to the redesign of social media itself. Endless scrolling, reels, shorts, and notification loops are engineered to be addictive, and we talk about what that does to attention, sleep, and mental health, especially for kids and teens. The European Union is pushing regulations that target the mechanics of addiction, not just content, and it may set the template for what comes next. We close with a few wild tech stories and a serious security lesson every business needs: revoke access before you terminate someone. If this helped, subscribe, share the show with a friend, and leave us a review so more people can find Tech Talk. Support the show [https://www.buzzsprout.com/2165515/support] Follow us on Facebook at https://www.facebook.com/RefreshComputers/ Track us on X at https://x.com/RefreshStores

05-16-26 You Are More Vulnerable Online Than Ever So Update Everything

Send us Fan Mail [https://www.buzzsprout.com/2165515/fan_mail/new] AI is no longer just a productivity tool, it’s a force multiplier for cybercrime. We dig into Google’s Threat Intelligence Group warning about the first confirmed case of criminal hackers using AI to discover and weaponize a zero-day vulnerability, the kind of software flaw nobody can patch because nobody knows it exists yet. That changes the clock on security from “weeks to respond” to “minutes to get hit,” and it’s why we keep hammering the basics that still work.  We walk through the simplest, highest-impact defenses for real people: turn on automatic software updates, patch your router, and stop treating smart home gadgets like toys. Every connected device is an endpoint, and attackers only need one weak link. We also talk about practical account protection with passkeys and why you should treat unexpected emails, texts, and scam calls as hostile until proven otherwise. If you want help checking your settings, we point you to the free tech support hotline and what to ask.  Then we hit a privacy shift you may have missed: Apple finally makes iPhone to Android texting secure with RCS and end-to-end encryption via iOS 26.5, closing a long-standing gap where SMS and MMS were basically digital postcards. From there we zoom out to the bigger AI landscape, including Google’s Gemini push across Android and the messy business side of AI with the Elon Musk vs Sam Altman lawsuit over OpenAI’s original mission, plus why Grok may be losing momentum as competitors gain fans.  If you get value from the show, subscribe, share it with a friend who keeps postponing updates, and leave us a review so more people can find Tech Talk. What security habit are you changing this week? Support the show [https://www.buzzsprout.com/2165515/support] Follow us on Facebook at https://www.facebook.com/RefreshComputers/ Track us on X at https://x.com/RefreshStores

05-10-26 Who Should Decide What AI Can Do?

Send us Fan Mail [https://www.buzzsprout.com/2165515/fan_mail/new] The government wants a seat at the table before the most powerful AI ships, and Big Tech is starting to say yes. We walk through the news that Google, Microsoft, and xAI plan to let the US Department of Commerce review new frontier AI models ahead of release, supposedly to catch national security and cybersecurity risks. That sparks the bigger question we cannot ignore: where does “safety review” end and information control begin, and how might it change depending on who is in charge? Then we bring it down to a practical decision you can make today: what computer should you actually buy in an AI world? We make the case that you do not need an expensive, brand-new “AI machine” to use tools like ChatGPT, Grok, or Gemini. A well-built, properly remanufactured refurbished Windows 11 PC can deliver a great experience at a fraction of the cost, especially when it comes with real testing, real support, and a real warranty. From there, we dig into the next wave: AI agents. Google’s Remy concept inside the Gemini app points toward a 24/7 assistant that can take actions like emailing, booking, ordering, and managing your calendar. That convenience is real, but so are the risks when an agent needs access to your accounts. We close with the money trail behind all this: Samsung hitting a trillion-dollar market cap, the AI chip boom driving demand for memory, and data centers scaling so fast that the costs eventually land on consumers. Subscribe, share the show with a friend who loves tech, and leave a review with your take: should frontier AI be reviewed before release or pushed forward at full speed? Support the show [https://www.buzzsprout.com/2165515/support] Follow us on Facebook at https://www.facebook.com/RefreshComputers/ Track us on X at https://x.com/RefreshStores

05-09-26 Google Installed A Hidden 4GB AI Model In Your Browser Without Asking

Send us Fan Mail [https://www.buzzsprout.com/2165515/fan_mail/new] A hidden 4GB file on your computer sounds like malware, but this time it can arrive through a normal Chrome update. We break down what Google’s Gemini Nano is, why the weights.bin file is showing up for Chrome users and many Chromebook owners, and why people are upset that it installs without a clear “yes” from you. We also talk through what “on-device AI” actually means, what it’s supposed to help with, and the real-world problem: even if you toggle it off, the storage hit can still be there. From there, we zoom out to performance and value. New PCs often ship with bloatware that burns storage, nags you with pop-ups, and quietly slows the machine down. We explain why a quality refurbished computer with a clean Windows install can feel faster on day one, plus simple ways to audit the tech in your home or small business. We also share how our recycling and trade-in options can help you clear out old gear responsibly and put credit toward an upgrade. Then we get into AI you actually choose to use. ChatGPT’s latest “5.5 Instant” update claims 52.5% fewer hallucinations, better memory, and clickable sources, but we keep it practical: AI is a research assistant, not an oracle, and it still needs fact-checking for medical, legal, and financial questions. Finally, we cover the Canvas ransomware breach tied to Shiny Hunters and why this is prime time for phishing scams aimed at students, parents, and educators, including the password changes you should make right now. Subscribe, share the show with a friend who uses Chrome or Canvas, and leave a review so more people can find Tech Talk. Support the show [https://www.buzzsprout.com/2165515/support] Follow us on Facebook at https://www.facebook.com/RefreshComputers/ Track us on X at https://x.com/RefreshStores

05-02-26 An AI Model Finds Thousands Of Hidden Zero-Day Flaws Before Criminals Do

Send us Fan Mail [https://www.buzzsprout.com/2165515/fan_mail/new] A brand-new AI model goes hunting for hackers and what it finds is honestly unsettling: thousands of high-severity zero-day vulnerabilities, including flaws that sat unnoticed for decades. We talk through Anthropic’s Project Glasswing, why the partnerships matter, and how AI-powered cybersecurity could help patch the holes before bad actors get there first. If you’ve ever wondered how much unseen risk is baked into the software running your bank, your hospital, or your everyday devices, this conversation puts real examples to that fear. Then the mood shifts from “future tech” to “right now.” The Shiny Hunters ransomware and data extortion group is back in the headlines with claims involving major brands like Carnival, Amtrak, ADT, Zara, and more. We break down what these attacks look like, why cloud data and third-party platforms keep becoming the weak link, and how phishing emails still beat expensive security programs. Most importantly, we share concrete steps you can take if you think your information was exposed, including checking HaveIBeenPwned, changing passwords the right way, and when it makes sense to freeze credit. We also get into a surprisingly practical change coming from the European Union right to repair movement: by February 2027, smartphones sold in the EU must have user-replaceable batteries. We debate what that could mean for iPhone and Android design, durability, warranties, and e-waste recycling and why tech rules made in Europe often ripple into the United States. If this helped you, subscribe, share it with a friend who needs a security reset, and leave a quick review with the one topic you want us to tackle next. Support the show [https://www.buzzsprout.com/2165515/support] Follow us on Facebook at https://www.facebook.com/RefreshComputers/ Track us on X at https://x.com/RefreshStores