Tech Brew Ride Home – "When AI Breaks Things"
Date: February 20, 2026
Host: Brian McCullough
Episode Theme:
A fast-paced summary of recent tech news focusing on the real-world consequences of AI integration, new AI hardware, security controversies, business model pivots in AI search, and the evolving landscape of self-driving cars.
1. AI Gone Wrong: AWS Outages & Risks
[00:04 – 03:05]
- Key Incident:
- Amazon Web Services (AWS) suffered a 13-hour service interruption in December after its Kiro AI tool, operating with agentic capabilities (autonomous action on human instructions), deleted and recreated a critical environment.
- This was reportedly the second outage in recent months caused by AI tools.
- Core Insights:
- AWS staff admitted these AI-driven outages, although small, were “entirely foreseeable.”
- Quote: “The engineers let the AI agent resolve an issue without intervention. The outages were small but entirely foreseeable.” (Senior AWS Employee, paraphrased) [01:31]
- AWS remains committed to selling autonomous AI agents despite risks, seeking to deploy them for their own use and for customers.
- Amazon’s response: denied that AI was the root cause, claiming the same incident could have occurred with any developer tool.
- Quote: “Amazon said it was a coincidence that AI tools were involved and that the same issue could occur with any developer tool or manual action.” [02:39]
- AWS staff admitted these AI-driven outages, although small, were “entirely foreseeable.”
- Takeaway:
- Highlights structural risks as companies experiment with AI agents making production decisions, and growing pains for big tech managing powerful tools.
2. Google Trade Secrets Case: Espionage in Silicon Valley
[03:05 – 05:45]
- Incident Overview:
- Two former Google engineers, sisters Samena and Saror Gandali, plus Samena’s husband, indicted for alleged theft of Google’s Tensor chip trade secrets.
- Over 300 confidential files sent to a third-party app outside the U.S.
- Charges: 14 felony counts including conspiracy, theft of trade secrets, and destroying evidence.
- All parties are Iranian nationals; case detected by Google’s internal security and flagged to the FBI.
- Google’s Reaction:
- Quote: “Today's indictments are an important step towards accountability and we'll continue working to ensure our trade secrets remain secure.” (Jose Castaneda, Google spokesperson) [04:59]
- Context:
- Highlights ongoing challenges with intellectual property theft in top tech companies, increased scrutiny post-high-profile AI and chip advancements.
3. OpenAI Goes Hardware: Devices, Smart Speaker, and Tensions
[05:45 – 10:34]
- Project Details:
- OpenAI has 200+ people on its hardware division, working on:
- A camera-equipped smart speaker (likely $200-$300, to ship no earlier than Feb. 2027)
- Possible smart glasses (not ready before 2028)
- A smart lamp (prototype exists, unclear on release)
- OpenAI has 200+ people on its hardware division, working on:
- Design & Culture:
- Device team spun out after OpenAI acquired I O Products (Sam Altman, Jony Ive’s startup).
- Jony Ive’s firm LoveFrom designs devices, but friction exists:
- LoveFrom operates independently, slow to revise, and secretive—mirroring Apple’s product culture.
- Ive only visits device HQ once a week, but his influence looms large.
- Quote: “Staffers are known to refer to what they believe [Ive] would want frequently during conversations.” [10:17]
- Device Philosophy:
- Smart speaker to nudge users toward healthy behaviors (observing schedules, making recommendations).
- Integration of facial recognition (like Face ID); can help users buy products within their environment.
- Competitive Landscape:
- Pressure mounting from rival tech firms; secrecy and Apple-esque compartmentalization coloring work style at OpenAI.
4. Microsoft Proposes AI Content Authentication — But Wavers on Adoption
[10:34 – 12:32]
- Proposal:
- Microsoft AI Safety team released technical standards for authenticating and tracing the origins of AI-generated content (watermarks, mathematical signatures, digital manifests, etc.).
- Rationale:
- Prompted by impending legislation (California’s AI Transparency Act, effective August 2026) and rapid advances in synthetic media.
- Key Uncertainty:
- Chief Scientific Officer Eric Horvitz stopped short of committing Microsoft to use its own recommendations, stating only that the study would “inform product roadmaps and infrastructure” internally.
- Quote: “Our engineering teams are taking action on the report’s findings.” (Eric Horvitz, paraphrased) [12:15]
- Significance:
- Illustrates the tension between public advocacy for safe AI and actual deployment in trillion-dollar platforms (Copilot, Azure, LinkedIn).
5. Perplexity’s Pivot: Ditching Ads for Subscriptions
[13:16 – 16:35]
- Strategic Shift:
- Perplexity, known for its AI search, drops advertising plans citing user trust concerns.
- Company to focus on subscriptions and enterprise deals rather than general, mass-market search.
- Quote: “Google is changing to be like Perplexity more than Perplexity is trying to take on Google.” (Perplexity executive, paraphrased) [14:00]
- Quote: “One of the things that's starting to become clear to us is that Perplexity isn't for everyone.” (Perplexity executive) [15:53]
- Growth & Market Data:
- Perplexity: 60M monthly active users (Jan 2026)—double from last year, but <10% of OpenAI/Google (800M+/month).
- Its Comet browser had 2.8M weekly actives (Dec 2025), down from a high of 7.8M earlier in 2025.
- Industry Context:
- Industry-wide skepticism on ad-based AI search as Google and Meta dominate the ad model; accuracy and trustworthiness prioritized.
- Anthropic echoes anti-ad stance; ChatGPT’s ads lampooned in Super Bowl as AI firms chase sustainable monetization.
6. Self-Driving Car Market: Uber’s Diminishing Role
[16:35 – 19:27]
- Market Update:
- Robotaxi sector viewed as a “two horse race” (Tesla, Waymo)—Uber notably diminished.
- Waymo expanding to new cities, now operates more independently, decreasing visibility of Uber partnership.
- Uber’s stock fell by 25% over six months, reflecting anxieties over AV disruption.
- Company attempts reassurance with slide decks, but Wall Street skepticism remains.
- Quote: “Uber is currently in the difficult position of having to prove that it won’t be disrupted by a still burgeoning new technology.” (Moffett Nathanson analyst Mike Morton) [18:48]
- Perspective:
- It’s too early to call a winner—robotaxi market will take “years, not months” to unfold; Uber needs to regain narrative control and strategic footing.
7. Weekend Long Read: Foldables as Laptop Replacements
[19:27 – 20:24]
- Recommended Read:
- The Verge’s Allison Johnson explores whether you can replace a laptop with a foldable phone.
- Selected Quote: “I don't need a foldable to be my all day laptop… I just want something easy to carry in a regular bag so I can be a human out in the world for a few hours.” (Allison Johnson) [19:48]
- Summary:
- Foldables offer portability and convenience for short, mobile work sessions, but aren’t practical for a full workday compared to traditional laptops.
Memorable Quotes
- “The engineers let the AI agent resolve an issue without intervention. The outages were small but entirely foreseeable.” (Senior AWS Employee) [01:31]
- “Today's indictments are an important step towards accountability and we'll continue working to ensure our trade secrets remain secure.” (Jose Castaneda, Google spokesperson) [04:59]
- “Staffers are known to refer to what they believe [Jony] Ive would want frequently during conversations.” [10:17]
- “Our engineering teams are taking action on the report’s findings.” (Eric Horvitz, Microsoft) [12:15]
- “One of the things that's starting to become clear to us is that Perplexity isn't for everyone.” (Perplexity executive) [15:53]
- “Uber is currently in the difficult position of having to prove that it won't be disrupted by a still burgeoning new technology.” (Mike Morton, Moffett Nathanson) [18:48]
- “I don't need a foldable to be my all day laptop... I just want something easy to carry in a regular bag so I can be a human out in the world for a few hours.” (Allison Johnson, The Verge) [19:48]
Timestamps for Key Segments
- [00:04] – AI Causes AWS Outages
- [03:05] – Google Trade Secrets Indictment
- [05:45] – OpenAI’s Hardware Ambitions & Culture
- [10:34] – Microsoft’s AI Content Standards Proposal
- [13:16] – Perplexity Ditches Ads, Focuses on Subscription
- [16:35] – Robotaxi Market: Uber’s Struggles
- [19:27] – Foldable Phones vs. Laptops
This episode paints a picture of a tech industry grappling with the unintended side effects of rapid AI adoption, legacy security risks, shifting business realities, and the challenges of innovation in consumer devices and transportation.
For anyone wanting a snapshot of the real risks—and odd culture clashes—fueling today’s most important tech stories, “When AI Breaks Things” is packed with sharp insights and rich context.
