Tech Brew Ride Home – November 6, 2025
Main Theme:
A rapid-fire update on major developments in AI and tech, headlined by Apple’s reported partnership with Google to power the next-gen Siri using Gemini, OpenAI’s government ambitions, the unveiling of Google’s new AI chip, and two new AI-powered consumer products.
Key Discussion Points & Insights
1. Apple Likely to Use Google’s Gemini to Power Siri
[01:15 – 06:01]
- Apple is reportedly finalizing a deal to pay Google ~$1B/year for access to Gemini, Google’s 1.2 trillion parameter AI model, to supercharge a revamped Siri ("the new Siri," targeted for spring 2026 release).
- The Gemini model’s parameters vastly outnumber Apple’s current 150B parameter cloud AI.
- Quotes & Details:
- “The iPhone maker is banking on Google’s help to rebuild Siri’s underlying technology, setting the stage for a new slate of features next year.” (Brian quoting Bloomberg, [01:42])
- "The hope is to use the technology as an interim solution until Apple's own models are powerful enough." ([02:30])
- Why Gemini?
- Apple tested Gemini, ChatGPT (OpenAI), and Anthropic's Claude, ultimately choosing Gemini for interim needs.
- How will it work?
- Siri’s “summarizer” and “Planner” functions (for synthesizing info, executing tasks) handled by Gemini.
- Other Siri features to remain in-house.
- Model runs on Apple’s own private cloud compute, data remains walled off from Google.
- Management and strategy:
- Project led by Mike Rockwell (Vision Pro lead) and Craig Federighi.
- Project codenames: Siri’s new AI backend—“Glenwood”; new Siri itself—“Linwood”.
- Public branding/subtlety:
- Apple treats Google here as a “behind the scenes technology supplier,” distinct from their public Google Search deal.
- No direct Gemini chatbot integration into main OS; separate from prior talks.
- Long-term>
- Apple is not planning on Gemini as a permanent solution, seeks to build its own 1T param model by next year.
- “Apple still doesn’t want to use Gemini as a long term solution, despite the company bleeding AI talent... Management intends to keep developing new AI technology and hopes to eventually replace Gemini with an in house solution.” ([05:16])
2. OpenAI’s Financial Outlook and Government Backing
[06:01 – 09:35]
- IPO Status & Data Center Financing
- CFO Sarah Friar says “IPO is not on the cards right now” ([06:30]), emphasizing a focus on scaling, R&D over immediate profits.
- Hints OpenAI is seeking US government backing (“backstop financing”) for its massive AI chip/data center expansion.
- “The depreciation rates of AI chips remain uncertain, making it more expensive for companies to raise the debt needed to buy them.” ([07:31])
- Fryer envisions support from “banks, private equity, maybe even governmental... any such guarantee can really drop the cost of the financing but also increase the loan to value.” ([07:42])
- Business Growth
- Enterprise sales are now 40% of OpenAI’s revenue (up from 30% in January).
- Revenue growth seen in sectors like financial services & healthcare.
- Computing power: will reach 2 GW at year-end, up from 200 MW two years ago.
- “We are talking about country-sized deployments,” Fryer said ([09:20])
- Profit Potential
- “If I had to get to breakeven, I have a healthy enough margin structure that I could do that by pulling back on investment.” ([08:40])
3. Microsoft AI Roadmap & Caution on “Humanizing” AI
[09:35 – 12:32]
- CEO Mustafa Suleyman at Microsoft AI:
- While praising OpenAI, stressed Microsoft’s ambition to develop its own voice, image, and text AI models—“self-sufficiency beyond OpenAI.”
- Warns against anthropomorphizing AI:
- Quote:
“AI is going to become more human like, but it won’t have the property of experiencing suffering or pain itself and therefore we shouldn’t over empathize with it.” ([10:57])
- Quote:
- Human values, control are emphasized:
- “We want to create types of systems that are aligned to human values. By default that means they are not designed to exceed and escape human control.” ([11:23])
- Product/Healthcare Focus
- Copilot (their chatbot) relies on OpenAI now, but Microsoft pushes own models before 2032 (contract expiry).
- Health partnerships (Harvard Health) and new AI diagnostic tools — reportedly diagnosing disease at “four times” the accuracy of some doctors, at lower cost.
- “AI diagnostic tools are very close to being market ready, he said.” ([12:18])
- Microsoft’s models focus on “containment,” avoiding giving any impression of consciousness.
4. Google’s Ironwood: Next-gen AI Chip
[12:35 – 14:54]
- Ironwood TPU Launch:
- Google's 7th generation in-house AI chip.
- “More than four times faster than its 6th gen TPU...” ([12:50])
- Handles LLM training, AI agents, real time apps — up to 9,216 chips in a pod.
- Custom silicon offers price, performance, efficiency advantages vs. Nvidia GPUs.
- Google's 7th generation in-house AI chip.
- Adoption & Infrastructure Race:
- Anthropic to use up to 1 million Ironwood chips to run Claude model.
- Google cloud business: Q3 cloud revenue $15.15B (+34% YoY) — fierce race vs. Azure (+40%) and AWS (+20%).
- Growth & Demand:
- Increased 2025 capex forecast to $93B.
- “We are seeing substantial demand for our AI infrastructure products including TPU based and GPU based solutions,” — Sundar Pichai ([14:30])
- Increased 2025 capex forecast to $93B.
5. Two New AI Consumer Products
A. Sandbar “Stream Ring” (AI Smart Ring) [14:54 – 16:53]
- A $249+ wearable AI ring: transforms whispered or spoken notes into transcribed text inside an app; aims to capture fleeting “shower thoughts”.
- “The Stream Ring is an AI powered smart ring designed to turn fleeting inner monologues into organized notes and conversations.” ([15:35])
- Built by alums of Control Labs & Meta.
- Designed for privacy (audio not stored), on-device and cloud LLM blend, haptic controls.
- Includes media controls; waterproof; all-day battery.
- Integrates ElevenLabs for voice cloning playback, supports alternate voices.
- Launching with 3 months Pro, ships in summer 2026.
B. Hopscotch Labs “Bee Bot” [16:53 – End]
- New app by Foursquare founder Dennis Crowley: location-aware, audio-based social experience (“AI DJ for your life”).
- “The Bee Bot DJ… will then be able to talk to you as you go throughout your day and alert you to interesting events, landmarks or updates from friends who happen to be nearby.” ([17:48])
- Connects to headphones/smart glasses, surfaces local info based on interests/contacts.
- Employs a mix of LLMs and synthetic voices.
- No badges/mayorships or gamification, but aims for the original Foursquare’s “playful spirit.”
- “Powered by a TikTok style algorithm, Crowley says, but one that’s focused on what’s happening nearby and in real life.” ([18:35])
Notable Quotes
-
Brian (host), summarizing Apple’s move:
“The hope is to use the technology as an interim solution until Apple’s own models are powerful enough.” ([02:30]) -
Sarah Fryer, OpenAI CFO:
“IPO is not on the card right now... I don’t want to get wrapped around an IPO axle.” ([06:30])
“We are talking about country-sized deployments,” ([09:20]) -
Mustafa Suleyman, Microsoft AI CEO:
“AI is going to become more human like, but it won’t have the property of experiencing suffering or pain itself and therefore we shouldn’t over empathize with it.” ([10:57])
“We want to create types of systems that are aligned to human values. By default that means they are not designed to exceed and escape human control.” ([11:23]) -
Sundar Pichai, Google CEO:
“We are seeing substantial demand for our AI infrastructure products including TPU based and GPU based solutions.” ([14:30])
Timestamps for Key Segments
- 01:15 – 06:01: Apple’s Siri/Gemini partnership (deal details, strategy, context)
- 06:01 – 09:35: OpenAI’s finances, IPO, and government support pitch
- 09:35 – 12:32: Microsoft AI independence ambitions, “AI as non-human” stance, healthcare focus
- 12:35 – 14:54: Google Ironwood TPU announcement, cloud infrastructure race
- 14:54 – 16:53: Sandbar Stream Ring demo and details
- 16:53 – End: Hopscotch Labs Bee Bot app overview
Memorable Moments
- The parallel between the original Foursquare’s serendipity and Bee Bot’s aim to create “playful” real-world audio interactions. ([17:35])
- The candid reveal of just how much AI infrastructure now requires “country sized deployments” ([09:20]) and the possible role of governments as financial backers.
Tone
Casual and informative, with a “water cooler” vibe that breaks down high-level business and technical news for a time-crunched, tech-savvy audience.
For anyone who missed the podcast: This episode delivers a jam-packed, digestible round-up of the biggest AI and tech news stories, with direct quotes and lively narration that brings out the personalities and stakes behind the headlines. From Apple and Google’s blockbuster partnership to the latest in consumer AI gadgets, it’s a brisk and engaging listen—or read.
