Tech News Weekly 408: OpenAI DevDay – ChatGPT OS?
Date: October 9, 2025
Host: Micah Sargent
Guests: Amanda Silberling (TechCrunch), Sabrina Ortiz (ZDNet)
Overview
This episode unpacks some of the week's most fascinating tech stories, centering on:
- The arrival of a new AI pet (Casio’s Mofflin) and the role of “emotional” AI companions.
- OpenAI’s Dev Day and how ChatGPT is evolving into an “operating system” for apps and AI agents.
- Developments around government attempts to compromise smartphone encryption in the UK.
- The rapid, risky expansion of “Buy Now, Pay Later” services.
Segment 1: Amanda Silberling & Casio’s AI Pet Mofflin
[00:00 – 17:00]
The Mofflin: What Is It?
- Amanda was sent Casio’s new AI pet—the Mofflin—for review.
- Describes it visually: “Imagine like, an orange-haired guinea pig...with so much hair that it's hard to see the eyeballs.” [02:30, Amanda]
- Listeners liken it to Star Trek’s Tribble, but Amanda prefers “guinea pig.”
Personality Development & Interactivity
- Development Curve: Takes “25 to 50 days to ‘develop a personality.’”
- Interactivity depends on user input – petting, talking, carrying around, etc.
- Sensors & Hardware: Two microphones, gyroscope, various touch sensors. No camera.
- Can operate without its companion app, but app adds features and personality tracking.
“The only way you can elicit a negative reaction...is putting him on his back. He does get upset.” [08:21, Amanda]
Social Experiments
- Amanda took Mofflin to Pilates: “Bringing an AI pet around is very odd behavior.” [07:02]
- Introduced Mofflin to a friend’s Yorkie—elicited curiosity and some trepidation from the dog.
Who Is It For? Price & Value
- Target market is ambiguous—possibly kids, hypoallergenic pet seekers, or elderly in memory care.
- Retail price: $429 ($430) – considered steep.
“That’s nearly $500...You could get a Switch 2 or you could get this.” [12:50, Micah]
- Amanda reflects: “It doesn’t feel that much more advanced than a Furby or a Fur Real Friend...it’s like a toy.” [14:17]
Data Privacy & Security
- Initial privacy concerns due to microphones and Bluetooth connectivity; company claims voice data is used but encrypted.
- Early app tests with cybersecurity editor Zach Whitaker found no obvious privacy issues; more research planned.
“It was promising...the app seemed pretty benign...bare bones in a way that was pleasantly surprising.” [20:54, Amanda]
Memorable Quotes & Moments
- “I can’t hurt this guy. But when he goes on his back, he does get upset.” [08:21, Amanda]
- On the price: “As much as I think this thing is adorable...I don’t want it for that price.” [13:22, Micah]
Segment 2: UK Encryption Backdoor Battle
[24:39 – 28:59]
The Issue
- UK government’s renewed request for “backdoor” access to iPhone and iCloud encrypted data (now focused only on British users to avoid prior diplomatic conflict).
- Apple resisted, pulling advanced security features from UK users rather than compromise encryption globally.
“Weakening encryption anywhere means weakening encryption everywhere.” [27:34, Micah]
Takeaway
- UK’s targeted approach still risks collateral access to non-UK data.
- Ongoing global debate: security, privacy, and government access.
Segment 3: OpenAI Dev Day—ChatGPT Evolves
[32:12 – 47:28]
Guest: Sabrina Ortiz, ZDNet
Major Announcement: ChatGPT as an App Platform
- ChatGPT will now allow third-party apps (Spotify, Canva, Zillow, Coursera, etc.) to run natively inside the chat interface.
- Unlike prior “GPTs,” the new apps can provide full custom UIs within ChatGPT—expandable to full screen.
“You could...access everything you would in the app...but within the chatbot.” [32:42, Sabrina]
- Example: Ask Canva via ChatGPT to design a poster and work on it without ever leaving the interface.
Operating System Vision
- OpenAI executives (notably Nick Turley, per media Q&A) explicitly described ChatGPT’s trajectory as becoming an "operating system" for AI-enabled interaction with software.
“He said it could be as soon as a six month window where...ChatGPT would feel meaningfully different.” [39:14, Sabrina]
Developer Energy & Agentic AI
- Dev Day energy was electric: “Oohs and aahs” as Sam Altman introduced new features. [34:10]
- “Agentic AI” & AgentKit: Building and deploying custom AI agents showcased – demonstration of an agent built “in under seven minutes.”
- Lowers the barrier for product teams to integrate advanced AI workflows.
“It places the ability to build in more people’s hands...and the ability to build more quickly.” [41:30, Sabrina]
Transformative Integrations & Industry Impact
- Beyond creative tools: Use case like Zillow—using conversational narrowing for home search.
- Encourages industries to rethink user experiences as “conversational” and AI-powered.
“[Zillow’s integration is] not about creating something new, but actually creating an experience that's easier for the user...” [43:14, Sabrina]
Hype vs. Lasting Value
- Sam Altman called the industry “bubbly”—alluding to hype and speculation.
“Do I think everything they're putting out is in the value camp? Not necessarily, but they're the first ones doing it...they are helping distinguish what is bubbly and what is actually here to stay.” [44:55, Sabrina]
Key Quotes
- “Combining the apps you know and love...with ChatGPT’s natural language processing...you’re getting best of both worlds.” [36:59, Sabrina]
- On operating system shift: “Six months is plenty of time for developers to get building...I don’t see why companies wouldn’t want to develop an application for ChatGPT...” [39:14, Sabrina]
Segment 4: The Risks of “Buy Now, Pay Later”
[47:32 – 65:41]
Explosion in Use
- BNPL exploded from $2B (2019) to $120B (2023); half of Americans have used it, especially <35 age group.
Hidden Dangers
- While “democratizing luxury,” BNPL can snowball into confusion, debt, and high interest (up to 36%).
- Algorithms encourage overspending via psychological tricks and temporal reframing—making $365 feel like “$1 a day.”
- “Regulatory wild west” – minimal oversight, skirting credit card and Truth in Lending rules.
Personal & Social Stories
- Anecdotes: From $430 handbags to groceries, designer concert tickets – people accrue debts without fully tracking repayments.
“Users can finance everything from a burrito all the way up to...vacations...it appears...these services operate in a regulatory wild [zone].” [47:32, Micah]
- TikTok content normalizes BNPL debt as “self-care” or “investing in yourself.”
Skeptical Perspective
- Host’s caveat: While BNPL can help with genuine needs (personal mattress anecdote), the real risk is losing track as users hop between services without seeing the total owed.
- FICO soon to track BNPL usage, which may shock many who treat it like a loophole.
“It can become a slippery slope very easily.” [51:45, Micah]
- Noted difference: Services do show payment schedules, but juggling multiple accounts can make budgeting hard.
Final Reflection
- BNPL can be helpful, but also “incredibly predatory.”
- “I will not look down on people making use of Buy Now Pay Later services. I will look down on how Buy Now Pay Later services can be incredibly predatory.” [58:50, Micah]
Notable Quotes
-
On Mofflin’s price:
“That’s nearly $500...You could get a Switch 2 or you could get this.”
— Micah Sargent [12:50] -
On ChatGPT’s new app platform:
“Combining the apps you know and love...with ChatGPT’s natural language processing...you’re getting best of both worlds.”
— Sabrina Ortiz [36:59] -
Encryption Debate:
“Weakening encryption anywhere means weakening encryption everywhere.”
— Micah Sargent [27:34] -
BNPL caution:
“It can become a slippery slope very easily.”
— Micah Sargent [51:45]
Episode Structure: Timestamps & Content Flow
- 00:00–17:00 — Amanda Silberling on Casio’s Mofflin AI Pet
- 17:00–24:39 — (Ads Omitted)
- 24:39–28:59 — UK’s renewed iCloud encryption backdoor demand
- 32:12–47:28 — Sabrina Ortiz on OpenAI Dev Day, ChatGPT as App “OS”
- 47:32–65:41 — Buy Now, Pay Later industry cautionary tale
- (Advertisements and show outro skipped)
Tone & Takeaways
- The tone throughout remains upbeat, curious, and occasionally irreverent.
- There’s a pragmatic skepticism toward AI gadgets’ real-world usefulness and toward hyped tech trends, coupled with genuine appreciation for innovation.
- The hosts and guests blend personal anecdotes, industry perspective, and practical concerns—especially around privacy, regulation, and user risks.
- Audiences leave with a nuanced understanding of new AI and financial services: both their promise and pitfalls.