Tech News Weekly 416: "AI Pricing on Instacart"
Date: December 11, 2025
Host: Micah Sargent
Guests: Amanda Silberling (TechCrunch), Derek Kravitz (Consumer Reports)
Episode Overview
In this episode, Micah Sargent is joined by Amanda Silberling from TechCrunch as co-host and later by Derek Kravitz from Consumer Reports. The show tackles several major tech stories, including the rise of AI usage in dating apps, a new minimalist wearable from Pebble, collaborative AI standards, and an in-depth investigation into Instacart’s AI-driven price variation.
Main Topics & Key Insights
1. The CEO of Hinge Leaves to Launch an AI Dating App
Segment: 02:57–13:14
Discussion points:
- The CEO of Hinge has stepped down to create "Overtone," an “early stage dating service” using AI and voice tools to foster more personal connections.
- Amanda and Micah discuss the evolution of dating apps, especially how algorithms and monetization strategies (like Hinge’s “rose jail”) have changed user experiences (04:50).
- Amanda:
"Hinge still is very popular and it is considered one of the better ones, but it's like one of the better lumps of coal in a pile of coal for the thematic metaphor." (05:52)
- Amanda:
- The perceived paradox of adding more AI to something as inherently human as dating. Both note user fatigue with dating apps and skepticism toward using AI to automate even more of the process.
- The growing trend: Founders of popular dating apps spinning off AI-focused projects (e.g., Whitney Wolfe Herd with Bumble).
- Amanda on AI doing the dating for you:
“What if I had an AI and my AI could talk to your AI? And then we don't even have to go on a date to know if we're compatible because our AIs went on a date for us…” (07:54)
- Amanda on AI doing the dating for you:
- Concerns raised about authenticity, potential emotional coldness, and philosophical questions about whether AI can/should help us “find love.”
Memorable moment:
Amanda’s “Black Mirror” comparison regarding AI dating agents:
"...hand scanning to pay for groceries seems quaint compared to, like, my AI is going to go date for me." (08:42)
2. Pebble’s Minimalist Voice-Capture Smart Ring
Segment: 14:22–26:29
Discussion points:
- Eric Migicovsky (Pebble’s founder) launches the Pebble Index 01: a $75 smart ring intended for quick voice memos, not health tracking or always-on AI.
- Unique approach: No always-listening features, no biometric sensors—just a microphone and button for on-demand recording/transcription.
- Micah:
"It’s an interesting counter programming move in an industry that's obsessed with doing everything everywhere all at once." (16:47)
- Micah:
- Amanda expresses initial enthusiasm:
"I was surprised that I really like this idea of the index ring..." (19:30)
- Amanda nearly ordered it but instead found a productivity app (Akiflow) and customized her workflow to replicate much of the ring’s function on her phone.
- Both hosts criticize the ring’s non-rechargeable, disposable battery.
- Micah:
“It does not have a rechargeable battery... once it’s dead, it becomes e-waste.” (23:04)
- Micah:
- General consensus: Devices should prioritize on-demand, privacy-respecting functionality (over “always on”), but e-waste concerns must be addressed.
- Anticipation that more discrete-purpose wearables will arrive on the market.
Memorable moment:
Amanda humorously describes being moments away from buying the ring in a coffee shop, only to pause for privacy/security concerns.
3. The Rise of the Model Context Protocol (MCP) – Standardizing AI Connections
Segment: 36:29–41:58
Discussion points:
- MCP (Model Context Protocol) is now an industry standard, enabling AI agents to reliably access and interact with various apps, data sources, and workflows.
- Origin: Began as an Anthropic engineers’ hackathon project, grew rapidly after industry adoption (OpenAI, Microsoft, Google, etc.).
- MCP now transferred to the Linux Foundation and governed by the Agentic AI Foundation (AAF).
- Key quote:
"It’s essentially a 'show me what you’ve got.'" – Connor Kelly, Product Marketing Manager, Anthropic (39:37)
- Key quote:
- Significance: Analogous to API standards in Web 2.0 and home automation protocols like Matter. Makes AI integration seamless for end users.
- Jim Zemlin (Linux Foundation):
“I’ve never seen anything like this. I can barely keep up with the number of inbound calls from organizations who want to be a part of this.” (40:54)
- Security, privacy, and future potential are highlighted as MCP proliferates.
4. AI-Driven Price Discrimination on Instacart (w/ Derek Kravitz, Consumer Reports)
Segment: 41:58–65:30
Background:
Consumer Reports led a massive study into Instacart’s use of AI (via its Eversight acquisition) to dynamically vary prices—meaning different users can pay different prices for the exact same cart of groceries.
How the Investigation Worked (46:06)
- Coordinated with 437 volunteers who simultaneously shopped for identical items and addresses in the Instacart app.
- Controlled for variables (location, timing, order contents) to isolate price-setting differences.
- Found evidence of “price buckets,” indicating algorithmic assignment of different buyers to different pricing groups, sometimes with 23% differentials for the same item.
Key stats:
- 74% of grocery items had different prices for different shoppers
- Variances for individual items could exceed 20%
- For a standard 20-item basket, price differences ranged from $114 to $124 (Seattle Safeway test); annualized, this could cost a family up to $1,200 extra.
Example:
- Skippy Peanut Butter: Variances from $2.99 to $3.69 across “buckets.”
- No correlation with loyalty, geographic area, race, gender, or income.
Memorable moment:
Micah:
"[...] you're charging me this for eggs and them that for eggs?" (48:53)
The Role of AI & Company Responses (55:59)
- Instacart’s Eversight software enables this dynamic/pricing “experimentation.”
- Instacart confirmed: 10 partners use this tech (but refused to name all 10).
- Target claims Instacart scrapes prices and marks up for operational costs, adding yet another layer of complexity.
Real-World Implications
- Algorithms group users into “sub-populations” and test their willingness to pay (via patents and interviews).
- Critical question for society: Should this type of “invisible price experimentation” even be allowed?
- Regulators (e.g., FTC’s Lina Khan) see an urgent need to address and potentially regulate AI-enabled price discrimination.
Derek Kravitz:
“The big takeaway... is even if you don’t use Instacart, their data—they sell it to everyone... they’re essentially creating the market in economic terms. They’re sort of creating the data out of thin air and then they’re selling it back to companies.” (61:34)
Notable Quotes
-
Amanda Silberling on Dating App Algorithms:
"It’s like one of the better lumps of coal in a pile of coal for the thematic metaphor." (05:52)
-
Micah Sargent on AI Dating Agents:
"That's sad. Is that not sad?" (08:22)
-
Connor Kelly (Anthropic) on MCP:
"It’s essentially a 'show me what you’ve got.'" (39:37)
-
Jim Zemlin (Linux Foundation):
"I've never seen anything like this. I can barely keep up with the number of inbound calls..." (40:54)
-
Derek Kravitz (Consumer Reports):
"For whatever reason, these folks were being placed into groups that all got the same price for the same exact items at the same exact time." (46:51)
"It’s invisible to us. We can’t really tell which group we’re in, but we do know that’s why." (62:38)
Timeline of Major Segments
| Timestamp | Segment / Key Topic | |-----------|---------------------------------------------------------------| | 02:57 | Amanda Silberling on Hinge CEO’s AI dating app | | 13:14 | Micah’s story: Pebble Index 01 smart ring | | 26:29 | Wearables discussion wrap up | | 36:29 | MCP – AI agent protocol & industry collaboration | | 41:58 | Interview w/ Derek Kravitz: Instacart AI-powered price disparity| | 46:06 | Methodology of Instacart price study | | 49:53 | Price differences on specific grocery items | | 55:59 | Eversight acquisition, Instacart & company reactions | | 60:58 | Real-world impact & possible regulatory action | | 65:30 | Show close / contact info |
Resources & Further Reading
- Consumer Reports – Instacart Pricing Investigation Report
- The Verge – Model Context Protocol & AI Standards
Conclusion
This episode delivers a nuanced, skeptical look at the AI-ification of everything—from our love lives to our grocery bills—while highlighting the need for user awareness, thoughtful regulation, and truly user-centric innovation. It’s an episode full of real-life stakes, surprising data, and plenty of laugh-out-loud asides.
Summary compiled to capture discussion highlights, pivotal quotes, and context for each major segment. For more, find the full show notes or episode audio/video at TWiT TV.