- Product Tapas
- Posts
- Perplexity Browser Security Flaw, Google Translate vs Duolingo, AI Unit Economics Reality
Perplexity Browser Security Flaw, Google Translate vs Duolingo, AI Unit Economics Reality
Plus: Escaping Feature Request Traps, Why Timing Beats Strategy and Vertical Adoption Curves
Alastair Preacher
August 29, 2025 • Estimated Reading Time: 15 minutes
In partnership with
We track Product so you don't have to. Top Podcasts summarised, the latest AI tools, plus research and news in a 5 min digest.
Hey Product Fans!
Welcome to this week’s 🌮 Product Tapas.
New here? We're your shortcut to staying sharp. Essential stories, practical tools, real insights.
For the best reading experience, check our web or app version and sign up for future editions here.
What's fresh this week? 🥬
Google ditches keynotes for Jimmy Fallon (peak Valley cringe?), Anthropic caves to enterprise whining, Harvard promises vibe-thinking glasses (sure), and 50% LLM adoption meets flat productivity growth. The future's messier than promised.
📰 Not Boring → AI enterprise push, Google's everything strategy, hardware wearables race
⌚️ Productivity Tapas → Anti-fraud tools, AI task completion, mobile agent control
🍔 Blog Bites → Smart AI automation frameworks (the 3/2/1 rule), why timing beats everything in strategy, and escaping the feature request trap
🎙️ Pod Shots → Brett Taylor & Clay Bavor reveal why AI adoption curves look vertical (spoiler: it's riding on 30 years of infrastructure investment)
Let's go 🚀
📰 Not boring
Google's AI Everything Strategy
Google's Made by Google event ditched the keynote for a Jimmy Fallon–hosted late-night show, but the real spotlight was on Gemini AI. It now powers the new Pixel 10 lineup (AI camera tools, live call translation, journaling, more), fuels smart home upgrades with Gemini for Home, and drives AI health coaching across FitBit, Pixel Watch 4, and even Pixel Buds. Plus, the new Tensor G5 chip brings 60% stronger AI performance.
AI Mode in Search gets new agentic features and expands globally (at $250 pcm)
Apple explores using Google Gemini AI to power revamped Siri
Hardware & Wearables Race
The Economics of AI
AI is reshaping not just products but the unit economics of tech companies — and the results look very different depending on business model.
Duolingo: AI is cheaper to run than expected, trimming margin impact to ~1pt while fueling engagement with Roleplay.
Snap: Infra costs per DAU edged up, but ARPU rose faster — showing AI can be a monetisation lever, not a drag.
Reddit: AI ad tools delivered +10% CTRs and 2x ROAS, helping push ad revenue up 84% YoY
Takeaway: AI is now a line‑item with leverage — sometimes as a cost saver, other times as spend that accelerates monetisation.
Everything Else
AWS CEO says using AI to replace junior staff is 'Dumbest thing I've ever heard'
India tech giant TCS layoffs herald AI shakeup of $283 billion outsourcing sector
LLM adoption among US workers is closing in on 50%. Meanwhile labor productivity growth is lower than in 2020. Perhaps that productivity spike isn't going to be quite what we expected….
The future of surveys and data: Digital Twins now achieve 85% accuracy in backfilling
JIRA continues to be the most-used product management software by a lot- except at small companies, where Linear is used almost as much as JIRA
Brave Browser flagged a concerning flaw in Perplexity's Comet browser that puts users' accounts and other sensitive info in danger. Oops
Anthropic bundles Claude Code into enterprise plans after months of complaints about usage limits
Google Translate takes on Duolingo with new language learning tools
Find out why 1M+ professionals read Superhuman AI daily.
In 2 years you will be working for AI
Or an AI will be working for you
Here's how you can future-proof yourself:
Join the Superhuman AI newsletter – read by 1M+ people at top companies
Master AI tools, tutorials, and news in just 3 minutes a day
Become 10X more productive using AI
Join 1,000,000+ pros at companies like Google, Meta, and Amazon that are using AI to get ahead.
⌚️ Productivity Tapas: Time-Saving Tools & GPTs
Roundtable - Proof of Human: Stop bots and fraud without slowing down real users. Score all your traffic in real-time, letting only verified humans pass through. Death to CAPTCHA
Cobot: A to-do app that also promises to do the task for you. 😂 Love it.
Omnara: Launch and control Claude code from anywhere (monitor, debug and guide your agent from your phone)
Remember. Product Tapas subscribers get our complete toolkit - 460+ personally tailored, time-saving tools for PMs and founders. Your shortcut to efficiency and what's hot in product management 🔥
Check the link here to access.
🍔 Blog Bites - Essential Reads for Product Teams
Strategy: The 3/2/1 Rule for Smart AI Automation Decisions
Craig Unsworth explores a practical framework for deciding what to automate with AI and what should remain human-driven. He introduces a simple yet effective decision-making tool based on three core pillars that prevents organisations from automating the wrong processes. Read the full article here.
💡 "That's why the challenge of automation isn't technical – it's judgmental. It's about knowing where automation genuinely adds value, and where it undermines it."
Key Takeaways
• The Three Pillars Framework: - Quality: Automation must deliver output as good as or better than capable humans; - Replicability: Same input must reliably produce same output for operational stability; - Explainability: Must be able to explain why the system made specific decisions for trust and compliance
• The 3/2/1 Decision Rule: - Three out of three pillars = Run the experiment immediately; - Two out of three = Worth considering but ask deeper questions about the missing element; - One out of three = Don't automate; fix the underlying process first
• Strategic Application Guidelines: - Start small with low-risk processes to build confidence and proof points; - Use framework as common language between executives, product teams, and operations; - Look for processes that only meet one pillar as signals for strategic process improvement work
• Implementation Best Practices: - Avoid locking into single AI vendor or tool for maximum flexibility; - Consider hybrid approaches where automation handles baseline tasks and humans manage edge cases; - Think of AI as composable infrastructure that can be mixed and matched as needed
• Risk Mitigation: - Automating wrong processes wastes resources and introduces errors at scale; - Poor automation can remove customer-valued nuance and lock bad processes in place; - Missing explainability creates trust issues with leaders, regulators, and customers
Strategy: Why Timing Beats Everything in Strategic Decision-Making
John Cutler explores how reframing urgency transforms both strategic thinking and prioritisation effectiveness. He introduces a powerful new definition of urgency based on value decay over time, revealing why most prioritisation frameworks miss the strategic mark. Read the full article here.
💡 "Urgency is the degree to which a delay reduces the long-term value of a decision or action."
Key Takeaways:
• Urgency Redefined: - Urgency measures how much delay diminishes long-term value, not just immediate pressure; - Market-tipping scenarios create high urgency (first to 30% share gains momentum advantage); - Expansion revenue typically has lower urgency unless tied to earnings guidance or investor expectations
• Strategic Timing: - Strategy is as much about when you do things as what you do; - Companies can control tempo by speeding up markets with rapid launches or slowing them through regulation; - Proactive urgency opens future options; regrettable urgency constrains them
• Prioritisation Framework Gaps: - Popular frameworks (RICE, ICE, MoSCoW, Kano) don't explicitly address time's impact on value; - Even WSJF/Cost of Delay reduces timing to single numbers rather than exploring value decay patterns; - Teams often equate urgency with anti-strategic thinking, missing timing's strategic importance
• Strategic Questions for Teams: - What is our window of opportunity and how does waiting affect it?; - How can we shift from reactive to proactive urgency?; - What bets can we place now to put time on our side later?
• Competitive Dynamics: - Consider how your timing assumptions differ from competitors'; - Look for ways to instill urgency in competitors while maintaining your own strategic patience; - Identify opportunities to delay market convergence in beneficial ways
• Organisational Reality: - Many companies complaining about lack of strategy are actually experiencing regrettable urgency from past decisions; - When strategic options are constrained, "doing everything" becomes the de facto strategy; - Weaving time impact into value discussions shifts conversations toward strategic thinking
Product Strategy: Escape the Feature Request Trap by Following the Breadcrumbs
Tom & Corissa from Crown & Reach explore how product teams can avoid the common pitfalls of feature request management by treating requests as clues rather than commands. They introduce the concept of seeking out the "problemplex" - the interconnected bundle of problems that lies beneath every feature request. Read the full article here.
💡 "Feature requests aren't perfect information or directly instructive. Neither are they worthless or to be ignored. They're breadcrumbs that will lead you towards the thing behind the thing – if you follow the trail far enough into the undergrowth."
Key Takeaways:
• Common Anti-Patterns: - The "teetering mountain" - saying yes to everything and trying to prioritise conflicting ideas; - The "oxbow lake" - creating rigid processes that force requesters to bypass you entirely; - Both approaches lead to building unused features and taking the blame
• Understanding the Problemplex: - Feature requests represent intuitions about resolving interconnected problems; - Requesters are limited by cognitive constraints and can only imagine familiar solutions; - The real value lies in uncovering the chain of perceptions, decisions and behaviours behind requests
• Investigation Techniques:- Ask "How do you currently...?" to understand existing workflow; - Break down requests into granular user interactions and decision points; - Explore workarounds people are already using as evidence of genuine need
• Practical Examples:- Intercom's map feature wasn't needed for analytics but for impressive slide presentations; - "Search like Amazon" can become 12+ micro-improvements targeting specific pain points; - Simple solutions often work better than complex feature builds
• Three Valuable Outcomes:- New, smaller, testable ideas linked to measurable behaviours; - Intelligence about existing workarounds that prove demand; - Opportunities to remix existing functionality rather than building from scratch
🎙️ Pod Shots - Bitesized Podcast Summaries
Remember, we've built an ever-growing library of our top podcast summaries. Whether you need a quick refresher, want to preview an episode, or need to get up to speed fast - we've got you covered.
Check it out here
🎯 How is AI Different Than Other Technology Waves? The Compounding Effect That's Changing Everything
We're witnessing something unprecedented in technology adoption. ChatGPT reached 100 million users in just two months—faster than any consumer product in history. But according to Brett Taylor (OpenAI Chairman, former Salesforce co-CEO) and Clay Bavor (former Google Labs head), this isn't just about speed. It's about compounding.
In another great conversation on Acquired, these two legendary product leaders—who between them helped create Google Maps, led Facebook's platform strategy, and built everything from Gmail to Google's AR/VR initiatives—reveal why AI represents a fundamentally different type of technology wave. One that's riding on the coattails of every previous innovation.
Their company Sierra, which builds customer-facing AI agents, is growing faster than any B2B software company they've ever seen. But the real story isn't just about their success—it's about how AI is leveraging decades of infrastructure investment to create adoption curves that look almost vertical.
Oh and finally, I’ve included only the intro below moving the full piece to a separate article to try to shorten the copy in the newsletter itself.
Let me know what you think on the content and new style!
Bret Taylor and Clay Bavor | Acquired
🎥 Watch the full episode here
📆 Published: Aug 18th 2025
🕒 Estimated Reading Time: 4 mins. Time saved: 60+ mins! 🔥
🏗️ The Compounding Technology Stack: Why AI Adoption Looks Vertical
Brett Taylor frames the current moment through an compelling lens: technology waves don't replace each other—they compound. Each new wave rides on the infrastructure built by previous ones, creating exponentially faster adoption curves.
"To put a PC on every desktop, you had to actually make a supply chain of PCs, lower the cost of chips," he explains. "We got to basically two billion PCs. Then the internet could ride on the coattails of the PC revolution. When smartphones came out, you had the internet already."
The numbers tell the story: The first website came online around 1991, but it wasn't until 2002—eleven years later—that 10% of the world used the web weekly. ChatGPT achieved the same milestone in just 25 months.
This isn't just about better technology. It's about infrastructure leverage. Where previous waves required building physical supply chains, distribution networks, and user education from scratch, AI companies can instantly tap into existing cloud computing resources, global internet connectivity, and billions of connected devices.
Key Takeaways:
Technology waves compound rather than replace—each builds on previous infrastructure
AI benefits from decades of internet, mobile, and cloud investment already in place
Adoption curves are becoming exponentially steeper with each successive wave
🧠 From Scarce to Plentiful: The Intelligence Revolution
Clay Bavor offers a first-principles way to understand AI's transformative potential: "What are you making plentiful that was scarce before, and how does that impact society?" ……….
Continued here and on the link below 👇️
Don’t worry - NO sub requirement, just making the newsletter more compact and the Pod Shots more easily searchable.
CLICK THE LINK BELOW TO CONTINUE READING 👇️
🎥 Watch the full episode here
That’s a wrap.
As always, the journey doesn't end here!
Please share and let us know what you would like to see more or less of so we can continue to improve your Product Tapas. 🚀👋
Alastair 🍽️.
Reply