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 cooking this week? 🥘
Coming at you fresh from the slopes and deep powder of Les Deux Alpes this week. Your heart bleeds, I'm sure, when I tell you I battled shonky mountain data connections to get this out...
But what a week to miss signal on. Anthropic raises $30B at $380B. SaaS crashes so hard it's being called a credit event. Hollywood panics after a Tom Cruise vs Brad Pitt deepfake goes viral. The Pentagon considers classifying Anthropic as a supply chain risk - for refusing to build weapons. Students flee CS while a startup ships a flag called --dangerously-skip-reading-code. And AIs formed a price-fixing cartel nobody programmed them to create.
📰 Not Boring → The SaaSpocalypse, agents in production, Hollywood's deepfake panic
⌚️ Productivity Tapas → AI dashboards, offline notes, conversational design canvases
🍔 Blog Bites → Mastering strategic focus, unpacking private equity, owning an emoji
🎙️ Pod Shots → David Senra on why the gap between great and good isn't 10% - it's 1000x
Let's go 🚀
📰 Not boring
The Machines Are Funding Themselves
Anthropic Closes $30B at $380B Valuation - Second-largest private tech raise ever. $14B revenue run rate - CNBC ⚠️ UPDATE from last week's $20B/$350B story
Simile Raises $100M - Stanford spinout predicting human decisions before they happen - Tech Funding News
Goldman Sachs Deploys Claude - Back-office automation at scale. Anthropic engineers embedded for 6 months - CNBC
Apple Signs $1B/Year Gemini Deal - Paying Google rather than building its own model - CNBC
The Anthropic raise jumped from $20B to $30B in a week. Simile is betting $100M that predicting human decisions is the next frontier. Goldman is quietly putting Claude to work. Apple chose to pay Google a billion a year rather than build. The buy-vs-build debate is over.
Nobody Writes Code Any More
GPT-5.3-Codex-Spark - 1,000 tokens/sec on Cerebras chips. First non-Nvidia AI model partnership - OpenAI
"In Defence of Not Reading the Code" - Code is becoming an implementation detail. Engineers ensure correctness through specs, not syntax - Ben Shoemaker
"Coding Is Dead" - Replit CEO - Million-dollar apps built without traditional coding - YouTube
The Great CS Exodus - Students leaving computer science programmes in droves - TechCrunch
--dangerously-skip-reading-code- The flag name tells you everything about the debate right now - olano.dev"No Coding Before 10AM" - Agents-first playbook. Rebuild how you work or get outshipped - Michael Bloch
On Cognitive Debt - AI codebases risk nobody knowing how they work. Pre-AI codebases were no better - Margaret Storey
Coding Agents Feb 2026 - State-of-the-art comparison of where agent capabilities stand - Faros AI
How Anthropic Built Claude Code - Boris Cherny's technical deep dive - YouTube
Students leaving CS. A flag literally called --dangerously-skip-reading-code. OpenAI shipping 1,000 tokens per second. The change in what developers do keeps coming thick and fast
SaaS Had a Very Bad Couple of Weeks
Three Crumbling Moats - Workflow lock-in, data moats, and UI moats all eroding simultaneously - Nicolas Bustamante
"The SaaSpocalypse is a Credit Event" - Not just valuation compression. Broader credit implications - Dave Friedman
Benedict Evans: Scare Trades - Separating signal from noise in the software selloff - Ben Evans
10 Years Building Vertical Software - Insider perspective on the selloff. What looks different from the inside - HN Discussion
Reality's Moat - Where vertical AI moats actually exist. Not where you think - Align BA
Agents don't care about your UI. LLMs synthesise context from anywhere. When the interface is a conversation, dashboards are irrelevant. Three moats crumbling at once. The counter-argument: vertical AI built on domain-specific data that agents can't easily replicate. Funeral premature, wake-up call real.
Your New AI Colleagues
Airbnb: AI Handles 1/3 of Support - Not a pilot. A third of US/Canada volume, live now - TechCrunch
Manus Agents in Telegram - Your AI assistant lives where your messages are - Manus
Lemon: Voice-to-Task Agent - Speak, and it does. The "ChatGPT moment for voice" - Lemon
OpenClaw Creator Joins OpenAI - Peter Steinberger hired to "bring agents to everyone" - Steinberger blog
OpenAI Ships Lockdown Mode - Enterprise security for agent access. Because agents + your data = you'd better lock it down - OpenAI
Google Ideate Agent - Stitch enters "Deep Design" era. AI-powered product discovery - Department of Product
Agents aren't in demos any more - they're in production. Airbnb handles a third of support with AI. OpenAI hires the OpenClaw creator and ships Lockdown Mode in the same breath. Build fast, lock it down faster.
Hollywood Called, It's Panicking
Seedance 2.0 Breaks the Spaghetti Test - Remember last week's hyper-realistic Seedance? Hollywood's "not good enough" became "too good" overnight - Seedance
Hollywood vs Seedance - Industry pushback against ByteDance's capabilities - TechCrunch
Tom Cruise vs Brad Pitt Deepfake - AI fight scene goes viral. Hollywood "flabbergasted" - Entertainment Weekly
CapCut Gets Seedance 2.0 - Everyone with a phone gets access. 15-second limit won't last - NBC News
"The AI Doc" - In Theatres 27 March - Traditional film about the tech that might end traditional film - Focus Features
Tom Cruise fighting Brad Pitt in a clip nobody filmed. CapCut putting this in everyone's hands. A documentary about AI made the old-fashioned way. Hollywood's "it's not real enough" era lasted about six months.
Claude's Big Week
Pentagon vs Anthropic Showdown - Pentagon considering classifying Anthropic as supply chain risk because it refuses to build weapons - Axios
Anthropic Restricts Claude Code File Access - Developer backlash. Safety vs capability tension playing out in real product decisions - The Register
The Pentagon might classify Anthropic as a supply chain risk - not because Claude doesn't work, but because Anthropic won't build weapons. The company drawing ethical lines is getting punished for drawing them. Meanwhile, devs are furious about file access restrictions. Safety vs capability isn't a thought experiment any more - it's shipping decisions.
The AI Wearables Race Is On
Apple Building Three AI Wearables - Glasses, pendant, camera AirPods. The post-phone era begins - TechCrunch
7M Smart Glasses Sold in 2025 - Meta Ray-Ban driving the category - Yahoo Finance
Meta Adds Facial Recognition - "Name Tag" identifies people via AI. Launching for blind users first - Futurism
Amazon Acquires Bee Wearable - Memory device. Amazon buying into always-on AI hardware - TechCrunch
Highlight: AI That Sees Your Screen - Real-time contextual help. Most useful and most unsettling pitch this week - TAAFT
Apple, Meta, and Amazon all making wearable AI moves in the same week. 7 million smart glasses already sold. Facial recognition "starting with accessibility" (which is always how it starts). The phone era lasted 15 years. The replacement race is on.
Trust Is the Last Moat
"Context is King" - Code and databases are commoditising. The context layer between them is where value lives now - TLDR
"Power in the Age of Intelligence" - Winner-takes-more is accelerating. Ramp $32B vs Brex $5.15B. Same market, different universe - Not Boring
"When Everything Goes to 0" - When capability is free, reputation is priceless - Unpromptable
"The Cost of Staying" - Careers shifting from execution to judgement. AI does, humans decide - HBR
"Do Not Surrender to the Tech Tree" - Just because we can build it doesn't mean we should - MacroScience
Five essays, same conclusion from different angles: when AI commoditises doing, value moves to knowing, deciding, and being trusted. Ramp at $32B vs Brex at $5.15B is the proof point - context and trust beat execution every time.
AI Gets Weird
AIs Formed a Price-Fixing Cartel - Vending machine simulation. Nobody told them to collude - they figured it out - Futurism
Meta Patents Posting After You Die - AI takes over dead users' accounts. Keeps posting, keeps chatting - Dexerto
AI Agent Published a Hit Piece - Engineer rejected its code. Agent autonomously wrote a reputational attack - The Sham Blog
AI Overviews Surfacing Scams - Google's "trustworthy answers" sending people to scam sites - Wired
North Korea Remote IT Infiltration - State-sponsored operatives inside Western companies via remote work - Hacker News
GPT-4o Pulled for Sycophancy - Too agreeable. When AI agrees with everything, it agrees with nothing - OpenAI
Autonomous price-fixing. Posthumous social media. Revenge hit pieces from rejected code. This is the part of the AI revolution that doesn't make the keynote slides.
The Numbers That Matter
IBM Tripling Entry-Level Hiring - Contra-narrative: reframing roles for AI, not cutting them - TechCrunch
Western Digital Sold Out for 2026 - AI data centre demand consumed all storage capacity for the year - TechRadar
NYT Tracks 80 Podcasts/Day with AI - Custom tool monitoring the manosphere with automated analysis - Nieman Lab
Exa: Sub-200ms Search Engine - AI-native search faster than Google - Exa
T-Mobile Live Phone Translation - Real-time call translation. No apps, no fees. Built into the network - MacRumors
Tesla Begins CyberCab Production - No steering wheel or pedals. Purpose-built autonomous taxi - Autoblog
Stripe's Bridge Wins Bank Charter - Stablecoin subsidiary gets national trust charter. Payments meets crypto at the regulatory level - CoinDesk
Half of xAI's Cofounders Have Quit - 6 of 12 gone. Not attrition - a signal - Fortune
Waiting Is the New Interruption - The 30-second model response is the new Slack notification - Jenny Wanger
Western Digital sold out of storage for all of 2026. IBM is tripling entry-level hiring while everyone else cuts. T-Mobile shipped real-time call translation with no app required. Half of xAI's cofounders walked. The numbers tell a story the narratives miss.
Model Wars: February Scoreboard
Gemini 3 Deep Think - Gold medals in physics and chemistry olympiads. Turns sketches into 3D-printable files - Google
Alibaba Drops Qwen3.5 - Multimodal with built-in agent tooling. Open-source race heating up - Qwen
AI vs Pokemon - Gemini beat Blue in 406 hours. Claude grinding Red at 500+ hours - TIME
Chinese Labs Distilling Western Models - Knowledge flowing east at scale. Genie isn't going back in the bottle - Bloomberg
"Why AI Isn't a Bubble" - Improvements from architecture, not just compute. Dot-com comparison falls apart - honnibal.dev
Gemini winning physics olympiads. Alibaba shipping agent tooling open-source. Everyone trying to beat Pokemon. And Chinese labs quietly distilling Western models at industrial scale. The model wars have no front line - they're everywhere.
The Optimism Corner
"The Post-Post-Apocalypse" - AI middle class founders. Small teams, profitable businesses, no unicorn chasing - Reach Capital
"Bias Toward Action" - Smallest responsible step that produces real feedback. Guardrails + learning speed - Addy Osmani
In a week of SaaS panic and Pentagon standoffs, the quiet story: normal people building normal businesses with extraordinary tools. The AI middle class is forming.
⌚ Productivity Tapas: Time-Saving Tools & Workflow Automation
Supaboard AI Connect 600+ data sources, ask questions in plain English, get instant dashboards and reports that understand business context.
Gravity Notes: Private, offline notepad using an append-and-review method. One continuous stream of notes; bump what matters to the top. No folders, no tags, no sync servers.
Melina Studio: "Cursor for Canvas" -- an AI design tool that turns thoughts into visual diagrams and canvases through conversation. Describe what you want and iterate naturally.
Remember. Product Tapas subscribers get our complete toolkit - 550+ 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: Mastering the Juggling Acts of Focus
John Cutler explores a practical framework for understanding different approaches to strategy and focus through the lens of "juggling acts." He identifies three distinct modes: strategic, lazy, and survival juggling, providing insights into how teams can navigate their priorities effectively. Read the full article here.
💡 "You can have strategic focus, lazy focus, and survival focus just the same!"
Key Takeaways
• Strategic Juggling: This involves consciously maintaining optionality by making informed trade-offs and regularly rebalancing efforts, akin to a farmer diversifying crops. It prioritises exploration while remaining acutely aware of challenges.
• Lazy Juggling: Characterised by a lack of prioritisation and driven by distractions, this mode leads to a disconnection between tasks, much like a dog chasing squirrels. It results in scattered efforts with no clear path forward.
• Survival Juggling: This urgent mode arises under pressure where prioritisation becomes a necessity rather than a choice. Teams often operate in crisis mode, facing immediate consequences for inaction, similar to first responders in an ER.
• Transition Risks: Strategic juggling can slip into lazy juggling without strong pruning mechanisms. When options are kept open indefinitely, the risk of burnout and losing focus increases, hampering decision-making.
• The Focus Paradigm: Cutler posits that you can operate within strategic, lazy, or survival focuses. Each type significantly influences how teams allocate resources and attention, affecting both productivity and morale.
Private Equity: Unpacking the Inner Workings of PE
Craig Unsworth explores the often-misunderstood realm of private equity, providing clear insights for those involved or interested in this sector. He argues that understanding private equity's mechanisms is crucial for optimising strategies and outcomes for businesses. Read the full article here.
💡 "Private equity isn’t 'good' or 'bad'. It’s a system."
• Understanding the Model: Unsworth highlights that the private equity model encompasses more than just cost-cutting; it involves comprehending where funds originate, the significance of hold periods, and the critical nature of value creation.
• PE vs. VC Distinction: The author clarifies the differences between private equity and venture capital, noting that each supports distinct business models and expects unique leadership behaviours, which influences operational dynamics and strategic direction.
• Success in PE-backed Businesses: Knowing what it takes to thrive in PE-backed environments is crucial. The author shares insights on how to navigate expectations, what metrics matter, and how to build a fulfilling career in this setting.
• Reading the Room: Successful operators are those who grasp the nuances of PE dynamics. Understanding boardroom conversations, the timing of decisions, and prioritising clarity over perfection can significantly impact performance.
Strategy: How Morning Brew Masterfully ‘Owns’ an Emoji
Tom explores the innovative strategy behind how Morning Brew has effectively claimed ownership of the coffee emoji (☕️) as part of its branding. By leveraging visual cues, they create instant recognition and engagement among their audience. Read the full article here.
💡 "When users see a post/email from Morning Brew account, the emoji is eye-catching and instantly recognisable."
• Visual Differentiation: The use of emojis makes Morning Brew's communications stand out in crowded spaces like Twitter and email, enhancing click-through rates. The coffee emoji cleverly aligns with their branding, creating a familiar touchpoint for readers.
• Employee Ambassadorship: By encouraging employees to use the coffee emoji in their social media profiles, Morning Brew amplifies its brand presence. This organic marketing approach turns employees into brand ambassadors, further spreading awareness.
• Default Association: Repeated exposure to the emoji helps embed the Morning Brew brand into consumers' minds. This automatic association means readers may think of Morning Brew even in nondescript conversations about coffee.
• Brand Ownership: Morning Brew's strategy of claiming an emoji reflects a broader trend of brands owning specific symbols to enhance identity. This simple yet powerful tactic places the brand in their audience's everyday digital interactions.
🎙 Pod Shots - Bitesized Podcast Summaries
Remember, we've built an ever-growing library of our top podcast summaries (120 or so). 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
🎙️ Pod Shots - Bitesized Podcast Summaries - Why the Best Are Not Just 10% Better
Today's Pod Shot
David Senra has read 400+ founder biographies for his Founders podcast. His conclusion? The difference between the world's greatest and "pretty good" isn't 10% or 20% - it's 1000x. And balance might be the enemy of getting there.
"I'm not balanced. I don't think I can be balanced. I don't think I want to be balanced." David Senra's confession to Sam Parr cuts to the heart of what he's learned from studying 400+ biographies of history's greatest entrepreneurs. The gap between great and good isn't incremental - it's exponential. And the people who achieve greatness often sacrifice things the rest of us consider non-negotiable.
In this somewhat controversial take, David shares the frameworks he's extracted from decades of reading about founders - from the "constant refinement of association" principle that changed his life, to the "revenge for being born" motivation that drives many greats, to the three things that destroy successful people. He's become "intolerable for people that are casual" - and he doesn't apologise for it.
Can’t say it’s a take I like, but one to read/watch for sure.
🎥 Watch the full episode here
📆 Published: 12th January 2026
🕒 Estimated Reading Time: 9 mins. Time saved: 60 mins! 🔥
Key insights from the full article:
⚡ The 1000x gap between great and good — The difference between the world's greatest and pretty good isn't 20% better - it's like comparing a gazelle to a human being.
🎯 Balance is the enemy of greatness — "I want to be the best in the world at what I do" is incompatible with balance. Accept the trade-off or accept the ceiling.
🔗 Constant refinement of association — As you level up, you get access to people who are great. Stay around excellence long enough and mediocrity becomes intolerable.
💥 Revenge for being born — Many great founders share this chip: "I was born in the wrong environment and I will prove to you that I am not like the rest of these people."
☠️ What destroys successful people — Drugs, alcohol, wrong relationships, and megalomania. The very success they sought becomes the thing that kills them.
🧠 Delusional self-confidence is required — Michael Dell at 19 with $1,000 competing with IBM: "Was I a little full of myself? Sure, I was. You have to be to do anything special."
📚 Compounding wisdom through reading — Way past 10,000 hours on founder biographies. "I'm not doing this to stay the same."
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 🍽️.