Key Takeaways
- Podscan has been acquired by Audiohook, with founder Arvid Kahl staying on to lead the next phase.
- Matic reports 99.999% visual SLAM across 16,000+ homes and 500M sq ft, positioning perception before manipulation in home robotics.
- AppSumo's $46M sales data shows AI content tools more than doubling in a year while infrastructure software declines.
- Tony Dinh measured Pro 500's Ultrafast mode consuming 20% of a weekly quota in ~1.5 hours, calling it not worth it.
- Dharmesh shared a conversion contrast: SMB free-to-paid at 1–5% versus enterprise free pilots auto-converting at 0.00000000001%.
- Derek Thompson's article reports hosting guests at home dropped from 42% in 1975 to 12% in 2026, a ~70% decline.
- Noah Kagan's breakdown of AppSumo data shows customer acquisition at 36% and done-for-you content at 23% of current top categories.
1. Acquisitions and Product Milestones
- Podscan has been acquired by Audiohook, according to founder Arvid Kahl, who described the deal as closer to a merger than an exit and will stay on to lead Podscan into its next stage. Kahl noted Podscan's technology already powered part of Audiohook's platform, and that having a customer become the acquirer brought existing trust and alignment. — via 1
- Matic's founding team argues the home robotics industry has its order wrong and chose to build "eyes" (perception) before "hands" (manipulation). Their visual SLAM reportedly reached 99.999% across more than 16,000 homes, making it the only scaled, pure-vision, fully autonomous unsupervised home robot, covering 500 million sq ft and 400,000 miles; the next phase will focus on manipulation. Arvid Kahl, who cited the post, said Matic is unlike any vacuum he has owned and called it the best he has used, emphasizing it was not an ad, and praised the team for sharing openly. — via 1
2. AI Tooling Economics and Developer Experience
- Tony Dinh argued ChatGPT for Business is poor value: lower quota floors, fewer features, higher price, and no Ultrafast mode, leading him to cancel the account and switch to a personal one. In a separate test, he found Pro 500's Ultrafast mode consumed about 20% of the weekly quota in roughly 1.5 hours, which he judged not worth it, saying he would return to Opus 5.5. — via 1 2
- Dharmesh shared a conversion comparison: free SMB users typically convert to paid at 1%–5%, while enterprise free pilot "customers" auto-converting to paid at pilot end came in at 0.00000000001%. — via 1
- Noah Kagan analyzed $46 million in AppSumo software sales to rebut the idea that software is dead. Current top categories: customer acquisition 36% (outbound, SEO, social), done-for-you content 23% (AI writing, voice, video), AI agents and app builders 12%, business operations 11% (calendars, CRM), and infrastructure 6% (email, hosting, backend). He noted AI content tools more than doubled in a year while infrastructure declined, and said what is dying is not software but software you have to manage. His three recommendations for developers: sell outcomes not tools ("10 booked calls" over "outbound platform"); either help customers acquire customers or do the work for them, not something in between; and if the product consumes tokens, charge by usage. — via 1
3. Social Trends and Founder Reflections
- Derek Thompson's article "You Are No Longer Invited to Dinner" reports that Americans have largely stopped hosting at home: combining Putnam's DDB Needham survey with a new Data For Progress replication, the share hosting friends or family at home at least monthly fell from 42% in 1975 to 12% in 2026, a decline of about 70%. The piece rules out two comforting explanations—dinner did not simply move to restaurants (solo dining is rising and in-person socializing is down overall), and other social activities did not replace gatherings (exercise and travel increased, but exercise is mostly solo, and time at home has risen sharply since the 2010s). It offers four explanations: dual-income household busyness (leisure feels like work, and men did not take over social planning as women entered paid work, though total leisure time actually increased); intensive parenting (parents, especially fathers, spend much more time with children, crowding out adult hosting evenings); shrinking friend networks (close friendships are declining, most visibly among those without college degrees, hosting is becoming a luxury, and non-hosts skew independent and non-graduates, with the speculation that weaker social ties may weaken party attachment); and homes being too comfortable (Americans spent nearly two more hours per day at home from 2003–2022, much of it on screens and short videos). The article closes with a "Lump of Leisure" theory: technology did not make human work unnecessary but made other people less necessary for entertainment, and social leisure is finite and easily substituted; the author frames it as a coordination failure rather than a personal problem—hosting a dinner requires scheduling, cleaning, food, allergy considerations, and the risk of boredom, while TikTok requires nothing. A footnote adds that if everyone is a bit busier, coordinating gatherings becomes multiplicatively harder, so a small increase in individual busyness can cause a large overall social decline. — via 1
- Dharmesh recounted that after running HubSpot for about 15 years, he became a beginner again this year, building a digital twin called Hal by hand with various AI tools, coding late nights and weekends, frequently making mistakes and feeling frustrated. He argued he must practice AI himself before he has the standing to advise founders to restructure their companies around AI. He also relayed Tom Brady's remarks at HubSpot UNBOUND: growth only comes from leaving your comfort zone, and Brady, with seven championship rings, is still seeking discomfort, so others have even less excuse—if AI makes you a little afraid, jump in. — via 1
- Alex Lieberman described the year after selling Morning Brew as one of the hardest of his life: the company had occupied his identity, fear-driven motivation vanished after the deal, and the post-exit struggle was hard to share because it looked like a "champagne problem," leaving him lonely, with sudden free time becoming a cage without direction. He drew three lessons: fear-driven motivation is powerful but short-lived, so he shifted to growth-oriented drivers like curiosity, growth, and play; time freedom does not equal happiness, and people need to create or have a clear creative purpose; and define your "genius zone" and values, and decide when money stops converting into happiness. — via 1
- Alex Lieberman also shared a post about engineers restructuring their approach as the frontier changes, which argued for deleting large numbers of tests that merely restate code; treating PRs and code review as more ritual than substance except for critical parts like payments and auth, letting models write code, test, review, and prove it with end-to-end demos; moving work to the cloud as a second unlock; using models as teammates for monitoring and releases; and no longer fearing performance optimization and large-scale refactors once there is closed-loop verification. — via 1
