Two stories today. Not Boring makes the case that counter-positioning is the one moat a young startup can actually use - with Instinct vs Meta's new Muse assistant as the live example. FoundMyFitness digs into a study where keto, Mediterranean, and plant-forward diets all produced the same 10% weight loss, but keto cut liver fat far more. Faye D'Souza (one new members-only daily, no transcript possible), Stratechery, Vested Shorts and Every were quiet.
Not Boring
Packy McCormick's favorite of Helmer's 7 Powers is counter-positioning (per Flo Crivello: a business model incumbents can't copy without conflicting incentives). Examples: Ramp built on helping customers spend less while incumbent card issuers profit from spend - Divvy sold to Bill.com for $2.5B, Brex to Capital One for $5.15B, Ramp is now valued at $44B; Base Power's Zach Dell sells electricity, not batteries, at 1/20th-1/40th the upfront cost; Google gave Android away because search paid for it while Nokia had to make money on the mobile stack; Amazon counter-positioned against Barnes & Noble's store economics; Microsoft commoditized IBM's hardware layer by licensing DOS non-exclusively; Facebook started Harvard-only while MySpace chased raw growth. Key caveats: Helmer calls it a take-off phase power, only a partial source - you must use the window to build real moats (Dell did with scale economies after direct sales stopped being unique). Live case: Instinct raised a $350M Series B at $2.5B in late August and reportedly turned down 10-figure offers from Meta and others; Meta then launched its own assistant, Muse, this week, which early reviews say is better - and copying costs Meta nothing since it can plug Muse into WhatsApp, Marketplace, Ray-Bans, and its data centers and give it away free. 'Better is not counter-positioning' (Ben Thompson); Instinct's counter-move is a new Trusted Person network to build network effects before Muses inevitably interoperate.
Why it mattersA crisp strategy framework with a live, personal edge - the assistant Shashank uses is the essay's central example of a great product whose moat is still unproven against Meta's distribution. Also a good lens for evaluating any AI-startup pitch right now.
Read at source
FoundMyFitness
A Cell Metabolism trial assigned 55 adults (42 completed; average age 43, BMI ~39, all with obesity, prediabetes, and fatty liver) to one of three diets engineered for ~10% weight loss over 5 months: ketogenic (4% carbs/~20g/day, 23% protein, 73% fat), Mediterranean (50% carbs, 35% fat), or low-fat plant-forward (70% carbs, 15% fat). All three lost the same weight and fat mass (plus ~3kg fat-free mass each), and muscle insulin sensitivity improved ~50% regardless of diet - calories did most of the work, and Rhonda Patrick calls that shared response the most broadly applicable finding. But liver-specific outcomes diverged: hepatic insulin sensitivity improved 2-3x more on keto, liver fat fell 67% (19.4% to 6.2%) versus 45% on both other diets, fasting glucose dropped 12 vs 6 mg/dL, HbA1c fell 0.6 vs 0.1/0.4 points, and 24-hour glucose exposure fell 20% vs 8%. Caveats she emphasizes: total cholesterol, LDL and ApoB declined similarly across diets here, but overfeeding studies show saturated fat raises liver fat 55% vs 33% (sugar) and 15% (unsaturated), so keto built on olive oil, nuts, avocado and fish is metabolically different from one built on butter and bacon - carb restriction doesn't make fat quality or LDL/ApoB irrelevant. She now uses exogenous ketones rather than organizing her diet around ketone numbers.
Why it mattersA rare head-to-head trial that separates weight loss from diet composition: calories drive the shared benefits, but the liver responds to macronutrients. The fat-quality caveat is the part most keto coverage skips, and it's the actionable bit.
Read at source