AI's Tipping Point: Agent Swarms, True-to-Life Video and Collapsing Costs
Raoul Pal and Emad discuss why new-generation models crossed a threshold; research data confirms the break in mathematics, forecasting and creative output.
IN A CONNECTED WORLD
Connect the dots.
Multiple perspectives. One place.
Your curiosity, your daily feed.
How it works
For you enables ranking by explicit follows and saved stories; marked-as-read history is used only with reading memory enabled. Last 48 hours first, then the last 7 days. Latest turns personalization off.
Your curiosity starts here
Kolibri 78B: Open-Weight AI You Actually Own
Aleph Alpha opened Kolibri at 78.1B parameters under Apache 2.0; 3.46B go active per token, aimed at regulated institutions.
The Completeness Axiom: Consequences That Make the Real Line Whole
Rational numbers miss lengths like the square root of two; the completeness axiom closes those gaps. This piece unpacks the chain of reasoning from a five-and-a-half-hour Calculus lecture.
Five Levels of Emotional Freedom: A Stoic Guide to Steady Relationships
A five-level path from phone-checking anxiety to calm self-possession, built on Stoic dichotomies and modern psychology research. Each level pairs a recognizable scene with one concrete weekly action.
OpenAI Revenue Reset Spooks AI Stocks as Nasdaq Lags
Wall Street split on October 8 as an OpenAI revenue accounting debate sank AI-linked stocks, leaving the Nasdaq down 1.3% while the Dow edged higher and oil surged.
Haiku 5.5 Redraws Cost Efficiency for Small Models
Anthropic launched Haiku 5.5 at 10 cents input per million tokens up to 100,000 tokens, scoring 72.4 on OSWorld 2.1 and beating rival Luna.
The quiet habits behind charisma: five mental shifts
Charisma grows less from appearance than from mental habits: dropping self-monitoring, holding silence, building confidence from evidence, releasing the need to convince everyone, and steering yourself under pressure form the core of inner authority.
Same Sticker, Different Bill: Haiku 5.5 vs Luna in 12 Runs
Ugur Keskekci gives the same three jobs to Haiku 5.5 and GPT-6 Luna; list prices match, yet the invoice and clock of 12 runs rewrite the cost logic of both models.
Haiku 5.5: Anthropic finally fixes its small-model problem
Theo's review finds Haiku 5.5 doubling its predecessor while matching Luna's short-context price; computer use and agentic coding tilt toward Anthropic.
When Machines Claimed a Millennium Prize Problem
On 8 September 2026 OpenAI announced a machine-checked solution to the Navier-Stokes existence and smoothness problem, one of the seven Millennium Prize challenges. The claim, built by a vast agent fleet in days, collided with a prior unpublished result by two mathematicians and ignited a fierce dispute over credit and method.
How Warmth and Competence Signals Shape First Impressions and Listening
Vanessa Van Edwards shows that first impressions settle before a word is spoken and that warmth-plus-competence cues carry the whole exchange, in an 18-point communication map completed by listening and strong delivery techniques.
Seven Quiet Signs of Genuine Love: How Devotion Shows in Daily Life
Ghana-based broadcaster Jessica Os argues that a woman's love is revealed through actions rather than words, listing seven behaviors that range from opening her social circle to listening with full attention.
Seven Stocks in a Selloff: Valuation Discipline and a Quiet Winner
With the market near records, Broadcom and KLA have fallen hard; the host ranks seven stocks on cash, multiples and expectations, crowning Waste Management shares at the top.
Acemoglu vs Anthropic: Who Decides What AI Does to Work
Nobel economist Daron Acemoglu urges worker-centered AI while Anthropic maps three economic paths to 2030, from mild gains to disruptive boom.
8 GitHub Projects to Go Viral and Make Money: From Launch Videos to Agent Teams
Greg Isenberg walks through eight open-source projects for coding agents — launch videos, customer research, voice cloning, slide decks, agent teams, free-tool marketing, idea interrogation, and design language. Each comes with a concrete try-it step and a way to turn it into income.
The Six Rungs of Money Freedom: From Paycheck to Lasting Impact
Starting from the gap between a 55,000-dollar salary feeling and the cash truly left in the bank, this piece opens a plain six-rung freedom frame and pairs each rung with a clear target and a small action.
Ten Oscillators Meet Thirty Years of SPY: Williams %R Takes the Robustness Crown
A rule-based test runs ten popular oscillators over more than thirty years of SPY data under identical rules. QSRSI posts the highest headline profit factor, yet Williams %R wins on the full robustness package.
How Hormones Reshape ADHD: Cycle, Pregnancy and Menopause
Estrogen fluctuations change ADHD symptoms and medication response across the month; new research is bringing the picture around the cycle, pregnancy and menopause into focus.
From Robot Brains to Nervous Systems: 3 Stocks Leading Physical AI
The MoneyFlows show splits the robotics and physical-AI stack into three layers and picks one stock per layer: Nvidia for the brain, Analog Devices for the nervous system, Keysight for testing. This article weighs the show's claims against reported earnings and current market research.
Math 2.0: When AI Published 722 Proofs Overnight
OpenAI releasing hundreds of AI-generated mathematical results overnight reignited Terence Tao's Math 2.0 warning. Host Alexander examines Tao's digestion and community thesis alongside the Hacker News debate.
Number One in Seven Hours: Inside an AI-Powered Google Ranking System
Julian Goldie shows a fresh article hitting number one in both classic Google results and generative AI answers within seven hours; the method rests on Search Console gap hunting, original fieldwork, and strict human review.
Stop Chasing Approval: Steady Confidence and Bonds That Hold
The narrator favors purpose, calm trust, and full presence over constant contact and jealous questioning. Rooted in Stoic thought, the video maps a path from approval-seeking to a language of respect.
Agents that reason about causes, not just correlations
Bareinboim shows fluent models failing counterfactual probes and argues agents must be built on causal world models.
Three Chip Stocks Still Cheap Inside a Record-Setting Index
The S&P 500 keeps setting records while its forward P/E sits at a one-year low; the balance sheets of Micron, Marvell and SanDisk show how that squeeze works in practice.
Among the noise, meaningful connections.