Lütke describes Shopify as a company already reorganized around agent AI and says that judgment, taste, and human responsibility will matter more in the future.

A clear primer on bipolar transistors, from NPN and PNP structure to biasing, saturation, package types, and common amplifier configurations.

Gregor Hohpe argues that a good architect reduces risk and complexity rather than imposing answers. His measure is simple: help others make better decisions.

A network engineer explains how certificates, internships, contract work, and salary negotiation helped him move from $12 an hour to a $120,000 offer.

Sutton argues that generative AI imitates well but discovers nothing. To reach artificial creativity, evaluation, selection, and shared goals are needed.

Why an old Go fan says the language is betraying its identity, and why he now sees Odin as the more coherent alternative.

Google’s $80 billion raise became a fight over AI’s real economics: revolution, hype, or both. The speakers argue over capex, adoption, and who captures the next layer of value.

Strauss Zelnick defends AI as an efficiency tool, but shrinks its creative claims: successes, he says, still come from taste and risk.

ThePrimeagen argues that people who truly love programming should stay close to the code. He also describes how AI changed the way he builds and enjoys work.

Ravikant y tres fundadores describen cómo la IA está desplazando el valor del código al juicio, de la salida a la fábrica que la produce.

Dax Raad, cofundador da Open Code, diz que a IA torna as equipes mais rápidas, mas não produz produtos melhores automaticamente.

Matuschak argues that code agents can break app silos and unleash new invention in interfaces. The risk, he says, is that technical culture keeps slowing what these tools promise to enable.

Theo Browne attaque Google sur l’IA, le terminal et le cloud : modèles chers, CLI défectueuse, confiance interne brisée.

Andreessen defends police tech, attacks wealth taxes, and explains why he thinks AI is the biggest productivity wave of his career.

She left Google with four offers in hand and explains why tech now rewards timing, referrals, and AI adaptation.

Sanfilippo explains why passing tests are not enough and why LLMs could become virtual QA. He also offers a critical reading of the Banne rewrite in Rust.

LeCun explains why he sees LLMs as a dead end for general intelligence and is betting on world models, Tapestry, and more local AI.

Nish spiega perché HydraDB punta sul contesto, attacca i vector database e racconta come un messaggio brutale abbia sbloccato la crescita.

Graham explains why founders should spend time in Silicon Valley and why Stockholm may benefit if they return.

Iñigo Medina argues AI will layer onto software rather than replace it, and that companies need history, not hype, to adapt.

Theo argumenta que o HTML está virando um formato melhor que o Markdown para agentes: mais legível, interativo e útil nos fluxos de trabalho.

Harvey’s CEO explains how AI is changing legal work, and why he thinks human judgment remains the decisive advantage.

White says technical training is not wasted, and that AI rewards people who adapt rather than worship programming.

Roslansky lê o mercado de trabalho como um reset de competências, não como uma simples destruição de vagas. Estes são os cargos que, segundo ele, realmente crescem.

Anthropic, OpenAI, and the Pentagon: the video tells the story of the fight over who controls military AI and where human responsibility ends.

DeepSeek v4 Flash runs locally on 128 GB, but Sanfilippo shows that the real bottleneck is prefill, not the answer.

Palihapitiya links AI, inequality, and platform power in one thesis: the central issue is who controls attention, money, and narrative.

Karpathy spiega perché vede l’AI come un nuovo paradigma di programmazione, non solo un aumento di velocità. E sostiene che il futuro dipende da verifiabilità, agenti e comprensione umana.

Ravikant argues that AI is already reshaping companies, hardware, and war, while the sector concentrates in a few hands.

गैलोवे एआई के सर्वनाश वाले नैरेटिव को चुनौती देते हैं और उसे मूल्यांकन, असमानता और सत्ता से जोड़ते हैं। उनके लिए असली सवाल यह है कि लाभ कौन समेटता है और लागत किस पर पड़ती है।

Fireship рассказывает о vulnerability in the Linux kernel, discovered by an AI agent, already exploited, and useful as a case study in AI-tool marketing.

Anthropic, Microsoft e Uber segnalano, secondo il commentatore, che l’AI sta entrando in una fase più costosa e meno permissiva. L’utilità resta, ma il conto economico diventa centrale.

Why a badly chosen shard key can skew data and traffic, and when hashing beats fixed ranges. Ben Dicken explains it with concrete examples.

Spiegel explains why durable consumer products are rare, how Snap tries to defend itself from copying, and why Specs is his bet on the next computer.

OpenAI’s miss, the AI capex boom, and the Musk lawsuit all point to the same bottleneck: power, not demand, is now deciding the market.

Zenz and Roner explain why agents are speeding up work but also complexity, and why Pi tries to keep the human in the loop.

Naval explains how coding agents are changing software, venture capital, and the iPhone — and why he thinks pure software is losing value.

Huang sostiene che i TPU non minacciano Nvidia perché l’azienda vende una piattaforma di calcolo più ampia, non un chip per l’IA.

Un playbook per avviare un business AI da zero: nicchia stretta, prodotto rapido, primi clienti subito, capitale solo se serve.

Ferriss argues that AI multiplies clarity and chaos at the same time, and that the real issue remains choosing the right game.

Bob Lazar and Luigi Vendittelli discuss the new film about S-4 and revive the old question: how much of the story is verifiable, and how much is still faith?

Sanfilippo explains why DeepSeek v4 Pro could be the first open-weights model really close to closed systems, and why price remains the decisive problem.

Flickering is a minimalist Arduino-based courtesy light, presented as a design gadget and offered in pre-sale.

Kleppmann explains why cloud and AI do not make systems simpler, but instead change which trade-offs you need to understand.

Antonio says AI has made polished coding tutorials less useful, so he is shifting to workshops that show debugging, architecture, and judgment.

Sanfilippo argues AI is already reshaping programming, but art still depends on human taste, selection, and judgment. He sees today’s worst failures in music, audiobooks, and slop, not in creativity itself.

Sanfilippo argues that in coding agents, the brand of the model matters less than the ability to switch it often. His thesis: try everything, pay for a short time, don’t fall in love with anyone.

Huang defends Nvidia’s moat: ecosystem, supply chain and CUDA. And he argues that cutting off China would weaken the United States.