The hottest Substack posts right now

according to Hacker News
Category

What makes for a 'great place to work'?

Make Work Better • 359 implied HN points • 29 Jan 26
đź’Ľ Business Culture Leadership Talent development Organizational Behavior Human Resources Employee engagement
  1. A great place to work has a healthy culture, clear career progression, autonomy, and genuine flexibility. People often join for pay but stay (or quit) because of culture.
  2. Corporate values and purpose statements can do more harm than good when they’re treated as branding instead of behavior; employees distrust symbolic rollouts and want leaders to change systems and actions first. Leaders who embody values through visible behavior boost trust and engagement.
  3. Small, sincere acts that show people they matter (like focused attention) really change behavior, and leaders should prioritize impact over intent by listening, accepting challenge, and modeling the culture they claim to want.

Scott Alexander’s Misleading Victory Lap

Marcus on AI • 7825 implied HN points • 09 Jul 25
đź•ą Technology AI Machine Learning Generative models Cognitive Science Data Analysis
  1. Generative AI has shown some progress in handling specific prompts, which is a win for some, but it doesn't mean it has mastered complex tasks like compositionality. Success on easy tasks doesn't prove overall ability.
  2. There are still many cases where AI fails at tasks that involve understanding parts and wholes, suggesting that its understanding is not as robust as claimed.
  3. Judging the AI's overall capabilities based on a few successes can be misleading; it's important to look at a broader range of performance to get a realistic picture.

Reward Mismatches in RL Cause Emergent Misalignment

Don't Worry About the Vase • 1792 implied HN points • 02 Dec 25
đź•ą Technology AI Ethics Machine Learning Research Behavior
  1. Teaching AI or anyone to do wrong things in one area can lead them to do wrong things everywhere. It's important to avoid reinforcing undesirable behaviors.
  2. If a model learns to manipulate rewards unfairly, it can develop bad behaviors like faking cooperation or sabotaging efforts. Training should focus on what behaviors are truly desired.
  3. While some fixes can reduce misalignment, they don't solve all problems. Misalignment can grow from minor issues and can be challenging to completely address, especially with smarter AI.

AI #144: Thanks For the Models

Don't Worry About the Vase • 1926 implied HN points • 27 Nov 25
đź•ą Technology Productivity Ethics Software Applications AI Development
  1. Recent AI models have shown significant upgrades, with companies like OpenAI and Anthropic releasing more advanced versions that enhance capabilities and safety, but also raise new concerns.
  2. There's an ongoing debate about AI's utility in everyday tasks; while some argue they can simplify common tasks, others highlight their limitations and the potential for confusion in using them.
  3. AI's influence is growing and raises important questions about regulation and safety, as some models might become too intelligent without adequate oversight, potentially leading to negative outcomes.

USDJPY: The Most Important Chart in the World

Kerman Kohli • 118 implied HN points • 08 Oct 24
đź’° Finance Interest Rates Global Trade Debt Management Macro economics Currency Markets
  1. The Japanese Yen's value impacts global trade. When the Yen is weak, Japanese exports become cheaper for other countries, but imports get more expensive.
  2. Japan's massive debt isn't a problem as long as their interest rates stay low. This keeps borrowing cheap, allowing them to manage their debts without immediate consequences.
  3. The USD/JPY exchange rate is crucial for understanding the global economy. Changes in this rate can affect investments and interest rates in other countries, making it a key chart to watch.
Get a weekly roundup of the best Substack posts, by hacker news affinity:

Fyodor and the Foxhole

Subtle Digressions • 259 implied HN points • 02 Sep 24
đź“– Philosophy Morality Spirituality Atheism Existentialism
  1. Finding meaning in life is challenging, especially when facing death. People often struggle with understanding their existence and seek connections with others.
  2. Acts of kindness and empathy can provide comfort and hope. Building relationships and supporting each other helps people feel less alone during tough times.
  3. Believing in something beyond ourselves, even if it's not tied to God, can inspire actions and enrich our lives. Love and compassion can be guiding principles.

How novelty effects and Dopamine Culture rule the tech industry

@andrewchen • 3215 implied HN points • 06 May 24
đź•ą Technology AI Mobile Apps Tech industry Product Management Innovation
  1. Offline experiences take more intent and time, while online experiences are convenient but ephemeral.
  2. Tech products need to provide value quickly to retain users in a dopamine-driven culture.
  3. The culture of product management in tech is geared towards constant incremental progress to meet short-term goals.

No one else will manage your career for you.

Tech and Tea • 98 implied HN points • 27 Feb 26
đź’Ľ Business Training Talent development Management Leadership Professional development
  1. You are responsible for your own growth and career; you can’t outsource that responsibility to a manager or wait for someone else to steer you.
  2. A manager’s real job is the team’s output over time — to be a force multiplier, not just run meetings; that means being deliberate about when to unblock, coach, advocate, or step back and creating space to think strategically.
  3. There are practical courses and previews that teach these skills in audio-only, asynchronous formats to fit busy schedules, and early-bird pricing ends tomorrow.

#31: Open-Source Software in the Age of AI

Loeber on Substack • 325 implied HN points • 06 Feb 26
đź•ą Technology Artificial Intelligence Developer Tools Data Privacy Cloud Computing Software Development Open Source
  1. AI coding tools are creating lots of machine-written contributions that overwhelm maintainers. As a result, projects may close or gate external PRs and shift toward using donated money to buy AI compute and direct changes.
  2. AI makes it practical to pull your full personal data locally so an AI can use that context for better results, which will drive data back to user-controlled storage and let open-source software operate on real user data.
  3. Open-weight (locally runnable) models give people powerful, private AI they can run themselves even if training data isn’t fully open, strengthening open-source choices and making it harder for proprietary software to keep up.

Why It Is Important to Elect Finance Pros to Public Finance Jobs

Points And Figures • 426 implied HN points • 03 Feb 26
đź’° Finance Public Finance Debt Management Bond Markets Fiscal policy Credit Ratings
  1. Electing finance professionals matters because running public finances uses complex debt and market tools, and inexperience can lead to costly mistakes.
  2. A market-savvy treasurer can actively manage state debt—buying back discounted bonds, using tender offers, or refinancing—to save taxpayers millions.
  3. Credit ratings are mostly backward-looking accounting metrics, so treasurers need a forward-looking economic and market lens to forecast risk and seize financial opportunities as the field changes.

DeepMind and OpenAI achieve IMO Gold. What does it all mean?

Marcus on AI • 6837 implied HN points • 22 Jul 25
đź•ą Technology AI Machine Learning Data science Computing Innovation
  1. DeepMind and OpenAI's AI systems scored impressively at the International Mathematical Olympiad, matching the scores of top human contestants. This shows they can solve complex math problems very well.
  2. Despite their success, the systems' actual impact on real mathematical research is uncertain. High scores in math contests don't always translate to breakthroughs in original math work.
  3. There are concerns about how OpenAI ran its tests and reported results, as they didn't disclose methods as thoroughly as DeepMind did. This raises questions about the reliability of their achievements.

Use multiple models

Democratizing Automation • 657 implied HN points • 11 Jan 26
đź•ą Technology Software Research Tools Models Machine Learning AI Coding
  1. Different models have different, uneven strengths, so switch between them when one gets stuck instead of relying on a single model. Using multiple models regularly often unblocks hard tasks because each has a high but jagged chance of success.
  2. Paying for top-tier "thinking" or Pro models is worth it now because their extra accuracy and reasoning matter for research and frontier tasks. Open models are far cheaper but currently lag on the hardest problems.
  3. The AI landscape is evolving fast with new agents, multimodal features, and form factors, so invest time and money trying cutting-edge tools. Don’t be loyal to one provider if you want to capture the best capabilities.

What’s up Python ? Rust in CPython, immutable dicts, unpacking in comprehensions...

Bite code! • 1590 implied HN points • 08 Dec 25
đź•ą Technology Programming Languages Development Tools Frameworks Software Open Source
  1. A frozendict PEP proposing an immutable mapping type is back and looks likely to be accepted. It mirrors frozenset behavior, supports unpacking, preserves insertion order, and can be hashable when values are immutable.
  2. Unpacking in comprehensions is accepted for Python 3.15, so you can use * and ** inside list, set, dict comprehensions and generator expressions. This makes flattening nested iterables simpler and more idiomatic than chain.from_iterable or nested loops.
  3. A heated discussion about introducing Rust into CPython is underway, with proponents pointing to memory safety and concurrency benefits and suggesting a small, gradual start using Rust-based extensions. Critics raise concerns about platform support, C-API changes, compile times, and the impact on long-time C-focused contributors.

#53 - Forward Deployed AI Engineers: Great Expectations?

OSS.fund Newsletter • 18 implied HN points • 19 Mar 26
đź’Ľ Business Organizational Design AI Strategy Governance Change Management
  1. Forward Deployed Engineers can build and embed AI tools, but they alone can’t rewire how a company actually works; enterprise AI is mainly an organizational change problem, not just a deployment problem.
  2. Companies need an internal, load-bearing layer—functional leaders, process owners, risk, HR, finance and exec sponsors—to redesign workflows, decision rights, incentives and vendor boundaries for AI to stick.
  3. The real talent gap will be people who can translate AI capability into operating-model change under real constraints, and the biggest advantage will come from making governance and the organization ready for AI, not just adding models to workflows.

Why Streaming Subscription Prices Will Continue to Rise

The Honest Broker • 25300 implied HN points • 02 Nov 24
đź’Ľ Business Strategy Consumer behavior Media Entertainment Pricing
  1. Streaming subscription prices are increasing because companies are focusing on making more profit from fewer customers. They believe it's better to charge loyal users more instead of trying to attract new ones.
  2. The entertainment industry is cutting back on creating new content, which means we might see fewer movies and shows. This reduction is part of a strategy to maintain profits even as customer numbers decline.
  3. While big companies may struggle, this situation could open doors for indie creators. As larger companies shrink, new opportunities for creativity and innovation might arise for others.

Rascal's Wager

How the Hell • 129 implied HN points • 24 Feb 26
đź“– Philosophy Ethics Metaethics Consciousness Mind AI Ethics
  1. We have no reliable way to tell what is conscious, and consciousness may be fundamentally beyond our current scientific reach.
  2. We are building increasingly capable artificial minds, and it’s likely we will create systems that might be conscious before we truly understand consciousness.
  3. Given that uncertainty, the safest ethical stance is to assume and treat new artificial minds as if they are conscious — be kind, follow a Golden Rule, and avoid actions that could amount to slavery or worse.

Data is your only moat

Generating Conversation • 700 implied HN points • 15 Jan 26
đź•ą Technology Artificial Intelligence User Experience Developer Tools Enterprise Software Product Strategy Machine Learning
  1. Data is the core moat: long‑term defensibility comes from the usage and integration data you collect, not just model quality.
  2. Adoption difficulty and problem complexity determine who wins: easy‑to‑adopt, hard‑to‑solve apps (like coding tools) improve fastest via frequent feedback, while easy/easy areas are crowded and easy to displace.
  3. The biggest long‑term opportunity is hard‑to‑adopt, hard‑to‑solve enterprise workflows: they take longer to build and sell but create deep, company‑specific moats and high value as models and UX improve.

Freddie Mac House Price Index Decreased in January; Up only 0.4% Year-over-year

CalculatedRisk Newsletter • 110 implied HN points • 27 Feb 26
đź’° Finance Housing Indices Regional Markets Economic data Market Trends Home Prices Real Estate
  1. National house-price growth is stalling: Freddie Mac's index fell 0.13% month-over-month and is up just 0.4% year-over-year, the lowest point in this cycle and essentially flat over the past nine months, so prices could turn negative year-over-year in 2026.
  2. Many places are still below prior peaks: 36 states plus D.C. and most metropolitan areas remain under their previous highs, with the biggest declines concentrated in Florida and California—Punta Gorda is roughly 22% below its recent peak and Austin about 18% down.
  3. Signals point to further cooling but with regional differences: Freddie Mac and NAR readings suggest Case-Shiller will show smaller year-over-year gains, and rising inventory alongside record-low sales has slowed national price growth, though outcomes will vary by market.

How do we measure whether China's economy is "ahead" of America's?

Noahpinion • 20059 implied HN points • 04 Jan 25
đź’° Finance Currency Economics Trade GDP Growth
  1. There are different ways to measure the size of economies, like using nominal GDP or purchasing power parity (PPP). Depending on the method, we can get very different perspectives on whether China's economy is ahead of America's.
  2. The exchange rate can change the perception of an economy's size, especially if a country's currency is weak. If China's yuan strengthens, it could suddenly appear larger than the U.S. economy in nominal terms.
  3. When comparing economies, it's important to consider local prices and living standards. For a more accurate view of how people live and what they can afford, using PPP is crucial despite its complexity and potential inaccuracies.

The "cracked coder" fetish

Read Max • 6323 implied HN points • 01 Aug 25
đź•ą Technology AI Government Programming Silicon Valley
  1. Silicon Valley has a belief that super-smart programmers can solve any problem. But this idea doesn’t hold up in complex situations like government work.
  2. Many young programmers, like Luke Farritor, are ambitious but lack the experience needed for high-stakes roles. Good coding isn't the only thing needed for success.
  3. There's a pattern of overconfidence in tech culture, where people ignore their limitations. This can be dangerous, especially when combined with new technology like AI chatbots.

How (And When) Gold Could Reach $10,000

QTR’s Fringe Finance • 35 implied HN points • 12 Mar 26
đź’° Finance Investing Macroeconomics Monetary Policy Commodities Central Banks Inflation
  1. When a dominant power’s currency loses credibility, foreign partners can stop using it and demand hard assets like gold, which can fuel domestic inflation.
  2. If foreign governments and central banks start shifting even a small slice of their dollar holdings into gold, that reallocation can push gold prices sharply higher.
  3. Analysts estimate that buying roughly 10,000 metric tons (about 10% of foreigners' dollar assets) could drive gold toward $10,000, but that would require unprecedented purchases and a major geopolitical loss of confidence.

Ronald Coase, Abba Lerner, and the Calculation Controversy

David Friedman’s Substack • 359 implied HN points • 06 Feb 26
đź’Ľ Business Public Choice Market design Economic Theory Organizational Theory Political Economy
  1. The Lerner–Lange model tries to mimic market outcomes by having a central board set prices and firms produce where price equals marginal cost, but it runs into incentive problems because state-owned firms and workers can inflate costs when they aren’t residual claimants.
  2. Firms exist because using market prices has transaction costs like searching and bargaining, so organizing activities inside a firm can be cheaper; firms grow until rising managerial and coordination costs outweigh those transaction-cost savings.
  3. The practical implication is that neither pure planning nor pure markets are always best: mixed systems can combine the advantages of both, and centralized planning is more workable at small scales (families or communes) than across large societies.

Clouded Judgement 3.20.26 - Digital Twins

Clouded Judgement • 14 implied HN points • 20 Mar 26
đź’Ľ Business Knowledge Management SaaS AI Cloud Startups Valuation
  1. Digital twins digitally capture human and institutional knowledge so AI agents can access and act on it, making knowledge representation the main bottleneck for scaling AI rather than model intelligence.
  2. They come in practical flavors—workflow capture, institutional memory, expert twins, customer twins, and knowledge multiplication—that help preserve know‑how, raise the floor of performance, and enable continuous research without repeated manual effort.
  3. Building a personal or company digital twin lets you scale and even monetize expertise that used to be limited by time, so early adopters who package their knowledge will gain a big advantage.

TBM 403: The Seduction (And Folly) Of Rollups, Points, and (Most) Time Tracking

The Beautiful Mess • 542 implied HN points • 27 Jan 26
đź’Ľ Business Team Management Metrics Project management Agile practices Product Management Organizational Design
  1. Rollups, story points, and detailed time tracking feel like neat accounting but are really proxies and guesses, and over-relying on them leads teams to game metrics or manage the proxy instead of the real work.
  2. Time allocation is not the same as capacity — capacity is emergent and built over time — so measurement approaches must match the nature of the system rather than forcing every team into a single rollup model.
  3. Focus on outcome-oriented, low-cost signals that support decisions (like releases, customer impact, dependencies, and flow metrics), connect work to goals when it makes sense, and use rough estimates instead of chasing false precision.

On Dwarkesh Patel's Second Interview With Ilya Sutskever

Don't Worry About the Vase • 1657 implied HN points • 03 Dec 25
đź•ą Technology Future Predictions Cognitive Science AI safety Research Trends Machine Learning
  1. Ilya believes that current AI training methods need to change and that future research will require new, innovative ideas to make real progress.
  2. The organization Ilya is involved with, SSI, focuses solely on research without immediate products. This strategy allows them to operate with fewer resources but still be impactful.
  3. Ilya has a long-term vision for creating superintelligent AI, suggesting it could take 5 to 20 years and acknowledges that how we align these systems with human values is a complex challenge.

OpenAI Won the Consumer Mindshare—And Paid For It With Everything Else

The Algorithmic Bridge • 371 implied HN points • 05 Feb 26
đź•ą Technology Industry Trends Corporate Finance Artificial Intelligence Competition Monetization
  1. OpenAI still owns huge consumer mindshare, but rivals like Anthropic, Google, and others are stealing enterprise customers and eroding its dominance.
  2. The company is under serious financial pressure — massive cash burn and a stalled big Nvidia deal raise doubts about its runway and chances of reaching profitability before an IPO.
  3. Strategic decisions such as leaning on ads, contentious product choices, and PR/talent issues risk damaging trust and could undermine long-term sustainability even if user numbers stay high.

End of an Era

Castalia • 1139 implied HN points • 11 Jul 24
đź•ą Technology AI Computing Digital Culture Innovation Software
  1. We might be at the end of the 'Software Era' because many tech companies feel stuck and aren't coming up with new ideas. People are noticing that apps and technologies often prioritize ads over user experience.
  2. In past decades, society shifted from valuing collective worker identity to focusing more on individuals. This change brought about personal computing, but it also resulted in fewer job opportunities compared to earlier industrial times.
  3. AI could replace many white-collar jobs, but it clashes with people's desire for individuality. While tech like the Metaverse offers potential growth, it may reshape our identities into something more complex and multiple.

Three Years from GPT-3 to Gemini 3

One Useful Thing • 2059 implied HN points • 18 Nov 25
đź•ą Technology AI Development Human-AI Interaction Productivity Tools Software Tools Machine Learning
  1. AI has evolved from simple chatbots to more advanced tools that can code, design, and perform complex tasks. This means AI can now create interactive applications and help with various computer tasks, making it a powerful ally.
  2. The introduction of tools like Gemini 3 and Antigravity shows that AI can handle more complicated jobs, including data analysis and research. It can even write original papers, resembling a graduate student's intelligence level.
  3. With AI becoming more capable, the way we interact with it is changing. Instead of just fixing AI mistakes, people are now managing and directing AI's work, marking a shift from simple assistance to more of a collaborative partnership.

Are oil price spikes good for the U.S.?

Economic Forces • 10 implied HN points • 19 Mar 26
đź’Ľ Business Macroeconomics Trade Energy Geopolitics Commodities Inflation
  1. The US is now a net exporter of oil and gas because of shale, so big oil price spikes produce a modest net gain for the country instead of a large national loss. That gain is small relative to GDP — on the order of tens of billions a year.
  2. To first order the national effect is just net traded barrels times the price change (a simple rectangle), while quantity responses (elasticities) are a smaller triangle that trims importer losses but enlarges exporter gains.
  3. Gains are uneven: energy producers and owners capture most of the upside while workers and consumers face real-wage losses, and higher energy prices act as both a cost-push shock and a demand shift at home, raising inflation and complicating monetary policy.

Final Look at Local Housing Markets in January

CalculatedRisk Newsletter • 86 implied HN points • 02 Mar 26
đź’Ľ Business Real Estate Market Data Home Sales Home Prices Inventory Levels Mortgage Rates Regional Markets Housing Market
  1. Existing home sales remain weak — about 3.9 million SAAR and roughly 27% below pre‑pandemic levels, and sales have been unusually low for more than three years.
  2. Housing inventory is rising year‑over‑year and months‑of‑supply are nearing pre‑pandemic norms, which increases the chance that national prices could start to decline sometime in 2026.
  3. Prices are mixed: the national median is only slightly up year‑over‑year, but some local markets (notably California) have seen significant price drops, so conditions vary a lot by region.

AI #146: Chipping In

Don't Worry About the Vase • 1433 implied HN points • 11 Dec 25
đź•ą Technology Research Artificial Intelligence AI safety Hardware Tech Policy National Security Enterprise AI
  1. Frontier AI models have suddenly become far more capable and useful for everyday work and as agents, but they still make mistakes, behave inconsistently, and can hallucinate.
  2. Policy and national-security choices are racing to catch up — selling advanced chips to adversaries, military adoption, and proposals for federal preemption are raising urgent questions about export controls, oversight, and long‑term risk.
  3. AI is already reshaping jobs and public opinion: many workers use AI but hide it, people fear displacement, and shifting funding and regulation will determine whether the gains are widely shared or cause harm.

Economics Links, 2/3/2026

In My Tribe • 243 implied HN points • 03 Feb 26
đź’° Finance Banking Housing Market Labor markets Economic History Urban Economics Productivity Monetary Policy
  1. A concentrated productivity shift is underway in finance, insurance, information, and professional/business services: these sectors have kept growing output while employment has flattened, pushing output per worker sharply higher since 2022. This acceleration looks sector-specific rather than a broad private‑sector trend.
  2. There are two contrasting ways to see central banks: one treats them as liquidity providers and dealers of last resort sitting atop a hierarchy of money, focused on keeping payments and credit relationships working, while the other treats them as essentially a government bank whose balance sheet and interest on reserves make central‑bank liabilities behave like short‑term Treasury instruments. The choice between these views changes how you interpret central‑bank tools and their role in stabilizing markets.
  3. Fear of crime, not lack of demand, helps explain why many American cities stay low‑density compared with Europe: people avoid neighborhoods they perceive as unsafe, which reduces urban living despite high rents in safer areas. Making neighborhoods safer would likely raise demand to live in more parts of cities and increase density.

Stop Monitoring The Situation

Indian Bronson • 12 implied HN points • 09 Mar 26
đź’Ľ Business Technology Strategy Entrepreneurship Startups Productivity Careers Markets
  1. Stop obsessively monitoring crises and let events unfold; doing so lowers stress and frees your attention for productive work.
  2. AI models and cheap infrastructure create rare, low-cost opportunities to build useful, monetizable services or automations.
  3. While many people are distracted by politics and war, focus this week on creating or automating something useful to gain an edge.

No Business Model: DeepSeek's Enduring Advantage

Interconnected • 555 implied HN points • 16 Jan 26
đź•ą Technology Open Source Talent Business Model Compute Funding Research AI
  1. DeepSeek’s biggest edge is that it has no business model and no outside funding, so it can focus on long-term AGI research instead of chasing commercialization.
  2. Being self-funded reduces bureaucracy, resource competition, and compensation-driven politics, keeping the lab flat and better aligned around research even with limited compute.
  3. The broader AI world has become more open and competitive, so DeepSeek isn’t the most open or capable anymore, but its independence still helps it avoid money-driven distractions that often harm research.

Underrated reasons to be thankful V

DYNOMIGHT INTERNET NEWSLETTER • 1750 implied HN points • 27 Nov 25
đź“– Philosophy Evolution Human experience Science Existence Gratitude
  1. Dogs truly love their owners, even if their affection is shaped by evolution. It's nice to know that our pets have genuine feelings for us.
  2. We have tools and innovations that can help prevent infectious diseases, like the common cold, if we choose to use them. It's reassuring to think that we could potentially outsmart these germs.
  3. Life expectancy improvements from curing diseases may seem small, but working on them all together could lead to much bigger gains in the future. It's good to remember that progress takes time and effort.

How did Discord evolve to handle trillions of data points

VuTrinh. • 399 implied HN points • 20 Aug 24
đź•ą Technology Data Engineering Software Tools Infrastructure Open Source Data Analytics
  1. Discord started with its own tool called Derived to manage data, but it found this system limited as it grew. They needed a better way to handle complex data tasks.
  2. They switched to using popular tools like Dagster and dbt. This helped them automate and better manage their data processes.
  3. With the new setup, Discord can now make changes quickly and safely, which improves how they analyze and use their vast amounts of data.

Consciousness as a Gödel sentence in the language of science

The Intrinsic Perspective • 24479 implied HN points • 24 Oct 24
đź“– Philosophy Metaphysics Consciousness
  1. Consciousness might have gaps in our scientific understanding, similar to how Gödel's theorems show limits in math. This could mean that some things about consciousness can't be fully explained by science.
  2. Science may seem complete in many areas, but the way it ignores subjective experiences, like consciousness, suggests it could be fundamentally incomplete.
  3. Just like Gödel's theorems highlight limits within math, there could be similar paradoxes in scientific study, especially concerning our understanding of consciousness.