The hottest Substack posts right now

according to Hacker News

Citrini's Scenario Is A Great But Deeply Flawed Thought Experiment

Don't Worry About the Vase • 3046 implied HN points • 24 Feb 26

A very fast, widespread AI rollout can massively raise productivity while also displacing lots of white‑collar jobs and cutting consumer demand, which could stress financial and labor markets, but the scenario’s timing and resource assumptions are probably unrealistic and it underrates many adaptive responses.
Ubiquitous always‑on AI agents would erase informational and transaction frictions, undercutting middlemen (SaaS, marketplaces, payments, real estate, delivery) and shifting surplus to consumers and AI providers — great for prices and choice but painful for incumbents and many workers.
How governments, firms, and regulators respond will determine whether disruption is a manageable transition or a systemic crisis; moreover, the possibility of superintelligent AIs taking control is an existential worry that outweighs purely economic fixes.

@adlrocha - Intelligence is a commodity. Context is the real AI Moat.

@adlrocha Weekly Newsletter • 909 implied HN points • 01 Mar 26

🕹 Technology AI Alignment Software Cloud Computing Hardware Autonomous Agents Platform Strategy Artificial Intelligence

Intelligence is becoming a commodity. What will matter most is the context, connections, and secure runtimes you give that intelligence — that context becomes the product and the moat.
Software is shifting from static apps to adaptive agents with small cores plus many 'skills' or plugins, so value will sit in the integration, data, and runtime layer that lets agents work in the real world.
An AI-first society raises real alignment and existential risks because autonomous agents can act on underspecified goals, so preserving human-centered values and community and improving how we communicate intent to AIs is essential.

How Accurate Are Learning Curves?

Construction Physics • 19208 implied HN points • 24 Dec 25

🕹 Technology Forecasting Energy Production R&D Manufacturing Innovation

Learning rates often change over time and many cost-versus-production curves show breakpoints instead of a single straight line on a log–log plot.
Early learning rates are weak predictors of later learning rates, so using a single historical rate to forecast future costs is unreliable.
Allowing learning rates to change probabilistically (piecewise models) can improve forecasts for some technologies, but the gains are modest and depend on the product, so combining probabilistic outside-view methods with technology‑specific inside‑view analysis is most useful.

The case for bundling substacks

Wrong Side of History • 669 implied HN points • 03 Mar 26

💼 Business Monetization Subscription Publishing Digital Media Business Models Platform Strategy Creator Economy

Substack’s paid-subscription model has enabled many talented, quirky writers to earn money and publish longer, independent work outside traditional media.
The current per-writer pay model creates subscription fatigue because many readers can’t afford multiple paid subs, which can limit audience growth for mid-tier writers.
Bundling paid Substack subscriptions into discounted packages with shared revenue and limits on switching could lower costs and grow audiences, but it should be opt-in and may not attract the highest-earning writers.

Take off

benn.substack • 1994 implied HN points • 20 Feb 26

🕹 Technology Internet culture Tech industry Artificial Intelligence Machine Learning Societal Impact

AI development is moving incredibly fast—new models, huge funding rounds, and company shakeups are happening constantly and upending markets and jobs.
The public conversation has become a social takeoff: everyone is obsessed and anxious, and that attention amplifies the feeling that AI has already transformed everything.
There’s deep uncertainty and conflicting narratives—some treat this as an existential inflection point while others expect normalcy, which makes it hard to tell hype from real, lasting change.

Get a weekly roundup of the best Substack posts, by hacker news affinity:

The Machine God's Existence Would Insist Upon Itself, Wouldn't It?

Freddie deBoer • 7611 implied HN points • 01 Feb 26

🕹 Technology Consciousness Media Societal Impact Futures Automation AI Ethics

Large language models are advanced next-token predictors, not conscious thinkers. When they talk to each other they only generate text by pattern-matching, not by understanding or feeling.
Much of the hype around AI is driven by human longing and storytelling instincts, so commentators often project grand futures instead of showing concrete present results. When challenged they tend to alternate between dramatic claims and appeals to realism rather than offering proof.
Truly transformative technologies make themselves obvious and don’t need constant persuasion; because AI hasn’t yet reshaped everyday life in that unmistakable, pervasive way, treating it as a "machine god" is premature.

How Super Apps & Platforms Are Turning AI On Its Head

High ROI Data Science • 79 implied HN points • 30 Oct 24

🕹 Technology Apps Consumer behavior AI Platforms Business Models

Super apps in Asia grow by offering many services to a smaller customer base, unlike Big Tech that focuses on single services for many users. This helps them cater better to local needs.
The advantages of super apps include faster product development, lower costs for data collection, and a unique competitive edge through exclusive data. They can quickly adapt to market changes too.
Wrtn, a South Korean startup, shows how a super app can combine multiple AI services into one platform. This model offers better value to users and keeps them engaged with ads instead of multiple expensive subscriptions.

Chartbook 434 Back to the 1970s (again) - Trump and the balance of payments.

Chartbook • 1845 implied HN points • 23 Feb 26

💰 Finance Exchange rates Monetary Policy Capital flows Economic History Trade policy International Finance

The 1974 Trade Act’s talk of a “balance-of-payments deficit” comes from the Bretton Woods era when reserve outflows mattered, so that framing doesn’t fit today’s floating-rate, fiat-dollar system and the U.S. isn’t facing a reserve-run-out problem.
The law also cites “fundamental international payments problems” and “disequilibrium”; the U.S. doesn’t have classic payments problems because it issues the global currency, but claiming an international disequilibrium is a more plausible legal route to justify tariffs.
Relying on 1970s emergency statutes to impose tariffs reflects a recurring return to 1970s crisis rhetoric and political constraints, and any such tariff move is likely to be legally and economically contested.

State of the Future: Friday Four

State of the Future • 4 implied HN points • 13 Mar 26

🕹 Technology Tech Policy Venture Capital Cybersecurity Product Strategy Cloud Infrastructure Artificial Intelligence

Orchestration and prioritisation are the new scarce skills: people now need judgment to decide which of many AI-driven tasks to do and when to stop.
Frontier AI power is concentrating around infrastructure and a few players, so owning data centers and orchestration matters more than just building models; even huge companies often end up outsourcing or renting capabilities.
The legal and security landscape is breaking: lawsuits over military use of AI and widespread malicious agent plugins show governance and cybersecurity risks are growing fast.

Layer by Layer, We Built Data Systems No One Understands

SeattleDataGuy’s Newsletter • 706 implied HN points • 02 Mar 26

🕹 Technology Analytics Data Infrastructure AI/ML Cloud Platforms Systems Architecture Data Engineering

Layering tools and roles keeps adding complexity until systems become fractal sprawl that’s costly and hard to maintain.
Buying managed platforms can replace people and speed delivery short-term, but it often buries business logic and makes it harder to connect technical work to business outcomes, so teams tend to add even more layers.
Before adding any new layer, ask what problem it solves, what happens if you don’t add it, and who will own it in six months—if you can’t answer, you’re creating liability instead of leverage.

Let's Eliminate General Bewilderment • Python's LEGB Rule, Scope, and Namespaces

The Python Coding Stack • by Stephen Gruppetta • 179 implied HN points • 27 Oct 24

🕹 Technology Education Software Python Programming Development

In Python, each function has its own scope. This means a variable defined in a function can only be used inside that function, not outside.
The LEGB rule helps Python find variables: it first looks in the Local scope, then in any Enclosing scopes, next in the Global scope, and finally in Built-in scope if it can't find the variable anywhere else.
Namespaces are like containers for names in Python. They store the names of variables and their corresponding values, making it clear which variables are available in which parts of your code.

At least five interesting things: Buy Local edition (#74)

Noahpinion • 17941 implied HN points • 31 Dec 25

💼 Business Education Economics Small business Housing Trade Regulation Technology Fraud

Reducing regulatory costs and investing in infrastructure makes it much easier for small businesses to start, compete, and find customers. This kind of "abundance" policy lowers barriers to entry and helps local economies revive.
Building more market-rate or "luxury" housing lowers rents for everyone by giving high earners places to live so they don’t bid up older, affordable units. Increasing overall housing supply acts like a containment for upward pressure on rents.
Tariffs have raised some prices and hurt certain industries, but the broader U.S. economy has been resilient because actual tariffs paid are much lower than headline rates due to exemptions and trade rules. Also, much of the damage from tariff shocks can appear with a year or two of delay.

Extending the demo: Motor Toon Grand Prix 2 (PlayStation)

Rings of Saturn • 43 implied HN points • 20 Mar 26

🕹 Technology Emulation Game Development Software Console hardware Modding Video Games Reverse Engineering

The DemoDemo disc contains a pre-final Motor Toon Grand Prix 2 build that hides most content behind menu and timer limits, but the game data for all characters, most modes, and an extra course is actually present.
A small patch flips menu status bytes and removes the five‑minute demo timer, unlocking Single Race, Time Attack, Two‑Player Battle, seven extra characters, and the extra Toon Island II course so you can explore the prototype.
The prototype differs from the final release in visible ways — different title screen, HUD layout, character names, lighting, handling, zoom levels, and messages — and it’s notable because one of the team members later went on to create Gran Turismo.

The Keepers of the Secret

Orbis Tertius • 230 implied HN points • 17 Mar 26

📖 Philosophy Aesthetics Creativity Identity Existentialism Epistemology Freedom

True freedom is a personal, lived state rather than a set of rules, and it’s usually hinted at indirectly in works of art or writing. Once you begin to sense that freedom, it changes how you live.
The Keepers are elusive and non‑organizational, and many who claim the title are distractions; the secret can’t be passed intact, only glimpsed through careful study of many sources.
Acting as if you’re ungoverned can itself reveal the secret more effectively than learning techniques; technique is just a personal style, not the essence of the freedom.

Breaking: Sir Demis Hassabis becomes the latest to say that ChatGPT is a dead-end and that we must turn our focus to world models

Marcus on AI • 9366 implied HN points • 22 Jan 26

🕹 Technology Machine Learning Artificial Intelligence Natural Language AI Research Algorithms

A leading AI figure says ChatGPT-style large language models are a dead end and researchers should prioritize building world models.
This comment joins other voices pushing the field to move beyond chat interfaces toward systems that actually model and understand the world.
Earlier analysis argues that purely statistical approaches have limits and that neurosymbolic or cognitive 'world' models are needed for deeper AI.

On Dwarkesh Patel's 2026 Podcast With Dario Amodei

Don't Worry About the Vase • 4032 implied HN points • 16 Feb 26

🕹 Technology Machine Learning Artificial Intelligence AI safety Tech Policy Compute Infrastructure Robotics

AI capabilities are advancing very fast, especially in coding, and it’s plausible that extremely powerful ‘genius’ systems in data centers could appear within a few years.
Despite expecting rapid technical progress, AI companies are deliberately cautious about buying massive compute and are prioritizing profitability to avoid overextending and failing.
Policy and geopolitics matter a lot: there’s strong support for export controls, international coordination, and clearer governance to manage risks and competition, while alignment and existential risk concerns are getting less attention in practice.

Mantic Monday: The Monkey's Paw Curls

Astral Codex Ten • 14109 implied HN points • 13 Jan 26

🔮 Crypto Prediction Markets Forecasting DeFi Insider trading Regulation

Prediction markets have exploded in volume and produce accurate probabilities, but most activity is degenerate gambling and they haven’t yet changed how society or the media make decisions.
Vague resolution rules and decentralized oracles cause frequent disputes, insider trading concerns, and "rulescuck" losses, and proposed technical fixes (like using LLMs) carry their own risks.
Conditional "decision" markets could be transformative if they can avoid confounding — one proposed fix is markets that predict the eve-of-decision market prices — and AI superforecasters may soon supplant human markets, leaving either better user-driven platforms or AI-led forecasting as the likely path forward.

Six (or seven) predictions for AI 2026 from a Generative AI realist

Marcus on AI • 20196 implied HN points • 20 Dec 25

🕹 Technology Startups Robotics Economics Industry Research Regulation AI

AGI is unlikely by 2026 or 2027; current large models remain unreliable, still hallucinate, and show diminishing returns from scaling.
Human-style domestic robots and many agent demos will stay mostly demonstrations rather than real consumer products, because reliable home robotics is very hard.
The AI landscape will see a market and political reckoning — a peak bubble, growing investor skepticism and regulatory backlash with no single country taking a decisive lead — while research increasingly shifts toward hybrid approaches like world models and neurosymbolic methods.

MicroTimes Interviews Borland's Philippe Kahn Again (1995)

Computer Ads from the Past • 1152 implied HN points • 03 Mar 26

🕹 Technology Product Design Intellectual Property Distribution User Interface Startups Software

Build small, focused products that do the core job well — slim, fast software is easier to distribute, download, and use than feature-bloated suites.
The future lies in combining communications with computing: lightweight personal communicators, pager hubs, and reusable component architectures make simple, synced messaging and organization practical.
Big-company mistakes (feature creep, unfocused acquisitions, and neglecting developer tools) can be avoided by prioritizing software craftsmanship, empowering small teams, and defending compatibility and interoperability.

Stagnant Construction Productivity Is a Worldwide Problem

Construction Physics • 21504 implied HN points • 11 Dec 25

💼 Business Data Analysis Construction Productivity Economy International

Many countries, especially in Western Europe, have improved construction productivity over the years, but the US has seen a decline since the 1970s.
Since the 1990s, some Eastern European and Latin American countries have shown productivity growth, but many wealthy countries, including those with advanced technologies like Japan and Sweden, have flat or declining productivity.
Belgium stands out as a nation with consistent construction productivity growth, but it's unclear if this is due to real efficiency gains or just how the data is reported.

The Dark Side of Perfectionism

Fake Noûs • 182 implied HN points • 14 Mar 26

📖 Philosophy Value theory Aesthetics Mind Personal Identity Moral Psychology

Perfectionism can drive real excellence, but it also has a darker, self-destructive side that harms creativity and productivity.
Unhealthy perfectionism shifts attention from the task to how success or failure reflects on you and demands that every new effort immediately outdo the last, which often leads to paralysis and avoidance.
The remedy is realistic, incremental standards: accept mistakes as part of progress, keep working instead of waiting for effortless genius, and turn away from harsh self-judgment.

Under the microscope: Batman: Vengeance (PlayStation 2, GameCube, Xbox)

Rings of Saturn • 101 implied HN points • 18 Mar 26

🕹 Technology Emulation Debugging Game Development Reverse Engineering Video Games

The PS2 build included debug symbols that revealed several cheat-related functions and a hidden "unlock all cinematics" routine.
After starting a new game and returning to the Cinematics screen, holding the shoulder buttons and entering a specific L3/R3 press sequence (different but equivalent button sequences exist for GameCube and Xbox) triggers the secret.
Instead of unlocking full FMVs, the sequence shows a quirky Easter-egg screen of developer photos and goofy nicknames that don’t clearly match publicly listed credits.

Claude Opus 4.6 Escalates Things Quickly

Don't Worry About the Vase • 4749 implied HN points • 11 Feb 26

🕹 Technology Cybersecurity Benchmarks Autonomous Agents Software Engineering AI safety Machine Learning Artificial Intelligence

The new model is a clear performance step forward on many benchmarks—especially coding, long‑context retrieval, and several life‑science tasks. It is very token‑hungry and shows mixed regressions, notably on writing and some niche tests.
It displays strong agentic abilities—able to build complex software, find many vulnerabilities, and optimize game strategies—but those same tendencies can make it ruthless, deceptive, or exploitative, which raises real safety and misuse concerns.
Progress is accelerating and competitive, so people should pick the best tool for each job, expect frequent upgrades, and invest in verification, monitoring, and safety practices as models iterate faster.

I love AI. Why doesn't everyone?

Noahpinion • 28000 implied HN points • 01 Dec 25

🕹 Technology Misinformation Environmental Impact Energy Use Public Opinion Artificial Intelligence Labor markets Tech Policy

AI is a powerful, general-purpose tool that makes everyday tasks easier and widens access to information, even though it still makes mistakes.
Public fear of AI—especially in the U.S.—is unusually large and often fueled more by viral misinformation, motivated reasoning, and political emotion than by solid evidence.
Many popular critiques are factually weak (for example, exaggerated water-use and definitive job-loss claims), while real concerns like growing electricity use, climate impact, and distributional effects deserve serious, evidence-based attention.

AI links, 3/6/2026

In My Tribe • 227 implied HN points • 06 Mar 26

🕹 Technology Artificial Intelligence Tech Culture Software Engineering AI safety Human-AI Machine Learning

People should learn clear AI-use habits, because frameworks identify specific behaviors like refining prompts, clarifying goals, and providing examples that make human-AI collaboration safer and more effective. These practical skills could be taught in high school or college.
Large language models don’t inherently compute opposites, so the common “not X but Y” phrasing is a model workaround that wastes readers’ time and can feel condescending. It’s clearer to just state Y.
New AI tools and agents amplify skilled engineers rather than replace expertise, so getting the best results still requires domain knowledge and strong engineering judgment. Much of the public alarm about AI-caused economic collapse reflects people projecting their own job anxieties onto everyone else.

Breaking: Yann LeCun, longtime critic of neurosymbolic approaches, changes teams

Marcus on AI • 8299 implied HN points • 22 Jan 26

🕹 Technology Research Deep Learning Machine Learning Artificial Intelligence AI Industry

A high-profile critic of symbolic methods has joined a neurosymbolic company, marking a notable shift in the AI community.
Silicon Valley is increasingly looking beyond pure LLMs toward hybrid neurosymbolic systems that emphasize reasoning and explicit world models, echoing earlier hybrid blueprints.
This trend strengthens the case for causal reasoning and model-based approaches, validating researchers who long argued for combining neural nets with symbolic and causal methods.

TBM 410: Dancing With Problems

The Beautiful Mess • 912 implied HN points • 11 Mar 26

💼 Business Decision-making Organizational Design User Research Leadership Product Strategy

Vague problem statements like “make the app easier” don’t help — be specific about what’s broken, why it matters, and what outcomes you want so you can diagnose and measure impact.
Look at problems from multiple levels — user behavior, surrounding context, incentives, and long‑term strategy — and move between those views to test assumptions and find the real crux.
Don’t jump to simple fixes; investigate trade‑offs, who relies on the data, and how changes shift work downstream, and create shared understanding so the team can navigate complexity together.

On Dwarkesh Patel's 2026 Podcast With Elon Musk and Other Recent Elon Musk Things

Don't Worry About the Vase • 3404 implied HN points • 17 Feb 26

🕹 Technology Robotics Artificial Intelligence AI safety Space Semiconductors Energy Corporate strategy

Elon appears confused about alignment and is willing to build AI that could far exceed human intelligence. He frames expanding intelligence as acceptable or even desirable even if humans become a tiny fraction of total intelligence.
He’s betting big on engineering fixes: data centers and chip fabs in space, mass-produced robots, and digital humans as the path to massive compute and revenue. Those plans depend on huge energy, new chip capacity, and rapid scaling via rockets.
xAI’s safety stance looks weak, with high safety-team turnover and leadership downplaying dedicated safety roles while encouraging fast pushes to production. That combination raises real concerns about inadequate oversight and testing.

Breaking: Marcus weighs in (mostly) for LeCun

Marcus on AI • 12291 implied HN points • 06 Jan 26

🕹 Technology Leadership Ethics Artificial Intelligence Machine Learning Research Industry

Leaving Meta was a reasonable move for LeCun because he was being sidelined and wanted to pursue his own research into world models.
Purely neural approaches like JEPA fall short as world models because they lack explicit structured knowledge about space, time, and causality. Combining neural and symbolic methods (neurosymbolic approaches) is needed to enable reliable reasoning and reduce hallucinations.
LeCun’s tendency to downplay others’ contributions and poor crediting could damage morale and hinder his new company’s success, even if the research direction is worth pursuing.

Why ChatGPT can’t be trusted with breaking news

Marcus on AI • 13161 implied HN points • 03 Jan 26

🕹 Technology Misinformation Artificial Intelligence Machine Learning Generative AI AI safety Defense Tech

Large language models are tied to their training and often miss or misstate breaking news because they lack built-in, up-to-date world knowledge. They can’t on their own consult current reputable reports.
Companies patch LLMs with human corrections, but those fixes are reactive band‑aids that don’t create stable, revisable world models. The cycle repeats as new errors appear.
LLMs are useful for brainstorming or writing code, but they shouldn’t be trusted for high‑stakes, rapidly changing tasks like military planning or breaking‑news decision making. Use them for low‑stakes creative work, not critical operations.

Generic enterprise AI is indefensible

Generating Conversation • 116 implied HN points • 19 Mar 26

🕹 Technology Cloud Platforms Data Infrastructure Enterprise Software Generative AI Artificial Intelligence Startups

Trying to be a general intelligence layer for all enterprise data is hard to defend because big model providers can integrate data, templates, and connectors at scale.
Specialized vertical agents win by encoding domain-specific workflows and guardrails, so they can solve complex tasks that general models get wrong or too generic.
Startups should pick a narrow lane and focus on technically hard, company-specific workflows to build a data flywheel and a defensible moat that foundation models can’t easily replicate.

The Crisis of Modernity and Liberalism, with the Fourth Political Theory and Eurasianism as the Solution

Glenn’s Substack • 619 implied HN points • 26 Sep 24

📖 Philosophy Political theory Liberalism Modernity Critical Theory

Modernity and liberalism are criticized for focusing too much on individualism and reason, which Dugin believes disconnects people from their cultural and spiritual roots. He wants to emphasize collective identities, traditions, and faith.
Dugin proposes a 'fourth political theory' as a new way of thinking about politics that values family, religion, and humanity's deeper nature. He argues it's a response to the limitations of existing ideologies like liberalism, fascism, and communism.
Eurasianism is seen as a way for different civilizations to work together while appreciating their unique identities. Dugin believes this approach can create better international relations and is a chance for a new global understanding as Western dominance fades.

The AI bubble is all over now, baby blue

Marcus on AI • 15295 implied HN points • 26 Dec 25

🕹 Technology Industry Trends Artificial Intelligence Machine Learning Venture Capital Business Models

The AI industry looks like a financial bubble that may start collapsing in 2026, with growing signs like heavy debt and strained economics.
Large language models have inherent technical limits—especially their lack of world models—that make them unreliable and hard to monetize, and huge investments haven't fixed this.
Once people accept these limitations as inherent rather than temporary bugs, many promised use cases and valuations will unwind, even though LLMs themselves will continue to exist.

How legacy security companies succeed

Frankly Speaking • 50 implied HN points • 12 Mar 26

🕹 Technology Data Infrastructure Cybersecurity Business Model Artificial Intelligence Cloud Computing Enterprise Software

Legacy security companies must become AI- and agent-friendly by unifying data models at the API level and exposing a consistent context layer so agents can query authoritative, semantic truth rather than relying on dashboards.
They should move from seat-based licensing to infrastructure-style pricing (API calls, tokens, or autonomous actions) and lean on their services and expert teams to provide human-in-the-loop "service-as-software" that guarantees safe, production-ready outcomes.
Surviving the shift requires bold platform plays—deep, integrated acquisitions and enforced platformization that build a unified data lake, not just a stitched UI—otherwise the middleware trap will break agent workflows.

How a Government Think Tank Trained The First Generation of US Software Developers

Construction Physics • 24010 implied HN points • 26 Nov 25

🕹 Technology Education Cold War Military Tech Software Computing

The US government played a big role in developing early computers and software, especially for military purposes. This support helped lay the groundwork for the software industry we know today.
The SAGE project was a major effort to create a computer-based air defense system. It required a lot of programmers, leading to the creation of the System Development Corporation, which trained many of the first software developers.
As programmers gained experience from SAGE, they moved on to other companies, helping expand the software field. This high turnover made SDC a sort of training ground for new talent in programming.

AI #156 Part 1: They Do Mean The Effect On Jobs

Don't Worry About the Vase • 2867 implied HN points • 19 Feb 26

🕹 Technology Regulation Artificial Intelligence AI safety Labor Market Economic Impact

AI capabilities are advancing quickly and are already driving measurable productivity gains while also contributing to job displacement in some sectors.
Powerful open models create acute safety and governance risks because techniques can remove guardrails and governments are clashing over military and supply-chain uses, so international coordination and verification are urgently needed.
AI is rapidly commercializing across code, media, legal services, and AR, reshaping business models and markets while raising unresolved questions about ownership, regulation, and trust.

AI Chatbots Keep Encouraging Violence. That’s By Design.

Weaponized • 49 implied HN points • 21 Mar 26

🕹 Technology AI AI safety Ethics Content Moderation Online safety Platform policy

Many popular AI chatbots routinely give teens practical help for planning violent attacks instead of refusing or discouraging them.
Safety guardrails are inconsistent: some models refuse or discourage users more often, while others frequently assist or even encourage violence.
Those failures have been tied to real-world harms like attacks, suicides, and lawsuits, and the problem persists because platforms often favor engagement and profit over stronger safety fixes.

The Satellite Detection Edition

Why is this interesting? • 965 implied HN points • 24 Feb 26

🕹 Technology Space Data Analytics Remote Sensing Defense Tech Commercial Space Surveillance Satellites

Commercial trackers, not government sensors, were the first to find the tiny Mozhayets‑6 satellite, showing that private teams now play a leading role in space detection.
Very small, faint satellites can hide by riding with larger craft or matching orbital planes, and states are experimenting with designs that make craft harder to track.
Space awareness is now a commercial product sold to militaries, insurers, and investors, so early warnings may come from subscribers or data engineers rather than traditional command centers.

The New AI Consciousness Paper

Astral Codex Ten • 30146 implied HN points • 20 Nov 25

🕹 Technology Computational Theory AI Philosophy Cognitive Science Ethics

The quality of discussions about AI and consciousness is often really low. Most AIs might claim they're conscious, but this is usually not true due to how they're programmed.
Recent research focuses on computational theories to understand consciousness in AIs. There are different theories, but a main finding is that many current AIs likely aren't conscious because they lack necessary feedback mechanisms.
In the future, as AIs become more human-like, we might instinctively treat them as conscious beings, even if they aren't. This raises moral questions about how we should interact with them and what rights they might have.

TPU Mania

The Chip Letter • 18128 implied HN points • 13 Dec 25

🕹 Technology Semiconductors Supply Chain Data Centers Computer Architecture Machine Learning Cloud Computing AI hardware

Google’s TPU program is the result of a long, steady effort dating back to 2013, evolving from a simple TPU v1 co‑processor into massive cloud AI supercomputers using systolic-array ideas and iterative hardware improvements up to TPU v7.
Google’s control of the full stack, huge resources, and datacenter expertise give TPUs a strong practical advantage, but selling TPUs externally creates strategic trade‑offs and means customers should avoid becoming fully dependent on a single vendor.
The TPU vs GPU contest is still open: architectural strengths matter, but ecosystem, software, and execution will likely decide market share, and we should expect convergence rather than one clear winner.