The hottest Algorithms Substack posts right now

And their main takeaways
Category
Top Technology Topics

YouTube Shorts Averages Over 200 Billion Daily Views

The Algorithmic Bridge • 626 implied HN points • 24 Jun 25
🕹 Technology Algorithms
  1. YouTube Shorts gets over 200 billion views every day, showing how much time people spend watching them. But that high number raises concerns about how this attention is being used.
  2. Many tech creators focus on making money rather than improving people's lives. This leads to products that can harm instead of help, especially for younger users.
  3. There's a growing pushback against technology that doesn't benefit society. People are becoming aware and rejecting approaches that prioritize profits over well-being.

Beyond Snowflake: A Problem-Solver’s Guide to Distributed ID Generation

Software Bits Newsletter • 51 implied HN points • 04 Jan 26
🕹 Technology Algorithms
  1. Memory allocator patterns — like per-node caches, hierarchical range grants, batching, and prefetching — transfer cleanly to distributed ID generation and let services hand out unique IDs locally with almost no coordination.
  2. There is no one-size-fits-all ID strategy: slabs and hierarchical ranges give extreme throughput and B-tree locality at the cost of wasted IDs and weaker global ordering, consensus gives strict global ordering and durability but costs latency and availability, and Snowflake-style schemes sit in between.
  3. The best engineering move is methodological: spot a related solved problem, extract its core principles (hierarchy, locality, batching, prefetching), and adapt them while accounting for distributed realities like partial failure and unbounded latency.

Algorithmic ranking is unfairly maligned

DYNOMIGHT INTERNET NEWSLETTER • 1156 implied HN points • 23 Jan 25
🕹 Technology Algorithms
  1. Not all algorithmic ranking is bad. Some algorithms can be useful if they align with what you want to see and achieve.
  2. A lot of current algorithms are designed to keep you engaged and make money for the companies, not necessarily to help you find what you like.
  3. We need better control over these algorithms to ensure they serve our interests, possibly through new technology or structures that prevent companies from taking that control away.

Music algorithms failed us

The Rectangle • 84 implied HN points • 12 Dec 25
🎶 Music Algorithms
  1. Streaming made the whole world of music easy to access, but recommendation systems focus on keeping you listening rather than showing truly new or surprising music.
  2. Algorithms reduce taste to data and similarity, so they mostly suggest songs that sound like what you already listen to and create echo chambers instead of serendipitous discoveries.
  3. Human curation and chance encounters used to surface more meaningful, unexpected music, and moving discovery away from people to machines has made finding real gems harder and more effortful.

@adlrocha - Is N=NP?

@adlrocha Weekly Newsletter • 64 implied HN points • 14 Dec 25
🕹 Technology Algorithms
  1. Complexity theory measures how much time and memory algorithms need so we can tell which problems scale feasibly and which become intractable. It separates problems that are merely computable from those that are practically solvable before resources run out.
  2. P contains problems solvable in polynomial time, while NP contains problems whose solutions can be verified quickly even if they seem hard to find. NP-Complete problems are the hardest in NP because every NP problem can be reduced to them, and NP-Hard problems are at least that hard but not necessarily verifiable quickly.
  3. If P = NP, many cryptographic systems would break because one-way functions would no longer exist. At the same time, P = NP would let us solve huge optimization and AI problems exactly and efficiently, radically changing many fields.
Get a weekly roundup of the best Substack posts, by hacker news affinity:

New Algorithms, Not New Particles

Bzogramming • 45 implied HN points • 31 Dec 25
🕹 Technology Algorithms
  1. Most practical technology is built from atoms, electrons, and photons, so discovering new high-energy particles isn’t what drives usable engineering; progress comes from better math, materials, and system design.
  2. Condensed-matter and materials science (like semiconductors and superconductors) are where real, applicable breakthroughs live, because emergent behaviors of many atoms produce useful properties we can actually engineer.
  3. The next big advances will come from new algorithms, mathematical tools, and using physical and biological systems as computational substrates (aided by ML), not from finding exotic particles; building bigger, smarter systems from known primitives is the path forward.

The Sliding Window Technique[Technique Tuesdays]

Technology Made Simple • 279 implied HN points • 28 Feb 24
🕹 Technology Algorithms
  1. The sliding window technique is a powerful algorithmic model used for problem-solving in coding interviews and software engineering, offering efficiency and practicality.
  2. Benefits of using the sliding window technique include reducing duplicate work, maintaining consistent linear time complexity, and its utility in AI feature extraction processes.
  3. Spotting the sliding window technique involves identifying keywords like maximum, minimum, longest, or shortest, dealing with continuous elements, and converting brute-force approaches into efficient solutions.

Navigating Attention-Driving Algorithms, Capturing the Premium of Proximity for Virtual Teams, & New AI Devices

Implications, by Scott Belsky • 530 implied HN points • 18 Nov 23
🕹 Technology Algorithms
  1. AI-powered algorithms are driving polarization by optimizing for attention-grabbing content, widening the surface area of topics that stoke anger.
  2. Our social media feeds are now sourced from algorithmic preferences rather than social networks, shaping the content we are exposed to.
  3. The benefits of physical proximity in fostering creativity and relationships for teams will lead to the emergence of new technologies and management strategies supporting hybrid and remote work environments.

Algorithms start out as magic. But then become captive to corporate interests.

Asimov’s Addendum • 19 implied HN points • 19 Aug 24
🕹 Technology Algorithms
  1. Google has been found to have abused its power to control search engine results, limiting competition. This means they had an unfair advantage to keep other companies from competing effectively.
  2. Algorithms that start off as amazing tools can end up being exploited for corporate gain. The way Google uses its algorithms looks like magic at first but turns out to serve its own business interests.
  3. To foster fair competition in the tech industry, we need more transparency and rules about how algorithms work. This could lead to better choices for users and support new companies to grow.

Do You Want Your Newsletter To Show Up In AI Summaries?

Notes from a Small Press • 31 implied HN points • 06 Jan 26
🕹 Technology Algorithms
  1. Newsletter creators are being asked to decide whether their newsletters should be included in AI-generated summaries, raising a choice about inclusion in AI features.
  2. The article is behind a paywall and requires a subscription to read the full content, but a 7-day free trial is offered for new readers.
  3. The page provides clear subscription and sign-in options so paid subscribers can access the full archives and article.

wind and rain and sun and data

Dan Davies - "Back of Mind" • 334 implied HN points • 19 Jan 24
🕹 Technology Algorithms
  1. Supply and demand for electricity become more unpredictable with an increasing proportion of wind and solar energy
  2. The profit motive drives the application of information processing power and bandwidth to solve energy planning problems
  3. Market trading and the profit motive are ways to match the variety of the energy problem with the regulatory system

💡Friends vs Stranger Feeds

First 1000 • 589 implied HN points • 02 May 23
🕹 Technology Algorithms
  1. TikTok introduced algorithmic feeds of strangers, which was a novel idea.
  2. Twitter followed suit with their version of the 'For You' page.
  3. TikTok is experimenting with more specific feeds, such as topic-specific and friends-only feeds.

Can GPT-4 *Actually* Write Code?

Tyler Glaiel's Blog • 567 HN points • 17 Mar 23
🕹 Technology Algorithms
  1. GPT-4 can write code when given existing algorithms or well-known problems, as it remixes existing solutions.
  2. However, when faced with novel or unique problems, GPT-4 struggles to provide accurate solutions and can make incorrect guesses.
  3. It's crucial to understand that while GPT-4 can generate code, it may not be reliable for solving complex, new problems in programming.

The Broken Leg Check

Abstraction • 34 implied HN points • 07 Jan 26
🕹 Technology Algorithms
  1. Do a quick "broken leg" check first because a decisive news event can resolve a question immediately and save the time and cost of running the full forecasting pipeline.
  2. Be cautious: a wrongly triggered broken-leg update is dangerous since proper scoring heavily penalizes confident incorrect forecasts, so false positives can wipe out gains.
  3. Treat it as an empirical trade-off: implement a news-based detector, clearly define what "overwhelmingly resolves" means, track when it fires, and tune thresholds, confidence damping, or disable it if blowouts outweigh the savings.

Decay Functions

Abstraction • 29 implied HN points • 09 Jan 26
🕹 Technology Algorithms
  1. A single probability for a time window needs a decay model because where the probability mass sits across the window determines how much chance remains as time passes.
  2. Probability can follow different hazard patterns—constant (linear decay), increasing (back-loaded, like last‑minute negotiations), decreasing (front‑loaded, like ceasefires), or event‑driven—and each pattern changes how fast the cumulative probability is consumed over time.
  3. The forecasting bot classifies which hazard applies (defaulting to constant when unsure) and uses that to update remaining probability as time elapses, but this is a refinement that can be misclassified and matters most for long‑horizon questions.

455: Rolex Pricing, Jensen & Nvidia, China Chip Ban 1.5, Costco, Google Maps to Optimize Cities, Bicycle for the Mind, and Beckham

Liberty’s Highlights • 452 implied HN points • 18 Oct 23
🕹 Technology Algorithms
  1. It's liberating to realize that most fields are understandable to an interested outsider, focusing on big ideas.
  2. Exploring new fields and combining knowledge from different areas can lead to rich and interesting discoveries.
  3. Taking calculated risks and thorough preparation can lead to successful outcomes in business decisions, like pushing all the chips in.

Under the microscope: Michelin Rally Masters (PlayStation)

Rings of Saturn • 43 implied HN points • 02 Dec 25
🕹 Technology Algorithms
  1. Many published cheats are wrong because the game expects a 13-character code, not the 12-character string that appears on most sites; the correct "unlock everything" code includes a final question mark.
  2. The code system is fairly complex: it uses a 32-character alphabet, three separate checksums, and a PRNG that shifts the alphabet to decode the first ten characters.
  3. Entering a valid code unlocks six bonus cars and enables the WRC and Legends single-player modes, but it doesn’t add any extra drivers beyond those listed in the manual.

The Sequence Opinion #778: After Scaling: The Era of Research and New Recipes for Frontier AI

TheSequence • 28 implied HN points • 25 Dec 25
🕹 Technology Algorithms
  1. Scaling up transformers with more data and compute drove past AI gains, but that straightforward path is hitting limits because high-quality pretraining data and scaling efficiency are finite.
  2. The field is shifting to an "age of research" where diverse experiments and new ideas, not just bigger models, will determine future breakthroughs.
  3. Progress will come from a toolbox of new recipes — like souped-up pretraining, novel architectures, and improved fine-tuning — that turn compute into faster learning, better adaptation, and fewer odd model failures.

The CAP Theorem of Clustering: Why Every Algorithm Must Sacrifice Something

Confessions of a Code Addict • 529 implied HN points • 29 Oct 24
🕹 Technology Algorithms
  1. Clustering algorithms can never be perfect and always require trade-offs. You can't have everything, so you have to choose what matters most for your project.
  2. There are three key properties that clustering should ideally have: scale-invariance, richness, and consistency, but no algorithm can achieve all three simultaneously.
  3. Understanding these sacrifices helps in making better decisions when using clustering methods. Knowing what to prioritize can lead to more effective data analysis.

GM Shaders: Noise 3

GM Shaders Mini Tuts • 157 implied HN points • 13 Jan 24
🕹 Technology Algorithms
  1. Simplex noise is a noise algorithm that is cheaper in higher dimensions and produces fewer artifacts than Perlin noise.
  2. Consider baking noise onto textures for optimization, especially in 2D for better performance.
  3. To make noise tillable, stick with a square grid and use modulo tiling for seamless textures.

Diffusion Models are Evolutionary Algorithms

Gonzo ML • 441 implied HN points • 09 Nov 24
🕹 Technology Algorithms
  1. Diffusion models and evolutionary algorithms both involve changing data over time through processes like selection and mutation, which can lead to new and improved results.
  2. The new algorithm called Diffusion Evolution can find multiple good solutions at once, unlike traditional methods that often focus on one single best solution.
  3. There are exciting connections between learning and evolution, hinting that they may fundamentally operate in similar ways, which opens up many questions about future AI developments.

REVEALED: How the Trump admin is covertly reconfiguring online algorithms

Weaponized • 113 implied HN points • 01 Aug 25
🇺🇸 U.S. Politics Algorithms
  1. The current strategy involves using government pressure to change how online platforms work, encouraging them to promote certain ideas while minimizing others. This means the algorithms are influencing what information people see without direct censorship.
  2. New rules require AI used by government agencies to be free of 'ideological bias,' which many argue isn't truly possible. This affects not only government tools but also private companies that want government contracts, shaping public information more broadly.
  3. This shift in online content management allows the government to appear neutral while controlling the narratives that are visible to the public. People may not realize that what they see online is being influenced by political agendas.

Controlling Rate-Limited Events - Implementation

Low Latency Trading Insights • 117 implied HN points • 11 Feb 24
🕹 Technology Algorithms
  1. The requirements for a rate-limiting algorithm include precise event counting, fast performance especially during market turbulence, and minimal impact on cache memory.
  2. Creating a rate-limiting algorithm using a multimap for counting events has inefficiencies; a better solution involves enhancements for optimal performance.
  3. A bounded approximation approach for rate limiting achieves memory efficiency by assuming a minimum time precision and implementing a clever advance-and-clear mechanism.

The Statistician Who Loved Machine Learning

Mindful Modeler • 279 implied HN points • 23 May 23
🕹 Technology Algorithms
  1. Leo Breiman emphasized the importance of both data modeling culture and algorithmic modeling culture in statistical modeling.
  2. Breiman advocated for being problem-focused over solution-focused, encouraging modelers to choose the appropriate mindset based on the task at hand.
  3. Understanding various modeling mindsets, such as statistical inference and machine learning, is crucial for effective modeling.

How to Learn Data Structures and Algorithms [Storytime Saturdays]

Technology Made Simple • 299 implied HN points • 22 Jan 23
🚌 Education Algorithms
  1. Understanding Data Structures and Algorithms is crucial for success in technical fields like software development.
  2. Many resources focus on DSA for coding interviews, but it's important to go beyond that to deepen your knowledge.
  3. Learning DSA effectively doesn't have to involve answering countless questions or watching numerous tutorials; there are better approaches available.

GM Shaders Mini: Dither

GM Shaders Mini Tuts • 196 implied HN points • 25 Jul 23
🕹 Technology Algorithms
  1. Dithering can help blend colors smoothly without mixing them
  2. There are three common types of dithering algorithms: Ordered (Bayer), White Noise, and Blue Noise
  3. Blue noise can provide a more even distribution of values compared to White noise

TheSequence Opinion #699: 2030 or Bust? The Compute Surge and the Bottlenecks Ahead

TheSequence • 84 implied HN points • 07 Aug 25
🕹 Technology Algorithms
  1. Artificial General Intelligence (AGI) might be possible by 2030 if we keep improving our computing power and models.
  2. However, there are worries that after 2030, we could hit limits with our technology that will require us to find new ways to innovate.
  3. We might need better algorithms and improved designs because just making computers bigger and faster won't be enough forever.

Return of Sam

Bojan’s Newsletter • 137 implied HN points • 28 Nov 23
🕹 Technology Algorithms
  1. Sam Altman returns as OpenAI CEO after being ousted.
  2. X solidifies its position as the go-to platform for tech news updates.
  3. Stakeholders at OpenAI now have a heightened sense of urgency and focus on accelerating projects.

GM Shaders Mini: Math

GM Shaders Mini Tuts • 157 implied HN points • 11 Sep 23
🕹 Technology Algorithms
  1. Alpha blending in shader programming requires blending colors and alpha channels separately.
  2. Weighted averages provide greater control for combining multiple elements together in shaders.
  3. Creating a simple 3D perspective effect in shaders involves scaling with a linear gradient.

GM Shaders Mini: Imagination

GM Shaders Mini Tuts • 157 implied HN points • 02 Sep 23
🕹 Technology Algorithms
  1. When working with shaders, think in terms of vector fields to direct the flow and create gradients.
  2. Consider the acceptable input domains and the output ranges of your functions to prevent errors and unexpected results.
  3. Utilize periodic functions for repetition, sine and cosine for waves and rotations, dot product as a ruler, and exponentiation for adjusting brightness levels.

Instacart's price discrimination isn't price discrimination

Economic Forces • 15 implied HN points • 16 Dec 25
💼 Business Algorithms
  1. Different price changes have different causes and effects: A/B tests, strategic randomization, dynamic supply-and-demand adjustments, and true price discrimination are not the same thing.
  2. The Instacart example looked like randomized A/B testing rather than pricing based on shoppers’ personal data, so treating every price change as evidence of algorithmic profiling confuses what might happen with what actually happened.
  3. Price discrimination isn’t automatically bad — it can raise output and sometimes help consumers, especially under competition — and banning price experiments won’t necessarily make consumers better off because low-price periods can outweigh high-price losses.

An algorithm for quality

Substack Blog • 774 implied HN points • 19 Sep 23
🕹 Technology Algorithms
  1. Algorithms can serve to feed the basest version of ourselves
  2. Technology platforms should aim to serve the best version of the user
  3. Reading with agency and curating your attention can lead to a more fulfilling experience

Compression - en francais

Squirrel Squadron Substack • 3 implied HN points • 04 Feb 26
🕹 Technology Algorithms
  1. Compression works by removing redundancy to make data smaller; lossless compression preserves every bit while lossy methods discard detail, and truly random data resists any meaningful shrinking. Recompressing already-compressed data usually fails and can make files bigger, so there are strict limits to how far you can compress.
  2. Information theory defines limits on compression and measures information by how short a program can reproduce the data (Kolmogorov complexity). Effective compression depends on clever representations and adaptive algorithms that capture structure in the data.
  3. Large language models behave like powerful compression-and-prediction systems that build compact internal models by learning to predict the next token. This predictive compression explains much of their useful, seemingly intelligent behavior and their value as productivity tools, even if they are not human thinkers.