Search
Showing results for “Algorithms”
Jump to a topic
AlgorithmsAn algorithm is a series of steps for completing a task. Broadly, this includes everything from the instructions in a recipe and the procedures used to solve mathematical problems to the software underlying search engines and the ways AI systems identify patterns in data. Colloquially, algorithms are associated with recommendation systems that provide users with personalized content based on their preferences, as seen in social media feeds.
Although algorithms have existed as long as routines, processes, and mathematics have, the term originated from Algoritmi, the Latin translation of al-Khwārizmī, a 9th-century mathematician whose book "Al-Jabr" gave algebra its name. In 1843, English mathematician Ada Lovelace—considered the first computer programmer—designed the first computer algorithm, which showed how a machine could perform a complex mathematical calculation. By 1936, Alan Turing proved that a machine could be designed to execute any conceivable algorithm, laying the foundation for modern computing.
Since then, algorithms have been developed to process data faster than can be done by hand, with complex systems able to automatically self-improve and become more efficient. Certain algorithms, such as those in hiring and lending, may reinforce existing biases because they rely on human programming or, in the case of machine learning, historical patterns in data.Explore Algorithms
What we've found
Algorithms built to detect credit card fraud can deny charges if they perceive the attempted purchase is unfamiliar or unusual based on historical data, even when a person has enough credit to make the purchase.Since fraudulent charges are rare compared with the much greater number of transactions across purchase histories, these models have been shown to be overly sensitive, lacking sufficient training data on what legitimate fraud looks like. Models also perform less reliably for groups in low-income areas or communities of color, given less available data from which to build an accurate baseline. The ConversationAn explanation of dynamic pricing algorithms and the factors that impact them Rather than static prices set by manufacturers or retailers based on the costs of goods, online businesses or box stores with electronic price tags can change prices in real time based on supply, demand, time, location, and competitor prices. These algorithms can be programmed to maintain or maximize profit margins based on shoppers' willingness to spend. The Wall Street JournalWhen you complete a CAPTCHA on a website to prove you are human, you may be unknowingly training algorithms for self-driving cars, whose systems require training data to help identify crosswalks, bicycles, and other outdoor objects. MediumVisualize how data becomes organized through a variety of sorting algorithmsThis interactive shows how different algorithms shift elements in a randomized set to order them from smallest to largest. By comparing the different procedures and their corresponding functions, users can explore how runtimes vary across algorithms that complete the same task. Simon Willison's ToolsThe origin of shortest-path algorithms and how they became optimized for navigation appsIn 1956, Dutch programmer Edsger Dijkstra spent about 20 minutes developing an algorithm to demonstrate how computers could quickly determine the shortest route between Rotterdam and Groningen. Since then, modified versions of Dijkstra's formulation comprise the processing backbone of systems such as Google Maps and Apple Maps. VeritasiumAn explanation of Turing machines—the foundations of digital algorithms and computersThe device reads data, overwrites or ignores it, and then moves on to the next piece of data based on a set of programmed rules. By imagining this data as zeros and ones—binary that makes up computer code—Turing machines became the theoretical models for how any algorithm could be run on a computer if the procedure could be coded into a set of digital rules. ComputerphileAn overview of 12 fundamental types of algorithms and the uses for eachThese include search algorithms to quickly process and find important information, cryptographic algorithms to secure communications, randomized algorithms—such as Monte Carlo simulations for estimating probabilities of outcomes—and greedy algorithms, which make optimal choices based on available data in a step-by-step fashion, as seen in navigation apps. LinkedInA breakdown of algorithms, how they are constructed, and their applicationsAlgorithms are instructions for accomplishing a goal and typically take the form of a step-by-step procedure. Given the conditions, these procedures can take various paths, increasing an algorithm's complexity. While traditional computer algorithms are programmed to account for all possible scenarios within a given process, machine learning systems build their own rules from data-driven patterns. Stanford Graduate School of BusinessEmergency room evaluations may be improved with AI algorithmsEmergency room care teams must make fast and crucial decisions. They must determine if patients need to be admitted or discharged and when to order more tests or not. Researchers have been exploring how often patients died in the week after they were sent home and evaluating how this and other aspects of emergency care decision-making could be improved with AI algorithms to assist decision-makers. Freakonomics Radio NetworkNSF-funded research into matching algorithms improved the success of organ transplantsInvestigations in game theory, including concepts focused on creating mutually beneficial partnerships—matching theory—led to the design of computer systems that efficiently matched kidney recipients and live donors, resulting in higher-quality matches. By 2019, these matching algorithms were used for multi-organ exchanges. UBSLearn how the algorithms behind Tinger, Bumble, Hinge, and other dating apps workFormerly based on the Elo rating system, which ranks chess players, the Tinder algorithm is built on user data, with more time spent on the app providing more data on user preferences from which to customize potential matches. OkCupid's match percentage algorithm relies on overlaps in users' search preferences and responses to a subset of over 4,000 questions. MashableSmart pricing algorithms can unintentionally work together to raise pricesResearchers have found that, through trial and error, algorithms designed to adjust prices based on market data can learn to mimic price-fixing tactics, even without explicit programming to do so. Over time, they "learn" that it is more profitable to avoid price wars and keep prices high when competitors lower prices. Quanta MagazineSocial media algorithms are shaping real-life language through memesThe evolution of the internet in the early 2000s supercharged the creation and spread of memetic content. Now, highly tuned algorithms on platforms like TikTok and YouTube influence the language, phrasing, and content of creators, which shapes how we use language in the real world. On with Kara SwisherHow VantageScore and FICO’s algorithms calculate credit scores differentlyVantageScore and FICO weigh various factors that determine your credit score differently. For instance, while your payment history makes up 40% of your VantageScore, it only makes up 35% of your FICO score. Business InsiderWatch how algorithms can be designed to train themselves and self-improve without human input, to the point where the optimized code is too complex for programmers to understand.Algorithms are pervasive in digital interactions, influencing what we see, buy, and perceive online. Driven by vast amounts of initial data, the creation and evolution of these algorithms involve testing, building, and iterating upon themselves without additional human input. CGP GreyUnderstanding the causes, effects, and prevalence of algorithmic biasFound in search engines, hiring and college admissions tools, facial recognition systems, and other software, algorithms act on the rules they are programmed with, which may be authored by individuals unaware of their own biases. Algorithms based on limited, incomplete, or inaccurate data are more likely to inherit and propagate existing biases. TEDx TalksYouTube's recommendation algorithm relies on over 80 billion daily pieces of informationThese signals include which videos a user has watched, searched for, and liked, as well as channel subscriptions and feedback selections, such as videos marked "not interested." Each user's viewing habits are compared with those of other users with similar habits, and this information is used to suggest additional content. YouTube Official BlogHow quantum computers can empower Shor's algorithm to break digital encryptionSecure communications largely rely on numeric keys generated by multiplying two enormous prime numbers together, and it is computationally impractical to determine these primes by trying to divide by every possible one. While Shor's algorithm can provide better guesses for this brute-force approach, it relies on a number that is difficult to determine on classical computers but much easier to calculate using the simultaneous computations enabled by quantum superposition. minutephysicsEnglish mathematician Ada Lovelace wrote the first computer algorithm a century before computers existed and hypothesized that computing devices would one day be able to compose music—foreseeing the generative AI systems of today. SciShowThe term 'algorithm' originated from the Latinized name of al-Khwārizmī, a 9th-century mathematician whose works introduced Arabic numerals and the decimal-based numeral system to Europe and whose treatise 'Al-Jabr' gave algebra its name. BBC IdeasThe mathematics of quantum-based cryptography, and why existing data is already at riskSeeing the potential of this technology, some governments and malicious actors are thought to have already started storing encrypted data, in the hopes that quantum computers within the next 10 to 20 years will be able to decrypt it. As of 2022, the National Institute of Standards and Technology had selected four algorithms for inclusion in a quantum-resistant encryption standard. VeritasiumThe different types of stablecoins include algorithmic stablecoinsIn addition to fiat-backed stablecoins, commodity-backed stablecoins, and cryptocurrency-backed stablecoins, algorithmic stablecoins use smart contract-based mechanisms to maintain a stable value by responding to supply and demand. KrakenInstead of .com or similar, web addresses on the dark web end in 'd.onion'Since the latest version of the onion routing protocol, dark web domain names consist of a random set of 56 alphanumeric characters followed by .onion (e.g., propublica.org → p53lf57qovyuvwsc6xnrppyply3vtqm7l6pcobkmyqsiofyeznfu5uqd.onion). The last character is always "d" due to the encryption algorithm's requirements. DigicertIf aliens visited Earth, it seems unlikely they'd be evolved biological creatures like usThe writer behind Mindless Algorithm, George McGowan, challenges various assumptions about aliens depicted in science fiction. One of those is that aliens are, like humans, individual entities that live, age, and die. McGowan suggests that the difficulty of intergalactic travel and the aliens' technological advancements make it more likely that they'd be an advanced form of AI or digital copies of biological entities. Mindless AlgorithmAphantasia involves a lack of mental images—but doesn't affect behaviorThe condition, generally known about for over a century, first received its name in 2015 by researcher Adam Zeman. An estimated 4% of people lack this ability to perceive mental images, and research suggests this deficiency does not have any practical impact on behavior. NatureHelp scientists at LIGO train AI to resolve glitches in observational dataVolunteers are provided with visual representations of signal abnormalities at various frequencies and tasked with categorizing them based on reference examples. Progressing through the tool unlocks more difficult images and additional features. Categorized images will be used to train machine learning algorithms for automated future classification. ZooniverseRapid environmental changes are exacerbating natural extinction processesHuman activities behind climate change, habitat destruction, pollution, and overhunting have resulted in extinction rates up to 1,000 times greater than normal, triggering massive biodiversity loss. This has led scientists to refer to the current geological age as the Anthropocene era, in which humans are the dominant force shaping nature. Stuff You Should KnowThe TikTok sale went through in January 2026The app's ownership is now majority American, though TikTok CEO Shou Chew, who is Chinese, remains on the board of directors. It's unclear what, if anything, has changed for American users, as the company has historically kept the specifics of its algorithm a secret. ABC NewsThe first algorithmic random number generator was developed in 1946Although previous methods of randomness involved physical tools with limited outcomes, such as dice, or relied on chaotic variables for number generation, such as electronic noise, John Von Neumann's middle-square method turned the problem into a mathematical computation. It has since been replaced by methods that generate numbers from unique initial values, such as the time on a clock. Interesting EngineeringConspiracy theories today are often born from online radicalizationNew York Times tech columnist Kevin Roose examines how individuals become radicalized while using platforms like YouTube. Hate crimes that took place before the COVID-19 pandemic inspired the podcast, which looks at online commentators, algorithms, and isolation, and possible explanations for how increased online activity can sometimes lead to violent crime. SpotifyScientists use ground-based radar data to detect turbulence within cloudsNCAR developed the NEXRAD Turbulence Detection Algorithm and the Graphical Turbulence Guidance system to provide forecasts up to 18 hours ahead using ground-based Doppler radar data. Software systems also utilize an automated observation and reporting package—the EDR system—found on Delta, United, and Southwest Airlines aircraft. National Center for Atmospheric ResearchViral pop artists create systems of participation, not just musicMusicologist (and Swift expert) Paula Clare Harper weighs in on the factors that make music go viral. Tracks designed for shareability that feed well into content algorithms on streaming platforms are able to grow in popularity, with traditional human gatekeepers sidelined. Big BrainsThe dead internet theory suggests that mostly bots are interacting onlineThe dead internet theory proposes that bots create most online content. These bots are believed to create social media posts to boost engagement and increase advertising revenue. However, the accounts doing the liking and commenting are also believed to be bots. The theory suggests governments use bots to manipulate public opinion. Live ScienceThe fast Fourier transform breaks down wave signals into discrete componentsThe algorithm was initially discovered by German mathematician Carl Friedrich Gauss in 1805, but his notes were not found until after his death. Today, FFTs form the basis for most compression software in media and are essential to technologies such as radar, sonar, and telecommunications. VeritasiumTikTok's personalized algorithm activates the brain's reward systemA 2025 Washington Post study found that 75% of the TikTok users it studied nearly doubled their time spent on the app within a month. Experts in design and psychiatry say this is due to the app's interface and algorithm, which, after learning about the user's preferences, recommend videos more likely to activate that person's dopamine reward system than non-personalized suggestions. The Washington PostAI is learning to be funnyExperts consider humor a particular challenge for large language models to learn, given the skill's complex linguistic play. In an experiment, a stand-up comedian performed half AI-produced jokes, and half human-produced jokes, to no discernible difference in the audience's laughter. Undark MagazineTikTok is transforming the English languageLinguists like Adam Aleksic have argued that the way internet slang—especially the kind that thrives on TikTok—enters the real world isn’t so different from how English evolved in its earliest days. Literary HubHow FICO and VantageScore weigh factors in credit history differentlyVantageScores are sometimes lower than FICO scores because the algorithms that calculate the two scores weigh the factors that influence them differently. For instance, VantageScores typically weigh payment history more heavily than FICO scores. ChaseCritics call WeChat a surveillance toolThis report from the Citizen Lab details evidence WeChat, with 1.4 billion users, is a powerful surveillance tool for the Chinese Communist Party. In particular, it finds the government uses non-China accounts to train censorship algorithms for use on domestic communications. The Citizen LabInside MLB's robo-umping experimentHuman umpires call roughly 94% of pitches right overall, but performance reduces on borderline pitches in high-leverage moments, the calls that shape game outcomes most. A professor and historian analyze early experiments with the full "robot ump" system that helped shape MLB's decision to implement a compromise "challenge" system. The ConversationFacebook's algorithm chooses content to display on the News Feed based on user dataIt employs machine learning to prioritize posts and ads based on user interactions, content type, and post timeliness, creating a personalized, engaging experience for each user. SocialBeeHow PageRank—Google's search algorithm—mimics academic citation analysisJust as research papers carry more weight if they are cited more often and cite major papers to reach their conclusion, Google's PageRank identifies search relevance based on the number and quality of links pointing to a page. Zach StarHow the YouTube algorithm works, according to MrBeastIn this brief interview clip, the YouTube savant "MrBeast" summarizes how the YouTube algorithm—which he relies on to make his content go viral—actually works. MagnatesMediaWhy TikTok's algorithm is likely driven by what you linger onWhile the inner workings of the engine driving the vast majority of the app's engagement remain a secret, it is understood to rely heavily on what content users spend the most time watching rather than swipe past. Experts believe the app can develop user profiles robust enough for its algorithm to consistently match engaging content to in as little as two hours. Wall Street JournalTikTok's algorithm has changed the way hits are madeSongs aimed at TikTok's algorithm—designed for viral dances or soundtracking common experiences—often find success on the Billboard charts. An early example was Lil Nas X's "Old Town Road," which used his memes and a country-trap beat from the Netherlands to rise to the top of the Billboard charts. The New York TimesNetflix's recommendation algorithm was born from a $1M competitionIn 2006, Netflix launched an open competition for data science teams to develop approaches to outperform its recommendation algorithm. Three years later, the prize was awarded after a group of competitors merged, developing a solution that beat Netflix's internal algorithm by more than 10%. Today, recommendations are based on what users watch, particularly via bingeing. ThrillistMachine learning and deep learning are successive subsets of artificial intelligenceML allows systems to learn from data, while DL uses complex neural networks to analyze large amounts of unstructured data. Neural networks enable capabilities like image and speech recognition through their use of algorithms to identify patterns in data. IBMQuantum computers harness weird physics to explore unsolvable problemsClassical computers manipulate bits as either 0 or 1, but quantum computers use qubits to hold combinations of both, enabling new algorithms. This new approach could one day help with breakthroughs in medicine, chemistry, and AI. WIREDExplore a visual guide to how quantum computers can crack RSA encryptionIn 1994, Peter Shor introduced an algorithm that showed a quantum computer could efficiently factor the product of two very large prime numbers, undermining the mathematics that makes RSA so secure. While using larger keys can delay the decryption time of quantum computers, this solution increases the time it takes to encrypt data, which may become impractical. ReutersHow algorithmic redistricting worksComputational methods are convincing a number of judges that some state district maps are drawn illegally. Here's a simple overview of how it works. The Washington Post
Try another search?