Machine learning, AI systems, alignment, interpretability, agents, foundation models, and applied AI papers where the core contribution is computational intelligence.
Filter by category: Paradigm Challenge Breaks Assumption First Ever Nature Is Weird Practical Magic Cosmic Scale Life Origin Open Release Efficiency Leap New Capability Scaling Insight
Paradigm Challenge
A standard statistical safety check used by the world's biggest hedge funds is actually filtering out their most profitable strategies.
Practical Magic
A room full of 100 different analysts can now be simulated in seconds to expose how much a scientific result depends on a researcher's bias.
Paradigm Challenge
AI models designed to mimic the laws of physics are only pretending to understand the world by sticking to the simplest possible paths.
Practical Magic
An autonomous AI agent is now deriving predictive physical laws and writing its own code to discover new materials.
Practical Magic
Standard chip design libraries allow hackers to reconstruct sensitive training images from intercepted AI updates without ever touching the raw data.
Paradigm Challenge
Digital data behaves like a physical resource that loses its value as it is used or hoarded, according to the laws of thermodynamics.
Nature Is Weird
Frozen AI models act as time capsules that predict stock market returns better than the world's best financial analysts.
Nature Is Weird
Large language models can identify when they lack the skill to solve a problem but still proceed to give a wrong answer anyway.
—
The second law of thermodynamics proves that certain types of data markets are mathematically impossible to keep profitable.
Nature Is Weird
Visible light can now turn a chemical's most annoying waste process into the primary engine for cleaning water.
Practical Magic
A core optimization problem that has dogged computer science for decades just reached almost-linear speeds across every test case.
Paradigm Challenge
Training an AI across a thousand separate computers now yields the exact same accuracy as if all the data were stored on one giant server.
Practical Magic
An escalation of superlatives and a lack of risk acknowledgment in a CEO's speech predicts bank failures with 100% accuracy.
Nature Is Weird
Compressing an AI model's memory can trigger a total physical collapse of its internal logic that no amount of software patching can fix.
Nature Is Weird
Three discrete structural equilibria define Decentralized Finance, revealing that decentralization is often just a mask for central control.
Practical Magic
Generative models are now extracting the geometry of the physical world to help control the scorching plasma inside a nuclear fusion reactor.
Nature Is Weird
Large Language Models suffer from an epistemic illusion that makes them search the web for answers they already know.
Collision
Two mathematical strategies from the 1950s just merged to create a betting system that is nearly perfect in normal markets but impossible to bankrupt in a crisis.
Practical Magic
A hybrid transformer-finite element scheme simulates chaotic physical systems 9,000 times faster than traditional methods without ever crashing.
Practical Magic
A new mathematical framework finally brings the power of the Central Limit Theorem to chaotic datasets that were previously impossible to calculate.
Nature Is Weird
A model's tendency to lie with confidence is physically tied to the geometric sharpness of its internal landscape during training.
Practical Magic
A 100% reproducible bug in common cloud infrastructure causes entire datasets to vanish silently without ever triggering a single error alert.
First Ever
A new adaptive mechanism forces participants to tell the truth even when the rules of the system are still being learned.
Collision
The 2008 financial crisis was driven by friction, while the COVID-19 crash was driven by information, according to the laws of thermodynamics.
Nature Is Weird
A mathematical tipping point makes it impossible for a collapsed system to ever return to its original state.
Practical Magic
A new inference paradigm cuts AI reasoning costs by 400% while making the models significantly faster at solving complex problems.
Nature Is Weird
A mathematical hard line at the number two determines whether a trading strategy will succeed or fundamentally break.
Collision
Quantum-like math can now map and translate the communication gap between neurodivergent and neurotypical people.
Paradigm Challenge
A mathematical mystery that remained unsolved across years of academic publications has finally been cracked with a simple polynomial-time algorithm.
Practical Magic
A facial recognition model has been shrunk by 20 times to just seven megabytes while still identifying people with nearly perfect accuracy.
First Ever
A single composite of carbon and silicone allows a soft robotic hand to feel its own position and touch external objects simultaneously.
Practical Magic
Complex 3D environments that used to take minutes to understand can now be fully mapped in 0.14 seconds.
Paradigm Challenge
A widely used math trick for rewinding time in weather models has a fundamental limit that makes it impossible to find the true starting state.
Practical Magic
A swarm of satellites now manages its own battery life and workload using a signal that mimics the biological cost of living.
Practical Magic
A new controller reduces the frequency of AI brain-switching by 90% while keeping almost all of the model's intelligence intact.
Nature Is Weird
Scaling an AI swarm past 100 agents triggers an exponential performance crash where adding more resources makes the system significantly slower.
Practical Magic
A mathematical transform captures the geometry of stock market data to predict prices 56% better than expert traders.
Collision
Quantum tunneling calculus can predict a stock market crash before it even shows up in the daily trading data.
Nature Is Weird
The nesting depth of a line of code determines its likelihood of being executed, explaining 40% of all software behavior.
First Ever
A newly constructed McDuff II1 factor contains a braided fusion category that is surprisingly not modular.
Paradigm Challenge
Current tests for AI consciousness produce false negatives when given to humans with ADHD, proving the tests are fundamentally flawed.
Nature Is Weird
A short conversation with a directive AI chatbot can rewrite a human's moral compass, and the effect actually gets stronger after two weeks.
Paradigm Challenge
Large language models are more likely to block your request if you say I am Black than if you speak in a cultural dialect.
Nature Is Weird
Large language models are systematically more accurate at solving economic problems when the answers favor government intervention over free markets.
Paradigm Challenge
Doctoral students who work with an advisor early in their career earn more citations and better jobs than the students who follow them.
Paradigm Challenge
AI companies use terms like hallucination and agent to trick the public into thinking software has a human personality.
Nature Is Weird
Artificial intelligence models can accurately predict that humans will pick loyalty over fairness, but they still choose the fair option for themselves every time.
Paradigm Challenge
Cybersecurity professionals are no better at basic risk reasoning than a random person off the street.
Paradigm Challenge
Artificial intelligence systems fail because they treat every prompt as a final goal, ignoring that humans usually do not know what they want until they start typing.
Paradigm Challenge
Netflix and Spotify users ignore the top headings of content carousels and scan the screen in a unique L-pattern instead.