A Contest to Beat Geopolitical "Superforecasters"
Four years ago, a team of researchers based at the University of Pennsylvania wowed the U.S. intelligence community by producing a superior new way to forecast geopolitical events. They were dubbed the "Superforecasters."
Driving the news: At a time the science of professional prognostication is sorely battered, the radical innovation arm of U.S. intelligence services is looking to best the UPenn team with a fresh big-money prize competition.
What's happening: IARPA, the research arm for the U.S. director of national intelligence, is offering $250,000 in prize money in a contest to forecast geopolitical events such as elections, disease outbreaks and economic indicators.
The Peculiar Blindness of Experts
...In 2005, [Philip] Tetlock published his results, and they caught the attention of the Intelligence Advanced Research Projects Activity, or IARPA, a government organization that supports research on the U.S. intelligence community’s most difficult challenges. In 2011, IARPA launched a four-year prediction tournament in which five researcher-led teams competed. Each team could recruit, train, and experiment however it saw fit. Predictions were due at 9 a.m. every day. The questions were hard: Will a European Union member withdraw by a target date? Will the Nikkei close above 9,500? Tetlock, along with his wife and collaborator, the psychologist Barbara Mellers, ran a team named the Good Judgment Project.
Identifying Trojans in AI
With the rise of adversarial AI, government researchers are looking for ways to automatically inspect artificial intelligence and machine learning systems to see if they've been tampered with. [...] The Army Research Office and the Intelligence Advanced Research Projects Activity are investigating techniques to spot and stop these Trojans in AI systems. Given the impossibility of cleaning and securing the entire training data pipeline, the broad agency announcement for the TrojAI program is looking to develop software to automatically inspect AI and predict if it has a Trojan. [...] Initial concept papers are due May 31, and the highest ranked applicants will be invited to submit a full proposal. Read the full BAA here.
IARPA Expands Research on Protecting AI Systems from Tampering
The U.S. government’s Intelligence Advanced Research Projects Activity (IARPA) is planning a pair of programs to prevent training data from being maliciously tampered with to turn artificial intelligence systems against their users.... One project, Trojans in Artificial Intelligence (TrojAI), seeks to create a warning system for machine-learning algorithm training data compromised by an adversary. That project was originally announced in December , and industry has provided feedback on it. Details of the second project will be revealed in a draft announcement later this year, but [IARPA Director Stacey] Dixon said that it will focus on protecting the identity of people whose images have been used to train facial biometric algorithms.
To upgrade quantum computers, researchers look to materials science
For decades, theoretical physicists and computer scientists have been compiling evidence that quantum computers will eventually leave our current top-of-the-line supercomputers in the dust…. Quantum computers have been limited to simple problems because of their hardware. The basic elements of quantum circuits, called qubits (for quantum bits), are still highly error prone. Truly useful quantum computers will probably need millions of robust qubits, a far cry from the tens of qubits operating in today’s machines. And if the ones we have today are misfiring, there’s no hope of a million-qubit system calculating anything with certainty. So scientists and engineers like Hong are shouldering the responsibility of building a better qubit.
“A qubit uses quantum mechanical phenomena to do things with information you can’t otherwise do,” says Brad Blakestad, a program manager for quantum computing at the Intelligence Advanced Research Projects Activity (IARPA).