Latest news with #DARPA
Forbes
16 hours ago
- Science
- Forbes
How To Quickly Improve Your Ability To Predict The Future
A few individuals have a heightened ability to forecast what will happen next. What traits do they ... More share? (Photo by) The need to predict is omnipresent. Every time you buy a stock, choose a partner, pick a president, or bet your brother-in-law that the 49ers will finally win it all, you're making a decision based on a prediction. And yet, despite all the big data, algorithms, learning models, and AI assistants, we're still not very good at predicting the future. But, turns out, there are a few techniques that will help you get better fast. The human desire to improve our ability to predict the future isn't new. Nostradamus was one of the original prognosticators to receive acclaim. Yet, on further reflection, his writings are so open to interpretation that they could be describing either the fall of Rome or the next global pandemic. In recent decades, predicting the future of everything has become a growth industry, especially in politics. Cable news needs experts who sound cocksure about everything, even if their accuracy is less than dart-throwing chimps. As long as the ratings are good, bring on the blather. A few individuals have a heightened ability to forecast what will happen next. What traits do they share? University of Pennsylvania professor Philip Tetlock has spent decades trying to answer this question: Spoiler alert: It's not fame, credentials, or wearing a bowtie on TV. Determined to find out what does make someone a good predictor, Tetlock launched a bold experiment. With funding from DARPA, he hosted forecasting tournaments known as the Good Judgment Project. Tens of thousands of ordinary people—teachers, engineers, pharmacists, and even a Canadian underwater hockey coach—competed to see who could best predict the outcomes of real-world events: Will the president of Tunisia go into exile next month? Will the price of gold exceed $3500 by the end of Q3? Tetlock identified a small percentage—about 2 percent—who consistently made remarkably accurate predictions. He dubbed them 'superforecasters.' They weren't clairvoyant. They didn't have access to classified information. But they do have certain traits in common: What You Think Versus How You Think Tetlock puts it this way: 'What you think is much less important than how you think.' Superforecasters don't get attached to their opinions. They revisit assumptions. They seek out dissent. One participant even wrote code to curate news articles from across ideological spectrums so he wouldn't fall into an echo chamber. They also tracked and scored their predictions over time, treating it not as a parlor trick, but as a craft. If you want to improve your ability to anticipate the future—and let's be honest, who doesn't—Here are a few suggestions: 1. Start with the base rate. Ask yourself: What usually happens in situations like this? Don't be seduced by the drama of outliers. Begin with the average. 2. Break it down. Instead of 'Will AI take my job?' ask: 'What tasks in my role are automatable?' Then assign probabilities to each. 3. Toggle perspectives. Use both the inside view (your specific context) and the outside view (what's happened in similar situations). 4. Stay flexible. Your assumptions are not sacred scrolls. Update them when new information arrives. Bonus points if you can admit you were wrong without needing therapy. 5. Use numbers, not vibes. Avoid vague terms like 'probably.' Go with: 'I'm 70% confident.' It sharpens your thinking—and makes you easier to argue with at dinner parties. 6. Keep a prediction journal. Write down your forecasts and your reasoning. Revisit. Learn. Repeat. (Optional: give yourself gold stars.) 7. Seek disconfirmation. Don't just look for information that proves you right. Hunt down what might prove you wrong. It's called 'growing.' 8. Diversify your info diet. Read widely. Follow smart people you disagree with. Cross-pollinate. Avoid becoming the human version of a YouTube algorithm. In the end, getting better at prediction won't make you omniscient, but it will make you wiser, calmer, and a better decision-maker. And maybe, just maybe, the next time someone at work says, 'Nobody could have seen this coming,' you'll be able to smile and say, 'Actually… I kind of did.'
Time of India
2 days ago
- Science
- Time of India
Germany turns live cockroaches into spies with AI-powered backpacks
Image: Reuters In a move straight out of a science fiction novel, Germany is pioneering the use of live cockroaches as covert surveillance tools. A tech startup called SWARM Biotactics , based in Kassel, is developing AI-powered "backpacks" that are mounted on real cockroaches, effectively turning them into biologically integrated spy agents. These miniature devices are equipped with sensors, cameras, and neural stimulators, allowing remote control and autonomous swarming in tight or inaccessible environments, according to Reuters. Backed by over €13 million in funding, this innovation is set to redefine the future of surveillance and battlefield intelligence. From lab to battlefield: How the spy cockroach works SWARM Biotactics' technology involves fitting Madagascar hissing cockroaches with ultra-lightweight, AI-enabled backpacks. These include: Tiny cameras for real-time reconnaissance Environmental sensors to detect gas, radiation, or heat Neural stimulators that send signals to the insect's nervous system to direct movement Wireless communication modules for operator control or swarm-level coordination This tech allows the bugs to slip through rubble, walls, or tight spaces where conventional drones fail—making them ideal for urban combat zones, hostage rescues, or disaster response, according to Reuters. by Taboola by Taboola Sponsored Links Sponsored Links Promoted Links Promoted Links You May Like Villas For Sale in Dubai Might Surprise You Villas in Dubai | Search Ads Get Info Undo Why cockroaches? A perfect spy in the shadows Cockroaches are small, durable, and capable of navigating cramped, complex terrain—traits that make them the perfect living platform for micro-surveillance. Unlike mechanical robots, they require no energy to move and can survive in extreme conditions, all while carrying payloads of up to 3 grams. Their movements can be guided using low-voltage impulses to the antenna or cerci, making them semi-autonomous with human override. Combined with AI, these biological machines are designed to act individually or as a swarm to map, monitor, or infiltrate hard-to-reach areas. Funded for the future: Germany's defense-tech push SWARM Biotactics recently secured €13 million in funding, including €10 million in a seed round aimed at transitioning this bio-robotic tech from laboratory research to field deployment. According to Reuters, the initiative aligns with Germany's broader effort to integrate cutting-edge startups into its defense innovation ecosystem, much like the U.S. DARPA model. This startup is also collaborating with security agencies and research institutions to develop protocols for field operations, ethical deployment, and swarm integration in live missions. While the primary use is for military surveillance and reconnaissance, the company also sees applications in disaster relief, such as locating survivors in collapsed buildings or chemical hazard detection in industrial zones. These AI-equipped insects could soon become part of firefighting, urban planning, and search-and-rescue operations. A new era of surveillance is crawling closer Germany's cockroach-cyborg initiative signals a new frontier in espionage and robotics—one that merges biology with artificial intelligence. While still in its early stages, the technology has already captured the attention of defense strategists worldwide. As development progresses, these living surveillance agents might soon redefine how wars are fought and crises are managed—one silent scurry at a time. AI Masterclass for Students. Upskill Young Ones Today!– Join Now
TechCrunch
22-07-2025
- Business
- TechCrunch
National security meets next-gen tech at Disrupt 2025 AI Defense panel
TechCrunch Disrupt 2025 is where breakthrough ideas meet the real-world challenges that define the future — and with over 10,000 startup and VC leaders converging, there's no better place to have the hard conversations. One of the most urgent? How artificial intelligence is reshaping national defense, security, and critical infrastructure in real time. Enter AI Defense, a can't-miss panel discussion taking place on one of the two AI Stages, where leaders from government, venture, and the armed services will explore the high-stakes collision of innovation and national security. Why this conversation matters now With AI evolving at a blistering pace, the stakes for defense innovation have never been higher. From autonomous systems and decision intelligence to cybersecurity and battlefield readiness, the U.S. defense and intelligence communities are racing to build smarter, more adaptive technologies — without compromising ethics, oversight, or safety. This panel explores the delicate balance between intelligence and control, the role of startups in defense innovation, and the multi-billion-dollar opportunities emerging at the intersection of national security and AI. If your startup touches defense, government, cybersecurity, or AI infrastructure, this is a conversation you can't afford to miss. Where defense, intelligence, and innovation converge Onstage, you'll hear from Dr. Kathleen Fisher, Director of the Information Innovation Office at DARPA, who is leading the charge on tech that gives the U.S. and its allies an edge in the information wars of the 21st century. With a career that spans AT&T Labs and academia at Tufts University, Fisher blends research credibility with strategic vision, and she's shaping the next generation of defense-grade AI. Joining the panel is Sri Chandrasekar, Managing Partner at Point72 Ventures, and formerly a leader at In-Q-Tel, the CIA's strategic investment arm. Chandrasekar knows how to spot frontier tech that moves the needle, and he's built investment frameworks to support some of the intelligence community's toughest missions. Techcrunch event Tech and VC heavyweights join the Disrupt 2025 agenda Netflix, ElevenLabs, Wayve, Sequoia Capital — just a few of the heavy hitters joining the Disrupt 2025 agenda. They're here to deliver the insights that fuel startup growth and sharpen your edge. Don't miss the 20th anniversary of TechCrunch Disrupt, and a chance to learn from the top voices in tech — grab your ticket now and save up to $675 before prices rise. Tech and VC heavyweights join the Disrupt 2025 agenda Netflix, ElevenLabs, Wayve, Sequoia Capital — just a few of the heavy hitters joining the Disrupt 2025 agenda. They're here to deliver the insights that fuel startup growth and sharpen your edge. Don't miss the 20th anniversary of TechCrunch Disrupt, and a chance to learn from the top voices in tech — grab your ticket now and save up to $675 before prices rise. San Francisco | REGISTER NOW Rounding out the panel is Justin Fanelli, Chief Technology Officer for the Department of the Navy. Fanelli lives at the intersection of cybersecurity, innovation, and adoption, driving digital transformation across one of the most complex organizations in government. From his DARPA Service Chiefs Fellowship to his nationally recognized work in defense health and command systems, Fanelli brings the boots-on-the-ground perspective to how defense is done in the age of AI. Catch it live at TechCrunch Disrupt 2025 This AI Defense panel takes place on the AI Stage. Exact session time coming soon — but don't wait to claim your pass. TechCrunch Disrupt 2025 runs October 27–29 at Moscone West in San Francisco. Join 10,000+ startup and VC leaders for three days of conversations and connections that define what's next. Grab your pass now before prices increase after July.
Yahoo
20-07-2025
- Politics
- Yahoo
Did the Defense Dept. cancel a grant to Harvard, then pay it anyway?
A new court filing Tuesday shed some light on an unusual wrinkle in the U.S. government's fight with Harvard University. Harvard previously claimed the U.S. Department of Defense had terminated a $3.4 million grant for important research into biological threats, despite pleas from an official to maintain the grant for national security purposes — but then asked for work to continue and paid the grant anyway. A Defense Department official issued a court statement this week saying the grant — supporting research for the 'AMPHORA' program, which stands for Assured Microbial Preservation in Harsh or Remote Areas — wasn't canceled after all. That's even though it was included in a list of terminated Harvard grants released in May. Efstathia Fragogiannis is director of the contracts management office with the Defense Advanced Research Projects Agency (DARPA) in the Department of Defense. Fragogiannis wrote that in the days following a May 12 letter to Harvard announcing the cancellation of that grant and many others, DARPA officials sought and received an exemption for the grant for reasons of national security. 'On May 21, 2025, May 22, 2025, and June 27, 2025, DARPA informed Harvard via email that the agreement remains active and that it should continue to perform work on the AMPHORA project,' Fragogiannis wrote. 'As such, Harvard has continued to perform work pursuant to that agreement, for which DARPA has paid, including the July 8, 2025, payment for work performed from May 1 to 31.' Harvard had claimed the request for continued work and the payment it received were evidence that 'reinforces' its court argument that 'the government's categorical terminations of research funding were arbitrary and capricious.' Harvard had said in a court filing they inquired with the federal government about whether the AMPHORA grant is in fact still active, but they haven't received a response. In a filing released Tuesday, the government argued that 'Harvard's grants were ultimately terminated because of Harvard's categorical refusal to respond to the Government's concerns' over antisemitism. A Harvard University spokesperson did not immediately respond to a request for comment. The court filings come as part of an ongoing lawsuit in which Harvard disputes cancellation of billions of dollars of funding by the U.S. government. It's one of two lawsuits the university has against the Trump administration, the other relating to its ability to accept foreign students. More Harvard University news Trump admin invokes Civil Rights Act in argument for Harvard funding cuts Harvard extends hiring freeze, says Trump actions could cost school $1B a year Trump admin halted Harvard grant, but Defense Dept. still paid it, court docs say A reckoning: Trump's attacks are inspiring self-reflection in higher ed Read the original article on MassLive. Solve the daily Crossword
Yahoo
20-07-2025
- Politics
- Yahoo
Trump admin halted Harvard grant, but Defense Dept. still paid it, court docs say
An unnamed Department of Defense official previously 'pleaded' with her superiors to keep paying Harvard University for a particular national security grant related to biological threats. And according to court documents, they have — despite the grant's termination. In a July 11 filing from one of its federal lawsuits against the Trump administration, Harvard informed the court it had received payment from the Defense Advanced Research Projects Agency (DARPA) — part of the Department of Defense — for the 'AMPHORA' program, which stands for Assured Microbial Preservation in Harsh or Remote Areas. With emerging infectious disease hotspots expected to increase globally within the next 50 years, the program seeks to develop a system without refrigeration 'that can preserve any microbe (bacteria, fungus, virus) from any sample, environment, or clinical setting, increasing the DoD's ability to surveil for emerging threats and protect force health and readiness,' according to the agency's website. Despite the grant being listed as terminated beginning May 12, DARPA has since asked Harvard's AMPHORA researchers to continue testing samples and to update on the project's progress, the court documents state. Then on July 8, the government paid Harvard approximately $373,000 on the AMPHORA grant for work conducted by its scientists May 1-31. Read more: Trump admin renews demand for Harvard foreign student info: 'We tried to do things the easy way' According to Harvard, the payment 'reinforces' its court argument that 'the government's categorical terminations of research funding were arbitrary and capricious.' 'This new evidence provides further support for Harvard's arguments,' Harvard wrote in its filing last week. Earlier court filings in Harvard's lawsuit disclosed that an unidentified Defense Department official had 'pleaded' with her superiors in the Pentagon not to cancel Harvard's AMPHORA grant because doing so could pose 'grave and immediate harm to national security.' The official said Harvard was the top-performing team in the AMPHORA program, which involves several teams from varying research institutions. Defense Secretary Pete Hegseth reportedly directed the cancellation of the military grants, and the notification stated the AMPHORA grant 'no longer effectuates the program goals or agency priorities.' Harvard has inquired with the federal government about whether the AMPHORA grant is in fact still active, but they haven't received a response, they wrote. A spokesperson for the Department of Defense said the agency does not comment on ongoing litigation. More Harvard University news A reckoning: Trump's attacks are inspiring self-reflection in higher ed Harvard continues dismantling its DEI offices amid Trump attacks Trump admin renews demand for Harvard foreign student info: 'We tried to do things the easy way' Trump admin threatens Harvard's accreditation over antisemitism response Read the original article on MassLive. Solve the daily Crossword
