Categories
Artificial Intelligence

Artificial intelligence accurately predicts cardiac arrest death, researchers say | UPI  

Artificial intelligence-based technology forecasts cardiac arrests months in advance. This ground-breaking technology has the potential to save countless lives. 

A novel technique based on artificial intelligence and heart scans of patients can forecast if they will die of cardiac arrest. The method, termed Survival Study of Cardiac Arrhythmia Risk, develops a tailored survival estimate via neural networks. Arrhythmia-related sudden cardiac mortality accounts for up to 20% of all fatalities globally. Cardiac arrest occurs when the heart stops appropriately pumping blood, resulting in an abrupt cessation of blood flow throughout the body. According to a study, many persons who have a cardiac arrest fell ill days before the event. 

Source: https://www.upi.com/Health_News/2022/04/07/algorithm-predict-cardiac-arrest-death-study/2701649337182/  

Categories
Artificial Intelligence

Musk wants to start selling Tesla’s A.I.-powered humanoid robot next year, but his A.I. chief just went on sabbatical | Fortune

According to Elon Musk, a functioning prototype of the Optimus robot might be available by the end of next year, according to the German daily Welt am Sonntag. Tesla looks to be reinventing itself as an artificial intelligence firm first and a carmaker second as part of the company’s next stage of growth—Master Musk’s Plan Part 3. Tesla Vision’s director, Andrej Karpathy, takes a four-month break from his day job. The resignation comes as beta testers anticipate the release of Version 11, which will integrate city and highway driving software. 

IN JANUARY, Tesla CEO Elon Musk requested that Model 3 purchasers pay $12,000 for FSD software. Musk may have sent a subtle signal downplaying Karkarathy’s value during an interview with artificial intelligence specialist Lex Fridman. 

Source:https://fortune.com/2022/03/28/musk-start-selling-tesla-ai-powered-humanoid-optimus-robot-andrej-karpathy-sabbatical/

Categories
Artificial Intelligence

AI suggested 40,000 new possible chemical weapons in just six hours | TheVerge

Open-source toxicological datasets have the potential to expose real-world chemical hazards.

A drug-discovering AI created 40,000 potentially dangerous compounds in less than six hours. Several compounds resembled VX, the most potent nerve toxin ever made. VX paralyzes your lungs by immobilizing your diaphragm or lung muscles. The North Korean leader’s half-brother, Kim Jong-Nam, was assassinated in Malaysia after being poisoned with VX nerve toxin. Someone could easily mimic what we accomplished.

Not only for VX but for almost any other open-source toxicity dataset available. If someone were to put this together without any knowledge of chemistry, they would end up with unusable material. There are still some significant jumps between what the AI generates and transforming it into a real-world danger.

Source: https://www.theverge.com/2022/3/17/22983197/ai-new-possible-chemical-weapons-generative-models-vx

Categories
Artificial Intelligence

The Vulnerability of AI Systems May Explain Why Russia Isn’t Using Them Extensively in Ukraine | Forbes

The fact that Ukraine is employing face recognition software to track down Russian attackers and identify Ukrainians slain in the current conflict is significant since it is one of the few known applications of AI in the fight. Ukraine’s military ministry started utilizing Clearview AI’s image software on Saturday after Clearview supplied it for free, according to the CEO of the controversial American face recognition startup Clearview AI. Artificial intelligence algorithms and a large amount of picture training data gathered from social media and the internet are said to be at the heart of the very sophisticated identification tool. Apart from Russian influence efforts with their much-discussed “deep fakes” and misinformation-spreading bots, many observers have been astonished by the Russian military’s lack of known tactical usage (at least publicly) of AI. 

Source: https://www.forbes.com/sites/erictegler/2022/03/16/the-vulnerability-of-artificial-intelligence-systems-may-explain-why-they-havent-been-used-extensively-in-ukraine/?sh=40365fdd37d5 

Categories
Artificial Intelligence

AI can help historians restore ancient texts from damaged inscriptions | The New Scientist

An artificial intelligence program built by historians and DeepMind, a UK-based AI company, can assist reconstruct ancient Greek writings with 72 per cent accuracy. The AI can also pinpoint with more than 70% accuracy where the writings were written in the ancient Mediterranean world and date them to within a few decades of their agreed-upon origination date. All of this is an advancement above the AI’s previous iteration, which could only reconstruct old writings. Ithaca could also, 71% of the time, determine where in the Mediterranean a work was written, and it could date the manuscripts to within 30 years of their original creation date, as historians have previously proven. 

Source: https://www.newscientist.com/article/2311657-ai-can-help-historians-restore-ancient-texts-from-damaged-inscriptions/