The Chinese government has declared that AI produced news will be required to disclaim the fact that it is indeed AI manufactured. This will make fake news produced by AI a criminal offence.
The new law will come into affect January 1, 2020. The main concern stemming from the fake news generated by AI is the production of "deepfakes." Or, in other words, the merging and splicing of video and audio to produce news that never occurred. This direction is credited in large part to the presidential election of the U.S in 2016, where the emergence of fake news caused shock waves across the globe.
The Al Jazeera Article
The State of Current Affairs by Hector
Sunday, December 1, 2019
Wednesday, November 13, 2019
The "DOOM" Guy's AI Passion Project
Legendary coder and revered video game developer John Carmack has stated that he has begun a project focused on the development of AGI: Artificial General Intelligence.
AGI is a type of AI that learns and acts much like normal humans do. In other words, Carmack is pursuing an AI capable of learning, problem solving, and behaving much like us. If images of the classic "the AI has gone rogue" trope just flashed in your mind then you are on the money, except hopefully Carmack will lean on the side of the "good" kind of AI like R2D2. Given Carmack's prowess in the realm of video games and virtual reality, if anyone can make headway in such an endeavor I believe it to be him. Time will tell if such a journey will yield the advancement of mankind or its destruction.
The TechCrunch Article
AGI is a type of AI that learns and acts much like normal humans do. In other words, Carmack is pursuing an AI capable of learning, problem solving, and behaving much like us. If images of the classic "the AI has gone rogue" trope just flashed in your mind then you are on the money, except hopefully Carmack will lean on the side of the "good" kind of AI like R2D2. Given Carmack's prowess in the realm of video games and virtual reality, if anyone can make headway in such an endeavor I believe it to be him. Time will tell if such a journey will yield the advancement of mankind or its destruction.
The TechCrunch Article
Wednesday, November 6, 2019
Only AI Can Prevent Wildfires
A San Francisco-based AI company, Chooch AI, has begun developing an AI program that would be capable of identifying fires before they even have a chance of spreading.
The current issue with fire prevention is that it is often too slow. Oftentimes by the time a fire is reported it is too late, as the fire has had the opportunity to gain traction making the situation even more difficult. What Chooch AI is attempting to develop is an AI capable of analyzing satellite imagery to determine where a fire has the potential of starting. It would do this via infrared to detect signs of smoke and heat, as well as identify where certain plots of vegetation are too close to power lines. Another goal is to enable this program to be able to constantly and regularly make scans of California's vast forest ranges. The technology is expected to be at a usable phase by next fire season. With over $25 billion in damages caused by the wildfire in just these past few weeks, Chooch AI could provide the solution to this epidemic.
The Washington Post Article
The current issue with fire prevention is that it is often too slow. Oftentimes by the time a fire is reported it is too late, as the fire has had the opportunity to gain traction making the situation even more difficult. What Chooch AI is attempting to develop is an AI capable of analyzing satellite imagery to determine where a fire has the potential of starting. It would do this via infrared to detect signs of smoke and heat, as well as identify where certain plots of vegetation are too close to power lines. Another goal is to enable this program to be able to constantly and regularly make scans of California's vast forest ranges. The technology is expected to be at a usable phase by next fire season. With over $25 billion in damages caused by the wildfire in just these past few weeks, Chooch AI could provide the solution to this epidemic.
The Washington Post Article
Wednesday, October 30, 2019
StarCraft II Pro
An AI by the name of AlphaStar has just successfully become a Grand Master at the video game StarCraft II.
The game of StarCraft II has often been dubbed as one of the most complex video games out there. Easy to understand yet incredibly difficult to learn how to play like a pro, AlphaStay grew to its current level of ability in just 44 days. It has beat 99.8% of all of its opponents since its inception. The important thing to note is that the AI was able to grow to its current level by "watching" recordings of other players and constant practice against itself to iron out any weaknesses in its game play. The broader hope for such a program is that said technology could be applied to predicting the weather and self driving cars.
The Guardian Article
The game of StarCraft II has often been dubbed as one of the most complex video games out there. Easy to understand yet incredibly difficult to learn how to play like a pro, AlphaStay grew to its current level of ability in just 44 days. It has beat 99.8% of all of its opponents since its inception. The important thing to note is that the AI was able to grow to its current level by "watching" recordings of other players and constant practice against itself to iron out any weaknesses in its game play. The broader hope for such a program is that said technology could be applied to predicting the weather and self driving cars.
The Guardian Article
Wednesday, October 23, 2019
Robo Builders
Currently, the assembly of vehicle parts and the construction of architecture has been either done in separate parts of the world or susceptible to human error. In an article published in Science Daily, researchers at MIT have been hard at work in producing a brand new "builder bot" of sorts. The robot is dubbed BILL-E (Bipedal Isotropic Locomoting Explorer).
BILL-E is capable of erecting structures without the need for a human to intervene. The human gives the robot a plan that it is capable of building and then it handles the rest. The hope is that it might be able to stop the current practice of certain structures needing to be assembled in separate locations. An example would be a commercial airliner where its wings may be constructed an assembled in one area, but the fuselage in another. The hope for BILL-E is that it may be able to stop this practice, which could positively impact the environment as it would eliminate the need to transport the separate pieces for final assembly.
Another hope is that a robot like BILL-E will be able to build human-livable structures on Mars. It would eliminate the need for humans to begin building as soon as they arrive on the Red Planet and instead have some sort of established base to begin whatever research they are there for.
The Science Daily Article
BILL-E is capable of erecting structures without the need for a human to intervene. The human gives the robot a plan that it is capable of building and then it handles the rest. The hope is that it might be able to stop the current practice of certain structures needing to be assembled in separate locations. An example would be a commercial airliner where its wings may be constructed an assembled in one area, but the fuselage in another. The hope for BILL-E is that it may be able to stop this practice, which could positively impact the environment as it would eliminate the need to transport the separate pieces for final assembly.
Another hope is that a robot like BILL-E will be able to build human-livable structures on Mars. It would eliminate the need for humans to begin building as soon as they arrive on the Red Planet and instead have some sort of established base to begin whatever research they are there for.
The Science Daily Article
Tuesday, October 8, 2019
Computer Chip Showdown
The United States is the global leader in the production of computer chips and semiconductors. In a recent move that seeks to undermine the authority of another global powerhouse, the U.S has decided to stop selling chip components to several Chinese computer companies, specifically those in the AI development sector such as voice and face recognition software companies. This move is part of the larger power struggle occurring between both nations. China does seem to be in trouble as they domestically produce just 16% of the chip components they use. Beijing has noted that its goal is to increase that number to 40% by the end of next year and to 70% by 2025.
Despite the lofty goals set, China's AI sector could suffer a set back. Over the recent weeks, Chinese tech giants Huawei and Alibaba have unveiled new computer chips meant to increase the speed of machine learning. The further implication of such a development is the impact on the military technology sector in China. Looking like a preventative measure more than anything, the U.S seeks to halt the production of AI military software that could pose a threat later on. As much as it is a power move to constrict AI development in China, it is also a defensive measure to reduce China's geopolitical power by attacking its AI development.
The BBC Article
Despite the lofty goals set, China's AI sector could suffer a set back. Over the recent weeks, Chinese tech giants Huawei and Alibaba have unveiled new computer chips meant to increase the speed of machine learning. The further implication of such a development is the impact on the military technology sector in China. Looking like a preventative measure more than anything, the U.S seeks to halt the production of AI military software that could pose a threat later on. As much as it is a power move to constrict AI development in China, it is also a defensive measure to reduce China's geopolitical power by attacking its AI development.
The BBC Article
Wednesday, September 25, 2019
AI: As Good as Medical Professionals?
The U.K is currently supporting a 250 million GBP investment in AI technology focused on accurately diagnosing and analyzing images and scans through the NHS, National Health Services.
What the technology actually does is interpret medical imaging to give a proper solution, as accurate as possible, on what actions may be taken to resolve the issue. Currently, the AI programs are just as good as its human counterparts. The issue that many on the opposition propose is that there is simply not enough studies or experiments done to fully understand what this technology is actually useful for. It should be noted that the side-by-side comparison between the doctors and the AI were using images of patients in which the doctors had no prior knowledge of and were nearly as good as the AI in giving a proper prognosis.
The technology is still in its infancy, but the future may allow for more personalized, efficient relationships between the doctors and their patients.
The Guardian Article
What the technology actually does is interpret medical imaging to give a proper solution, as accurate as possible, on what actions may be taken to resolve the issue. Currently, the AI programs are just as good as its human counterparts. The issue that many on the opposition propose is that there is simply not enough studies or experiments done to fully understand what this technology is actually useful for. It should be noted that the side-by-side comparison between the doctors and the AI were using images of patients in which the doctors had no prior knowledge of and were nearly as good as the AI in giving a proper prognosis.
The technology is still in its infancy, but the future may allow for more personalized, efficient relationships between the doctors and their patients.
The Guardian Article
Subscribe to:
Posts (Atom)