AI is evolving faster than you think…
The controversy around the FaceApp and AI strides in Technology, Healthcare, Education & more
I usually cover the developments in the field of Artificial Intelligence as part of my regular feature titled Tech Diaries, but this week was so jam-packed with AI news that I had to do write a separate piece. One story that stood out from the rest was of the AI-enabled photo-editing App called FaceApp. It is owned by Russia-based Wireless Lab and has been around since 2017, but with the recent addition of a feature which lets you see your future self made it go viral last week. My Facebook newsfeed was full of people showing off their ‘Now & Then’ pictures — Looked enticing but I held off the temptation to generate an older version of myself.
So what’s the problem then? For starters, some reports suggest that the App has collected more than 150 million photos of people’s faces since its launch & according to its terms of service it can use the huge database in whatever way it wants. With the issue of Deepfakes already plaguing the social media and thinning the lines between right & wrong, this could further complicate issues. However, the company has clarified in a statement that it deletes most images from its servers within 48 hours & doesn’t share data with third parties.
And besides, companies like Google and Flickr & face-editing Apps like FaceTune have accumulated far bigger databases of photos without coming under any scrutiny. This is where the fear-mongering of the American Political machinery kicks in — Any new technology which has a nativity to Russia or China these days tends to spook them, as some Democratic members of the U.S Congress are now demanding an FBI investigation of the company. Similar hearings on the issues of Deepfakes have already been initiated.
Even if the intentions on part of the companies are malicious & lawmakers are able to censor them, these databases will keep cropping up with the abundance of public media on the Internet. These Apps just highlight the fact of how willing we are to lose control of our digital data. With that, let’s move on to see what else AI has been upto.
The only knowledge I had on the topic was from the Sci-Fi series ‘Star Trek,’ where the Alien races ‘Klingons & Romulans’ had the ability to cloak their space ships. Thanks to AI now, researchers are close to engineering a garment with unnatural properties, giving you the ability to move around undetected.
These so-called metamaterials are designed from artificial atoms and have specific structural properties. Earlier in March, Boston University researchers created a new material by combining Math with 3D printing — light & air could pass through it but the sound couldn’t. Now South Korean researchers have developed an AI which can design new metamaterials with specific optical properties detailed here.
Chip giant Intel has just unveiled new AI chips that can process data 1,000 times faster than the conventional chips. The project dubbed as Pohoiki Beach is a system that contains 64 of its Loihi AI processors — set of neuromorphic chips which seek to imitate the brains with their learning ability & energy efficiency. Use cases of these chips include improved maneuvering of prosthetic limbs on uneven ground & creation of accurate digital maps for autonomous vehicles. The new hardware platform expects to produce a system capable of simulating 100 million neurons by the end of 2019.
Smart Algorithms for Medical Discovery
We have seen great advancements in Healthcare over the past few decades, but researchers have to spend tons of money & lots of time to develop new drugs with a high failure rate. Developers have been experimenting with the use of Artificial Intelligence to discover new drugs. An AI developed by researchers at Australia’s Flinders University called Smart Algorithms for Medical Discovery (SAM) has produced a promising new flu vaccine by itself.
Who amongst us hasn’t tried his hands on Rubik’s Cube — the multicolored, three-dimensional puzzle which has perplexed so many over the past five decades. An AI system developed by the University of California at Irvine and named as DeepCubeA solved the puzzle in just over one second, beating the current human record by more than two seconds.
The AI utilizes a reinforcement-learning algorithm programmed by scientists and mathematicians. The most impressive part was that DeepCubeA solved the puzzle without any prior game knowledge or coaching. The AI solved the puzzle in only 20 moves, as compared to the 50 moves that a highly skilled human makes to solve it.
AI Immersive Classroom
Most of us know that the best way to learn a language is by speaking or being in an environment where others are speaking the language. IBM Research & Rensselaer Polytechnic Institute (RPI, New York) have now collaborated to bring a virtual environment to the classrooms by employing AI & VR technologies. Students get to haggle with the street vendors in Beijing or order food in restaurants in real-time. The experimental course at the University attempts to understand the effects of cognitive and immersive environments on learning.
While the debate on morality & Ethics of Artificial Intelligence drags on, the capacity and capability of AI continue to grow exponentially.