The Evolution of Artificial Intelligence
The tech we use nowadays is the outcome of a number of artificial intelligence milestones achieved by many unsung heroes. We travelled back in time and compiled a list of all significant artificial intelligence achievements that have allowed us to enjoy our current lifestyle. Let’s take a look at AI evolution over time. The timeline below highlights AI milestones that have been achieved throughout history. These achievements can be found in all areas of Artificial Intelligence. Because this is a growing timeline, new milestones will be updated on a regular basis.
Timeline
Facebook’s AI To Stop Suicide
Facebook’s AI To Stop Suicide plays proactive in detecting suicidal status posts using AI, especially Natural Language Processing (NLP), to scan and point out negative phrases of desperation and hopelessness by users. This AI was given a huge number of negative and false positive phrases, and only negative emotional phrases..Read more
Google Home Mini, A Smart Voice Assistant For Home
Google Home Mini is a smaller yet advanced version of Google Home Speaker with a reasonably lower price, i.e., $49 versus $129. It is a voice-controlled speaker that can perform multiple tasks like playing music, controlling smart home gadgets, managing to-do-lists, scheduling calendars, playing videos on Chromecast-enabled screens, adding items..Read more
An Algorithm AlphaGo Was Created By Google
Google DeepMind team developed an algorithm AlphaGo to beat the game GO’s top competitive players (the ancient far-eastern board game). AlphaGo program was trained using Monte Carlo Tree Search algorithm that finds moves stored in it previously. All the moves made in winning games were recorded and trained both by..Read more
Intel Movidious Launched First Neural Compute Stick (NCS)
Movidius, an Intel subsidiary, launched its first-ever Neural Compute Stick (NCS). NCS is the world’s first AI accelerator contained within a USB flash drive. This is used to run deep neural network processes natively (on edge). Otherwise, most neural network operations require a central server with huge computing power. NCS..Read more
BWIBots-The Visionary Robots
The researchers developed these visionary robots that leaned human performances. They worked with humans to learn how to cooperate during a particular task. BWI is Building-Wide Intelligence. The main purpose of this project was to design certain attributes in robots to train them for daily in house tasks, based on..Read more
Caffe2Go: An AI For Graphics Deployed By Facebook
Facebook uses an open-source deep-learning program, “Caffe2Go,” for image processing and real-time video editing. This AI can give your videos artistic an creative look. This application is available only on iOs and Android smartphones. Although your videos are transformed, they are sent to data centers for processing; this application can..Read more
Nanorobots developed by the team Drom Polytechnique of Montreal
The team created these nanorobots as a transporter-bot that could administer drugs and save surrounding organs and tissues. In a study, these nanorobots were used to deliver an anticancer drug to oxygen-depleted cancerous cells. These nanorobots were manufactured using 100 million flagellated bacteria with a compass. Anticancer drugs damage normal..Read more
OpenAI Released OpenAI Gym For Reinforcement Learning
The OpenAI gym is a platform that allows you to create programs that attempt to play a variety of video game-like tasks. This is often applied to reinforcement learning in Artificial Intelligence (AI). This platform is used to develop and compare reinforcement learning algorithms. It helps AI agents to learn..Read more
Andrew Dai Introduced Semi-Supervised Sequence Learning
Andrew Dai from Google Brain demonstrated a unique language model approach known as Semi-Supervised Sequence Learning. Basically it was a dual approach model that can train a language model with both supervised and unsupervised learning. First apprach was to predict upcoming word in a sequence, which is conventional approach in..Read more
The First Digital Reconstruction Project of the Somatosensory Cortex
The microcircuitry of the somatosensory cortex of the juvenile rat was done first time digitally. This Digital Reconstruction Project was a combined effort of the Blue Brain Project & the Human Brain Project. It was based on the detailed study of cortical dynamics, in-vitro reproduction, and in-vivo experimental results. Read..Read more
First Robotic Kitchen Debuted In Germany Trade Fair
MoleyRobotics is the company behind the idea of automated kitchen chefs. Robotic Kitchen is a robot set (RoboChef) with technical hands, an electric stove, oven, dishwasher, and a touch screen unit. It can perform multiple kitchen tasks and handle other kitchen equipment. The total estimated cost is $75000, but it..Read more
Neural Turing Machine Introduced By Alex Graves
Neural Turing Machine is a recurrent neural network model. This model employs a pattern matching mechanism of neural networks and processing power from programmable computers. It also incorporates some external memory devices. In simple words, Alex created machines’ short term memory. Read more December, 2014
Amazon Launched Its First AI Mediated Virtual Assistant Alexa
Alexa was first used in Amazon Echo smart speakers. It is an amazing program that can interact with voice commands. Plus, it can be used as a personal assistant to play/stop music, tell weather forecast, and set the alarm. This virtual assistant can also be used for home automation. Read..Read more
TrueNorth: the first Neuromophic integrated circuit was built
TrueNorth, IBM & SyNAPSE prepared the first neuromorphic integrated circuit to get one million individually programmable neurons. Plus, they had 256 individually programmable synapses. Synapse is the junction of real brain’s neurons and the human brain has more than 200 trillion synapses. In TrueNorth, artificial neurons are emulated using a..Read more
Abhinav Gupta from Carnegie Mellon University published the Never Ending Image Learner (NEIL)
Never Ending Image Learner (NEIL) is an online computer program that works nonstop to learn every aspect and relationship between and between images, respectively. It scans the images for core objects and their relationships. For example, A boy has eyes, ears, and hair. And boys have shorter hair than girls...Read more
HRP-2 Robot Built By Schaft Won DARPA’s Robotics Challenge Trials
Schaft Inc, a Google subsidiary of Japan, built an HRP-2 robot in DARPA’s Robotics Challenge Trials, held in Miami. 16 teams with their robots participated in this competition for eight different tasks crucial for disaster response. Those tasks were driving a vehicle, walking over an uneven waste surface, climbing a..Read more
Splinter Cell Blacklist: Top Rated AI Mediated Video Game Was Released
Splinter Cell Blacklist is third-person stealth and shooting game. Developed by UBI Soft, this game uses artificial intelligence in interesting manners. Your opponents (guards) note/ record your moves using deep learning and respond according to your move. In previous games, the opponents were not evolved according to your expert level...Read more
World’s First Talking Robot Astronaut, Kirobo
It was developed by the University of Tokyo, Toyota, and firm Dentsu. The talking robot waited for the arrival of astronaut Koichi Wakata. Kirobo showed emotions when he met Wakata and provided insights to researchers that machines could also be introduced with emotions. Plus, machines can also offer emotional support..Read more
Google Launched Quantum Artificial Intelligence Lab
Running machine learning algorithms is a complex and processing-rich task. Common computers are not suitable for such a burden. There is where quantum computers are required. NASA collaborated with Google and Universities Space Research Association to establish Quantum Artificial Intelligence Lab for advanced and deep computing. Quantum AI lab used..Read more
Google Released Google Now, A google Search Feature
Google Now is a search feature for android and iOS. The user’s search and preference data delivered its user with predicted information in the form of informational cards. It used information from users based on their habits and surroundings, like location, browsing history, contacts, etc. Google gradually dismissed it, but..Read more
AI Recognized The Cat From Totally Unsupervised And Unattributed Image Data
Jeff Dean and Andrew Ng from Google established a neural network of 16000 personal computers. All these processors were given 10 million unlabelled pictures as a training set from youtube videos screenshots. After running neural network algorithms, the AI recognized the cat from the picture without knowing it was a..Read more
A Robotic Arm of BrainGate System Controlled By The Mind Of Paralyzed Patients
The robotic Arm of BrainGate System was the most advanced innovation in the robotic industry in the year 2012. The paralyzed patients were introduced with a 4 millimeter-wide chip implanted in their heads. With it, they could control and command the robotic arm. The computer used in this system decodes..Read more
Spaun- The First Computer Model To Produce Complex Behavior
Spaun- The First Computer Model To Produce Complex Behavior was done by the University of Waterloo, Canada. Its engineers prepared a model that could lead to human performance in simple tasks. It was modelled with the human brain to function in a realistic way biologically. Nengo platform was used to..Read more