The world has seen a boom in the field of Artificial Intelligence in the past few years. The major reasons contributing to this is the availability of data and computing power.
a) Data: Huge data collected by different companies, IoT devices, etc
b) Computation: Availability of AWS, Google Cloud, NVIDIA GPUS, etc
A lot of research has happened in the field of AI in the last decade and society has witnessed many amazing use cases. In the last decade, AI went mainstream because of the availability of hardware, courses, platforms, big companies taking workshops, etc. What our AI community has achieved in the last decade has set a strong foundation for the future.
Let us try to understand the different breakthroughs made in this field year by year:
a) Foundation of DeepMind
DeepMind Technologies is a UK artificial intelligence company. It was co-founded by Demis Hassabis and had a big impact in the field of AI.
b) ImageNet Competition
ImageNet competition brought a lot of development in the field of computer vision. It is a competition designed for research in object recognition. It contains 14 million images hand-annotated for 20,000 categories like a balloon, chair, etc. It is held annually. The main goal of the competition is to reduce the error rate in identifying categories like chairs from the image. It was the brainchild of the renowned professor Fei-Fei-Li.
a) IBM wins Jeopardy
IBM Watson is a question-answering computer system. It beats two former Jeopardy champions. This was a big leap of AI in Natural language processing. It shows AI reached the level of understanding the different intricacies of human language.
a) Convolutions Galore
In one of the competitions of German Sign language recognition, AI build with the help of CNN’s beats the Human accuracy of 99.2 % The accuracy which AI achieved was 99.46%. A similar groundbreaking feat was achieved in the ImageNet competition where the University of Toronto reduced the error rate from 25% to 16%. It was this period when AI was getting better than humans in computer vision tasks.
a) Google Glass
Google Glass is a brand of smart glasses. It was developed by X with the mission of producing a ubiquitous computer. It is an optical head-mounted display in the shape of glasses. This created a pavement for a new hardware device that will be collecting data. Glass joined the companies of Laptops and mobiles. Today, along with them we have other data collecting partners too like AI voice assistants, Smartwatches, and lots of other sensory data collectors.
b) Boston Dynamics revealed Atlas
Atlas was revealed to the public by Boston dynamics. The future is robots being a family member. There are few like Alexa, Mop P who have already entered our houses and there are many who will be joining the family in the future. Boston dynamics today is a pioneer in the field of Robotics and it all started with the popularity of Atlas.
GANs is the baby of AI pioneer Ian Goodfellow. It is the next big thing in the field of computer vision. In simple terms, GANs generate new data from the existing one and has remarkable applications in the major industries. All the aging apps, reface apps,cartoonifying your images are the different applications of GANs.
b) Google acquires DeepMind
The acquisition of Deepmind by Google is a landmark moment in the field of AI. The rise of Google in AI after this acquisition is noteworthy. Facebook too was in the race but Google won it at a whopping amount of 400 million euros.
a) Tensorflow release
Tensorflow is an end to end open-source platform for machine learning. It is very popular because of its big community, lots of packages, tools, etc. The release of Tensorflow made deep learning go mainstream. Lots of people started trying their hands-on deep learning and many different applications and use -cases were seen growing.
b) Open AI founded
Elon Musk founded Open AI. Their work in the field of deep reinforcement learning is remarkable. Today Elon is one of the biggest drivers in the field in AI and Open AI is a big name in the AI community.
a) AlphaGo beats World Champion Lee Sedol
Go, a Chinese game is more complex than Chess. It involves lots of strategies, techniques. An AI learning all these techniques and beating a current World Champion 4/5 is a monumental achievement. There was one very interesting moment in the match when AlphaGo made a move and seeing it the world champion Lee Sedol was so perplexed that he left the arena.
b) Sophia Robot
Sophia is a robot developed by Hong-Kong based company Hanson Robotics. She became the first robot to get the Citizenship of Saudi Arabia. She can imitate facial expressions and engage in conversation with other humans.
c) Release of Pytorch
Pytorch is developed by FAIR(Facebook’s AI research lab). It is an open-source framework for building AI models. Currently, it is the most popular framework for deep learning.
a) AI developing its own language
FAIR i.e Facebook’s AI research lab trained two AI chatbots to converse with each other. While conversing those chatbots deviated from the primary English language and created their own language. Although the project was shut down, this incidence shows the AI reaching the human general intelligence.
To enable AI frameworks interoperability, many big companies like Facebook, Google joined hands and launched ONNX – the interoperability ML framework.
a) Self-driving car – Waymo
Waymo is a company that works on self-driving cars. With 10 million miles on public roads and 7 billion in simulation, they are on their way to make the most experienced driver.
b) Google’s BERT
BERT stands for Bidirectional Encoder Representations from Transformers. It can be used in solving many language tasks like questions – answering, sentence completion, etc.
With GANs going mainstream, people have started creating fake videos of famous personalities. It is very difficult to identify just by seeing if the video is real or fake. The technology is called DeepFakes.
a) AI beats Human Doctors in Cancer identification
Google researchers have created an AI system in detecting lung cancer and their accuracy is better than doctors.
b) Open AI beats DOTA2 esports world champions
Dota2 is not an individual game but a team game. The AI needs to coordinate with other AIs to beat any player in DOTA2. The factor of AI coordination and the win over the world champions highlights the real achievements of it.
The last decade set up a strong foundation for the coming years and as Andrew Ng stated, “AI will be the new electricity and will impact the lives of every individual in the coming years.”
Did you like this blog? Please let us know in the comments section! Additionally, you can read our AI / ML Blogs here.