Artificial Intelligence- A Highlight

Last month I attended a Seminar on Artificial Intelligence conducted by The Climber under Indian Youth Conclave held in Chennai. So here are some highlights and minutes of the seminar.

So to start with let's define the topic first.

Artificial intelligence (AI, also machine intelligence, MI) is apparently intelligent behaviour by machines, rather than the natural intelligence (NI) of humans and other animals. In computer science, AI research is defined as the study of "intelligent agents": any device that perceives its environment and takes actions that maximize its chance of success at some goal. Colloquially, the term "artificial intelligence" is applied when a machine mimics "cognitive" functions that humans associate with other human minds, such as "learning" and "problem-solving".

In Simpler words, AI is nothing but a machine with common reasoning capabilities.

Things to keep in mind:

1. The growth in technology is exponential.

2. Computation power doubles every 18 months.(Moore's Law)

Scenario before AI:

• Navigation was done with the help of compass and maps.
• Translations were done with the help of a dictionary.
• Email folders were overloaded with spams.

What is AI?

• Recreating something which can make a machine do human tasks.
• Computers doing work which requires human intelligence.

Some Raw Applications:

• NLP - Natural Language Processing
• understanding human language as it is naturally spoken i.e. understanding instructions in English or any other Human understandable language.

• Speech Processing -
• Speech Recognition - Speech to text
• Speech Synthesis - Text to Speech

• Computer Vision -
• image processing
• face recognition
• giving the computer an ability to see

• Knowledge Representation -
• possession of knowledge
• making knowledge useful for humans
• a good demonstration could be Wikidata - Wikidata is a free and open knowledge base that can be read and edited by both humans and machines. 

• Machine Learning -
• making a machine learn something through a process of trial and error.
• learning through experiences and dataset synthesis.

• Supervised Learning -
• learning through algorithms which are capable of upgrading
• closest path and most efficient algorithm
• Answers and variables are already given, the computer just needs to work on the methods to devise questions.

• Unsupervised Learning - 
• When final answers are not given.
• the data is clustered into groups so that some conclusion from the data could be drawn.
• For ex- Recommendations from Amazon is unsupervised learning.

• Deep Learning -
• Working on some specific feature by the use of available counterparts
• It's exactly same as human brain's functioning.
• It consists of breaking information into smaller pieces layer by layer on neural network.

 

The scenario with AI:

• gmail now filters up to 99% of spam emails.
• Each and every gadget is AI factory.
• Siri, Google Assitant, Alexa, Cortana all are AI based internet deduction applications.
• Google Translate made it really easy to translate text just by typing or clicking a pic too.
• Amazon Recommendations are based on AI grouping.
• Chess, Go, Othello, Backgammon, etc. are now based on an AI-powered engine developed by Google DeepMind

.

AI vs Jobs:

• depending upon different fields influence of AI on jobs can be felt.
• Will Robots Take My Job?- This website predicts the job security based on recent trends in AI influencing your job.
• Some recent AI takeovers:
• Robotic warehouses allow Amazon to offer super-fast shipping.
• Multiple companies are using drones for delivery.
• Robot authors novel in Japanese Novel Writing Competition.
• Sunspring- A short movie entirely scripted by AI.
• Robotic Lawyers overturned 160000 parking tickets already in opposition to Equifacts SSN security breach.
• Self-driving cars.
• More than 50% trade in the US is done by AI. So it is not wrong to say that the Economy of the world is powered by AI.
• MAR/IO
• Mar/Io is AI played version of the classic Mario. It is entirely based on predictions by an AI
• Here is the Video.

Future Scenario with AI:

We are at the Edge of change comparable to the rise of Human Life on Earth

       1. The Law of Acceleration Returns:

• A revolution like the Industrial revolution is coming back.
•  The Law states Business Insider

       2. Death level of Progress:

• It's basically the limit of human sustenance at sight of technological advancement.
• For ex - If a person from Industrial Revolution will see today's Technology he/she would be so amused that they would die of shock whereas if any person from early 1960s will see today's technology he/she won't die out of amusement.
• It is basically a time period in which people will die out of amusement due to technology.
• Next 20 years are very crucial. 
• It is estimated that the Death Level would come down to few weeks.

       3. Some future Advancements in developing stages:

• 2D to 3D conversions.
• Image Captioning (See this)
• Face Hallucination (See this)
• Image Style Transfer - Applications like Prisma and others for enhancing photographs using AI. (See this) 
• Image Colorization - colouring black and white photos using a neural network and deep learning. (Try it)
• Computer Programming.

       4. AI Caliber

• Artificial Narrow Intelligence
• Not much Machine and Deep learning involved. (See this)
• Artificial General Intelligence 
• Includes hardcore Deep Learning specific to the field (See this)
• Artificial Super Intelligence
• High-end algorithm processing tendency (See this)
• Neuralink
• An Elon Musk company targeting controlling computers just with brainwaves by inserting a bionic chip inside the brain. (Neuralink)

Some Fun facts regarding AI:

• We consider AI as not even intelligent than chimp but in reality, AI has surpassed even Einstein.
• AI will either cause Human extinction or make humans immortal.

Some unattainable AI responses yet:

• Emotional stability.
• Imparting knowledge about good and bad.
• Surety for human welfare.
• Self-improving AI at large scale.
• Rogue AI rectification.

At last, I would like to leave you all with some unanswered question...

1. Can AI be a Darwinian Error?
2. Will AI never overpower humans in spite of being superior to humans?
3. Can we tackle rogue AI?
4. Is AI threatening justified?