IIT Madras Offers Post Baccalaureate Fellowship In Artificial Intelligence & Data Science

By

-

June 17, 2021

The Indian Institute of Technology(IIT) Madras is offering a fellowship for post baccalaureates in various areas of artificial intelligence and data science for 2021 at the Robert Bosch Centre for Data Science and Artificial Intelligence (RBCDSAI) for 1-2 years.

RBCDSAI at IITM was founded in 2017 with 47 researchers on board who are among the best in the artificial intelligence and data science landscape. The fellowship opportunities started in 2019. RBCDSAI has increased its candidate intake from 4 to 15 in the last two years, paving the way for graduates interested in the field of research and development from deep learning and networking to endorsing artificial intelligence in transport, agriculture, and health care facilities. With the increasing trend in the intake, it is likely to have more openings in 2021 too.

The benefits of the fellowship do not only include the experience from the finest institutes in India but also gives the candidates access to high-performance CPU/GPU infrastructure along with a rewarding stipend of ₹40,000 to ₹60,000 (per month), which will be based on the performance and the qualification of the candidate. With a bonus of a fully-funded trip to an AI and data science national conference throughout the candidate’s fellowship program.

Anyone who possesses the following credentials all across the country can apply for it. The certificates required are: bachelor’s degree and should not be above the age of 27, and demonstrate excellence in academics all across their undergraduate program. Good to have high-level proficiency in any coding language (preferably python/R/MATLAB) to obtain an added advantage.

The applicants should submit a copy of their online CV (Dropbox/OneDrive/Google Drive, etc.), a short research paper stating their purpose of the fellowship (300-500 words), a list of their relevant areas of research in the past or ongoing, and also a list of any relevant artificial intelligence or deep learning online courses (from NPTEL, Coursera, Udemy, Linkedin, etc.). To apply or obtain more information on the fellowship, candidates can visit https://rbcdsai.iitm.ac.in/Post_Bacc/.

The applicants need to apply before the 20th of every month, and the shortlisted candidates will be interviewed in person in the first week of the subsequent month.

GM-Backed Cruise Secures $5 Billion Credit Line For Self-Driving Car

By

Dipayan Mitra

-

June 17, 2021

General Motors’ majority-owned subsidiary, Cruise, has secured a new US$5 billion line of credit as it intends to commercialize its autonomous ride-hailing business. On Tuesday, this new credit was announced, which is being given by General Motors’ financing division for the purpose of purchasing Cruise’s self-driving origin shuttles.

GM plans to begin production of the same by 2023 in an unit located in Detroit. Dan Ammann, CEO of Cruise, claims that this brings Cruise’s valuation to over US$30 billion. He also said, “This is an incredibly exciting time for Cruise.” General Motors acquired Cruise in 2016 and has brought many investors like Microsoft, Walmart, Honda, and others.

But this time, instead of attempting to raise funds from outside, the company decided to provide Cruise the required credit internally. According to Cruise officials, GM has already begun assembling 100 pre-production units of Cruise Origin vehicles that would be built in Q2 of 2021 for testing purposes.

The Origin was first introduced in January 2020. The car is specifically designed as a driverless car. Hence, it lacks many traditional components like a steering wheel and pedals. The company has become the first autonomous car manufacturer to get a permit from the California Public Utilities Commission, which allows Cruise to give passenger rides in prototype driverless cars.

Commercializing autonomous vehicles has turned out to be way more difficult than it was expected a few years back. The challenges have led to a consolidation in the autonomous vehicle industry after many years of enthusiasm touting the technology as the upcoming multi-trillion dollar market.

Cruise earlier delayed their plan of launching their ride-hailing services in 2019 in San Francisco for testing purposes. The company has been providing ride-hailing services to its employees for many years in San Francisco.

Pony.ai To Launch Robotaxis In California By 2022

By

Dipayan Mitra

-

June 17, 2021

Pony.ai is a robotaxi startup that mainly operates in the United States and China. Recently, the company has started testing driverless cars in California, aiming to launch its commercial service by 2022.

Pony.ai said that humanless car testing is happening daily on roads in Milpitas and Fremont, California. Alongside the United States, the company is also testing its robotaxis in Guangzhou, China.

Pony.ai’s CEO and CO-founder, James Peng, said, “Going completely driverless is key to achieving full autonomy and an indispensable catalyst to realizing our ambitious vision.”

Though Pony.ai is currently facing some regulatory hurdles as it needs the approval of the California Public Utilities Commission and California Department of Motor Vehicles for operating commercially, the company firmly believes that they would be able to launch the service in 2022.

Last month, the California state issued Pony.ai a permit to test a sum of six driverless taxis in a specific location that spans about 39 square miles. After Baidu, AutoX, Zoox, and Neuro, Pony.ai has become the 8th company to receive a driverless vehicle testing permit in the United States.

Pony.ai also began operating PonyPilot, which is a public robotaxi service in Irvine, California. Yamibuy, an e-commerce company, partnered with Pony.ai to start a new autonomous service called ‘last-mile delivery’ in the same area. According to Pony.ai officials, the delivery service was started to ease the supply of overwhelming online orders during the COVID-19 pandemic.

Last year, the company stated its valuation to be US$5.3 billion after receiving funding of US$1 billion from giants like Toyota. The company has also partnered with many industry-leading organizations like Hyundai, Bosch, and Toyota.

Dataiku Takes Its Machine Learning Platform To The Cloud

By

Dipayan Mitra

-

June 16, 2021

Dataiku Takes Its Machine Learning Platform To The Cloud

Dataiku has introduced a cloud edition of its AI development platform that will reduce the work involved in developing custom machine learning models.

Dataiku is a New York-based company that has received $250million in funding from tech giants like Alphabet’s CapitalG venture. After secondary finance, it achieved unicorn status in 2019.

The company’s Artificial Intelligence (AI) platform helps companies develop machine learning models to perform many tasks like preparing training datasets.

Earlier, the customers needed to set up a dedicated infrastructure to run Dataiku’s software. The new managed edition of the software eliminates the need for any physical infrastructure as it is an entirely cloud-based service. This platform allows the customers to choose how they carry out tasks. They can develop a custom algorithm or can let the platform choose one using the AutoML feature.

The AutoML feature works by scrutinizing the kind of data a customer wants to process using its AI software and then finding out the best algorithm to perform the task efficiently. The software has a built-in AI library from which the desired algorithm is selected. Many organizations wish to build their custom AI algorithms. For this, Dataiku offers tools that data scientists use to develop a custom network or better known as hyperparameters.

Florian Douetteau, CEO of Dataiku, said, “We started developing Dataiku Online to address the needs of small and midsize businesses, in addition to startups, who don’t rely on on-premise technologies or custom clouds the way our enterprise customers do.”

Dataiku believes that this cloud-based service will help them expand further in the industry. The company’s platform is already being used by giants like Unilever PLC. and General Electronics Co. They are currently targeting small and midsized firms as now there is no need to set up a dedicated infrastructure.

Amazon India Launches ‘ML Summer School’ For Students

By

Dipayan Mitra

-

June 16, 2021

Amazon India Launches ML Summer School For Students

On Sunday, Amazon India launched an integrated learning program to learn applied Machine Learning (ML) skills for students to make them ready for a career in technology. Amazon India, in a statement, said that considering the high demand for talent with this skill set across various industries, they had introduced the program- ML Summer School, which would help students learn in-depth about Machine Learning.

It is a three-day immersive program that would cover the fundamentals of Machine Learning and its practical implementation in the industry. The participants of the ML Summer School program will be identified through an online assessment.

Final year engineering students along with Masters, Ph.D. students across select tech campuses in the country, which includes Indian Institute of Technology (IIT) Delhi, IIT Mumbai, IIT Kharagpur, International Institute of Information Technology (IIIT) Hyderabad, Birla Institute of Technology (BITS), National Institute of Technology (NIT) along with a few more Universities can take part in the ML Summer School.

Participants will get to learn about advanced Machine Learning techniques like the Probabilistic Graphical model and Deep learning along with their applications which can be used to solve specific business problems like search ranking and online advertising, catalog quality, and demand forecasting.

Students will also get access to the Amazon Research Days (ARD) conference, where they will get to know about trends in technology from the best industry experts in the world.

For those, who have prior knowledge of ML, this course will act as a refresher while additionally providing them a hands-on practical application of ML in the industry.

VP Indian Machine Learning at Amazon Rajeev Rastogi said, “With the pace of advancements in ML, we are proactively helping students to learn about the latest trends in the field of ML and apply them to solve real-world problems.”

Amazon India, in a blog, said that its machine learning team is driving innovations that would not only help Indians but customers worldwide.

AI Will Now Judge Artist’s Popularity In Italian Museums

By

Dipayan Mitra

-

June 16, 2021

AI Will Now Judge Artist’s Popularity In Italian Museums

New AI technology introduced in Italian museums will judge the popularity of artists through the facial expressions of the visitors. On Thursday, researchers started a trial of a camera system augmented by artificial intelligence in three Italian museums to gauge the visitors’ reactions to different works of art.

Italy’s National Agency for New Technologies, Energy and Sustainable Economic Development invented this technology, which has been on trial in Rome, Parma, and Bologna for the past three weeks.

The AI-powered camera system can read the visitor’s expressions – happy, sad, surprised, angry, neutral- along with their gender, age, and eye movements. Researchers believe this would help the curators plan their exhibitions better by putting on air well-liked pieces of art over those that might not get much attention.

However, this technology has received some criticism from artists as they believe this would lead art into another race for ‘likes.’ Inside the museum, a visitor was spotted smiling at a 12th Century painting, which was reported back to the database.

Silvia Battistini, Bologna civic museum curator, said, “I was surprised at what a great opportunity this was. As a curator, you don’t usually get such relevant information about the general public’s point of view without a mediator of some sort.” Though museums in London and the USA have previously used Artificial Intelligence, this ShareArt experiment is one of its kind.

The technology was initially built to avoid potential Covid contagion after the museums reopen. Ricardo Scipinotti, one of the developers of this system, said that the system had recorded many neutral expressions. He added that the visitors need not worry about privacy, and the data is not stored anywhere.

The Head of curatorial at Design museums in London said while time sensors can help museums understand the visitor’s liking, facial recognition is not a correct way to judge someone’s preference, and expressions are not an exact science.

Telangana Government Launches Mission Revv Up for AI Startups

By

Dipayan Mitra

-

June 16, 2021

Telangana Government Launches Mission Revv Up for AI Startups

Telangana Government on Monday announced the Telangana AI Mission as a part of the ‘2020 Year of AI’ initiative. The purpose of this AI Mission is to make Telangana a global hub for Artificial Intelligence.

‘Revv Up’ accelerator program was also announced by the government, which will empower AI-based start-ups. Revv UP will start in the month of July with the sole purpose of developing the state into a global destination for AI and various other technologies.

This Accelerator program will enable start-ups to work alongside the government to develop solutions for complex business problems. Selected start-ups will also receive help from the government and industry to set up their business at a larger scale.

Jayesh Ranjan, Principal Secretary, Government of Telangana, said at the launch, “Government of Telangana is committed to its vision of positioning Telangana as the leader in AI. After becoming the first state in India to launch an actionable policy framework for Artificial Intelligence in June 2020, we are now happy to set in motion the first edition of the Revv Up accelerator program to hone innovative AI ideas under the T-AIM. This accelerator will be a torchbearer for future Artificial Intelligence accelerators in the nation, and we are excited that NASSCOM will execute the envisioned strategy of T-AIM to identify and promote innovative and impactful AI solutions to solve real-world problems”.

This special program is meant for AI-based start-ups in Hyderabad or is willing to set up their office in the city. Telangana AI Mission is encouraging AI start-ups with a good use case to take part in this program, which would boost the growth rate of start-ups, resulting in a higher overall growth rate of the industry.

NASSCOM’s President, Debjani Ghosh, congratulated the entire team of T-AIM for the launch of this program for AI start-ups. She also said that NASSCOM would provide the best support possible to make this program a success.

Why is the world facing a chipset shortage in 2021?

By

Dipayan Mitra

-

June 16, 2021

Why is the world facing a chipset shortage in 2021

Whenever we think of chipsets, we tend to associate them with smartphones, but that is not the reality. Every electronic device we use has semiconductor chipsets in it. From the rise in smartphone prices, supply shortage of home appliances to delayed delivery of cars are because of this unexpected shortage of semiconductor microchips.

Recently Samsung Electronics has suspended the production of their flagship smartphone because of this shortage. The root cause of the shortage is an amalgamation of many factors as car manufacturers closed their plants last year amidst the coronavirus pandemic. They now are competing against the booming industry of consumer electronics for microchip supplies.

During the pandemic, the electronic industry experienced a huge demand as consumers stocked up mobile phones and laptops. The automobile industry also experienced a higher demand of cars than expected, which further tightened the supply.

Automobiles being the first affected industry, the shortage has now spread across many other electronic sectors like smartphones, air conditioners, microwaves etc. This shortage caused manufacturers to stock up their supplies, which has resulted in a price hike of even the cheapest components of chipsets leading to an increased price of final products. The demand for 5G phones increased rapidly and the manufacturers struggled to increase the production.

Tech giants like Qualcomm and Foxconn have warned their clients about microchip shortage, which has affected their supply line. Majority of the chipset manufacturing is done in Asia by companies like Samsung and Taiwan Semiconductor Manufacturing Co. Ltd.

Companies that produce chipsets are extremely expensive to build. Expanding their existing production capacity is equally hard as it includes induction of many complex hardware, which is very time consuming.. Experts say that it could take more than a year to successfully expand the chipset manufacturing industry. Joe Biden, the President of the US has announced $37 billion to fund the execution of expanding microchip manufacturing in the country. China is also offering subsidies to local companies to develop machineries in order to reduce its dependence on western technologies.

Hugging Face Launches Their Free NLP Course

By

Dipayan Mitra

-

June 15, 2021

Hugging Face Launches Their Free NLP Course

Hugging Face launches a free Natural Language Processing (NLP) Course on their website. The course will give in-depth knowledge and understanding about NLP using libraries from the Hugging Face ecosystem and Hugging Face Hub. The course is meticulously designed by Hugging Face’s industry-leading Machine Learning engineers like Mathew Carrigan and Sylvian Gugger.

The course is divided into three sections that progressively get more advanced. As of now, the first section of the course has been released. The second section is to be released in the mid of 2021, followed by the third section, which is expected to release in early 2022.

The first section will give knowledge about Transformer models and its usage. Transformer model is an architecture for transforming one sequence into another. Many giant organizations like Google AI and Facebook AI use the Hugging Face Transformer model and contribute to the community by sharing their models.

The first chapter of the course doesn’t require any prior technical knowledge. The chapter talks about what transformer models can do and how one can use them in practice. The following three chapters require previous knowledge of Python and some basics of Machine Learning and Deep Learning. The next set of 4 chapters will teach the very fundamentals of datasets and Tokenizers.

After basics, the course dives into classic NLP tasks. On completion of chapter eight, one will be able to deal with basic NLP problems independently. The next set of four chapters gives even more profound knowledge about architecture, including memory efficiency, long sequences, and lots more. By the end of the first section of the course, one can handle complex NLP problems and solve them.

In this world of technology, Machine Learning is growing at a meteoric pace. This course by Hugging Face will impart valuable and useful knowledge to someone willing to enter the Machine Learning industry.

Google’s New AI Can Design Chips Faster Than Humans

By

Dipayan Mitra

-

June 15, 2021

Google’s New AI Can Design Chips Faster Than Humans

Google recently announced that they had developed an artificial intelligence (AI) software that can design micro chipsets faster than humans. The company on Wednesday said to the journal Nature that a chip made by humans would take months to design, can be made in less than 6 hours by its new AI software.

Google said that this AI has already been used to develop Google’s tensor processing unit chips, which are used to run AI-based tasks. Google’s co-head of machine learning, Anna Goldie and Azalia Mirhoseini, said, “Our method has been used in production to design the next generation of Google TPU.”

In other words, Google is using its AI software to create even more complex AI systems. Precisely, this AI software of Google can draw the blueprint of a chip, which includes plotting where components like CPUs, GPUs, and memory are placed on the silicon die in relation to one another.

The positioning of their micro components is of huge importance as it affects the chipset’s processing speed and power consumption. Humans would take months to design these blueprints accurately, but Google’s deep reinforcement learning system can do the job with minimum effort.

This AI software is trained to determine which component combination works best to make the chipset most efficient. The software was fed more than 10,000 chip blueprints for it to learn what works best. Human chip designers generally layout components in neat lines. On the contrary, Google’s AI uses a more scattered method to design chipsets.

Yan LeCun, Facebook’s chief AI scientist, wrote about this research on Twitter, “very nice work, this is exactly the type of setting in which RL shines.” Nature editorial on Wednesday hailed this research as an “important achievement” that would be very helpful in speeding up the supply chain.