Language is not just a communication tool but an expression of different cultures, societies, and opinions across the globe. Nonetheless, language is also the barrier separating them. Thanks to translation technologies and artificial intelligence (AI) taking over the linguistic world, people can now read in their preferred languages. The world would lose a significant portion of its cultural treasures if it weren’t for translation and, more recently, technologies for translation.
Like other technological developments, translation technologies have evolved too. Currently, the most frequently used method of translation is via machines, Machine Translation (MT). Other methods like Computer-Assisted Translation (CAT) Technology are also prominently used. These technologies have undoubtedly offered seamless communication capabilities that people have wanted for ages. Likewise, they still have undeniable limitations.
All tools and technologies used for translation work on different principles and consequently deliver different results. Some offer more accurate results, while others are compatible with a more significant number of languages. Moreover, all high-end translation tools are not accessible to billions of people and are incompatible with hundreds of languages. People cannot openly participate in online conversations and communities in their regional/native languages.
To remove some of these barriers and make people a part of the future metaverse, AI researchers at Meta have created ‘No Language Left Behind-200’ or NLLB-200, an AI model to enhance machine translation capabilities for most of the world’s languages. The company claims that the model translates 200 languages with higher accuracy by an average of 44%. These languages include lesser-known African languages like Kamba and Lao (55 in total) and languages from other parts of the world. Such languages are incompatible with other existing translation tools.
No Language Left behind (NLLB) is a part of Meta’s long-term efforts to build language and machine translation tools. Launched in February 2022, the project builds advanced AI models to learn and decipher languages based on fewer examples.
The NLLB-200 is made to truly serve everyone, as other AI systems are not designed to cater to hundreds of local languages and provide a real-time speech-to-speech translation. Covering 200 languages is a step forward in overcoming data scarcity and acquiring more training data in local/regional languages. The new AI model also aims to overcome some modeling challenges of expansion faced by the company in previous years.
It is not the first time that Meta has developed a translation model. It released the 100-language M2M-100 translation model in 2020 with improved architectures and data acquiring practices. The AI company has now scaled to another 100 languages in NLLB-200. It can be used to advance other technologies, developing assistants for languages like Uzbek and creating subtitles for movies in Oromo/Swahili. There are endless possibilities to extend its application and democratize access for people in virtual worlds.
Meta trained NLLB-200 on FLORES-200, a dataset that enables AI’s performance assessment in 40,000 different language directions. The dataset measured NLLB-200’s performance in each of the 200 languages to be highly accurate.
Adding to the upsides, Meta is open-sourcing the model and the FLORES-200 dataset to all developers. It has also open-sourced the model training code. The company has also provided a demo to show the application of this open-source translator. The sole reason behind providing open-source access is to help researchers improve their work and translation capabilities via machines. Since inaccessibility is a major drawback of other language translation technologies/tools, Meta’s AI would make technology accessible to ordinary people.
Further, NLLB-200 will aid in promoting native languages and enabling people to read things without an intermediary language. Languages like Mandarin, English, and Spanish dominate the language webspace. Many people from other countries or regions cannot get the sentiments or context of things written in languages other than their own. NLLB-200 will bridge this gap and add meaning to the text, as people can now read in their preferred language.
As an incentive to use the AI model impactfully, Meta is awarding up to US$200,000 grants to researchers and nonprofit organizations. These researchers/organizations are invited to use NLLB-200 to translate underrepresented languages.
Meta has also collaborated with Wikimedia Foundation, a nonprofit organization, to offer translation services on Wikipedia. The model would help reduce the disparity between English publications on the website and those in other languages, especially those spoken outside of America and Europe. For instance, there are only 3,260 Wiki articles in Lingala, a native language spoken by 45M people in the Democratic Republic of Congo, against 2.5M Wiki articles in a language like Swedish, spoken in Sweden and Finland by much lesser people.
Even though the AI model has enhanced accuracy and meaningful translation of more languages than before, there is an endless scope for improvement. 200 languages cannot cover the entire language space. Additionally, the company faced several challenges in expanding the model from 100 to 200 languages. Since many of these languages are regional, the challenge is to acquire data from low-resource datasets. The model starts overfitting if trained for extended periods due to data scarcity. Such challenges would only scale as the number of languages increases. Long story short, there is a long road ahead for translation technologies, but NLLB-200 takes us one step forward in the right direction. Meta plans to strive for a more inclusive and connected world by breaking down linguistic and technological barriers and empowering people.
Meta, the American multinational technology conglomerate that owns Facebook, Instagram, and WhatsApp, has recently announced that it is suing the US subsidiary of a Chinese tech company Shenzhen Vision Information Technology Co., on the grounds of data-scraping from Facebook and Instagram.
Legal actions are being taken against Octopus Data, the US offshoot of a Chinese national high-tech enterprise Shenzhen Vision, which the parent website claims to have launched in 2016.
Meta also revealed that it is suing a Turkish individual identified as Ekrem Ateş, who allegedly set up automated accounts on Instagram to scrape data from about 350,000 Instagram profiles. The social media giant alleges that Ateş published the scraped data to their own websites, or so-called ‘clone sites’.
In its most general form, data scraping refers to a technique in which a computer program extracts data from the output generated from another program. Data scraping commonly manifests itself in the form of web scraping, which is the process of using an application or automated tool to extract valuable information en-masse from a website.
Social media giants and internet companies such as Meta are common victims of web-scraping. Despite a potential threat to sensitive private information, data scraping or, in this case, web scraping is, to one’s surprise, legalized in the US.
Almost three months earlier, a US court reaffirmed in a ruling that web-scraping is legal, which came as a verdict of a long-standing legal battle between LinkedIn, a Microsoft-owned platform, and Hiq Labs, a data science company. The latter had scraped private information from LinkedIn to assist its customers in predicting employee attrition. The court ruled that the action of scraping publicly accessible information does not infringe the Computer Fraud and Abuse Act (CFAA). CFAA is a cybersecurity law governing computer hacking in the US.
The decision by the US court came as a relief for professionals and amateurs across the industrial sectors, including journalists, researchers, and archivists whose jobs regularly involve scraping publicly available data. However, it sparked legitimate privacy and security concerns among users about how their publicly accessible data is harnessed without their permission.
It seems that in lieu of the decision for the ‘LinkedIn vs. Hiq Labs’ case, Meta is pursuing matters against Octopus Data through the Digital Millennium Copyright Act (DMCA) instead of targeting the entities under the Computer Fraud and Abuse Act (CFAA). The DMCA is more focused on intellectual property and copyright infringements than hacking.
In its accusations, Meta affirms that Octopus Data charges a fee to its customers to grant access to a software product named ‘Octoparse,’ which can launch scraping attacks. Besides, the customers can pay Octopus Data to scrape websites directly. For the software to work, the customers are required to give access to their accounts, which allows the software to scrape data that is usually only available to logged-in users. The personal data includes email addresses, phone numbers, birthdates, Facebook friends, Instagram followers, and more.
In a blog post, Jessica Romero, director of platform enforcement and litigation at Meta, wrote that their lawsuit alleges Octopus Data violated Meta’s Terms of Service and the Digital Millennium Copyright Act. She added that the accused did so by engaging in unauthorized automated scraping of data and by attempting to conceal their scraping to avoid being detected and blocked from Instagram and Facebook.
In its court filing, Meta specifically points toward certain parts of Section 3 of its Terms of Service, which state that users own the intellectual property rights such as copyrights and trademarks in any content they may create and share on any Meta company platforms. The terms of service also state that users have rights to their content and are free to share it with anyone or wherever they want.
Further down the section, it also mentions that the user cannot collect other users’ content or information. It also forbids users from accessing Facebook using automated means such as harvesting bots without Meta’s prior permission.
This lawsuit against the Chinese tech company Octopus Data comes in light of Meta’s recent victory in a similar data-scraping case, which was filed about two years ago against BrandTotal, an Israeli company. The latter offered a browser extension that scraped data from Facebook users. Unlike the ‘LinkedIn vs. Hiq Labs’ case, the court agreed with Meta’s claims that BrandTotal breached Facebook’s terms of use. The court also stated that BrandTotal violated the CFAA or Computer Data Access and Fraud Act (CDAFA) for California by hacking password-protected pages using automated user accounts.
Web-scraping, or in broad terms, data scraping, is a practice as old as the internet itself, and the possibility of getting rid of web-scrapers seems almost impossible, considering its prevalence. However, considering the potential threat it poses to the users’ private data, actions are being taken in favor of the victims. By targeting some of the top data scraping offenders at an individual and corporate level, Meta seems to be warning others from following suit. In light of Meta’s recent victory, it is likely that the outcome of the ‘Meta vs. Octopus Data’ case would be in the social media giant’s favor.
Apollo Hospitals has partnered with Singapore-based organization ConnectedLife to avail an artificial intelligence (AI) tool to predict the risk of cardiovascular diseases. This new advancement will allow doctors to intervene early in treatment by predicting the disease risk.
The Singaporean organization ConnectedLife provides digital solutions for condition management, wellness, and other health-focused applications. Under this tie-up, ConnectedLife will use the information from the Apollo Hospitals’ database to arrive at a risk score of cardiovascular diseases for its patients.
According to ConnectedLife’s founder Daryl Arnold, the data from the hospitals would be secure, and the company would adhere to the standards set by the Singapore government. He added that the data would help physicians develop a personal care plan to provide preventive care and digitally monitor the patients proactively.
The joint managing director of the Apollo Hospitals group, Sangita Reddy, said that the collaboration with ConnectedLife amalgamates artificial intelligence and machine learning technology with reliable and easy-to-use risk prediction tools that provide indications for early action. She added that the partnership would boost research to understand health risk scores and that Apollo would soon expand the collaboration to other non-communicable diseases.
ConnectedLife captures and analyses patient-reported data from wearable devices like Fitbit using AI to provide patient health and wellness insights. The data is then shared with the health care providers to help them to develop a care plan. The technology offers near real-time information, including details such as exercise, sedentary time, heart rate, breathing rate, and sleep, among others.
According to the director of Fitbit Health Solutions International, Steve Morley, the program provides patients with a better view of their health metrics to help them better manage their cardiovascular health.
C3 AI, an Enterprise AI software firm, announced its selection by Raytheon Technologies to deliver next-gen AI and enhance MLOps. The platform shall be used as a ready-now solution for the US Army’s Tactical Intelligence Targetting Access Node (TITAN) program.
Raytheon Technologies is competing with others in designing TITAN, a tactical ground solution that serves as the Army’s groundwork solution for multi-domain operations. TITAN tracks and locates threats to aid in precision targeting and advancing the Department of Defence’s strategy for the Joint Force C2 (JADC2).
To offer enhanced MLOps and AI capabilities, it will use C3 AI’s ML/AI model operations for the most effective third-party models across the TITAN enterprise. It will include both connected edge and cloud-based environments.
Thomas M. Siebel, Chairman and CEO of C3 AI, said, “This work combines Raytheon Intelligence & Space’s expertise in aerospace and defense with C3 AI’s proven expertise in enterprise AI to support critical national security interests through next-generation technology.”
TITAN will use high-altitude, aerial, space, and terrestrial sensors data to ingest targetable data while providing multi-source intelligence support for commanders. The leveraged AI capabilities will support pattern-of-life sensemaking and automated target recognition to help operators make sense of the available data and be able to prosecute a target.
The former CEO of the public policy think tank of the Government of India, National Institution for Transforming India, or NITI Aayog, Amitabh Kant stepped down on June 30, 2022. His tenure was extended three times before the government expressed interest in bringing a new shift in the NITI Aayog’s approach and a greater focus on social sector schemes and redistribution measures.
Kant was appointed CEO on February 17, 2016, and oversaw several high-profile projects, including the Aspirational Districts Program, National Monetization Pipeline, Aspirational District Program, Production Linked Incentive Scheme, Asset Monetization, and Transformative Mobility, among others. Kant served as the Department of Industrial Policy & Promotion Secretary prior to his transfer to the NITI Aayog. Additionally, he served as the chief executive for the Delhi–Mumbai Industrial Corridor project, and he was the one who came up with the “Incredible India” campaign.
He graduated from St. Stephen’s College with a degree in economics and Jawaharlal Nehru University with a master’s degree in international relations. Kant has also received the Sir Edmund Hillary Fellowship from the Prime Minister of New Zealand. In 1980, he enlisted in the Kerala cadre of the Indian Administrative Service. He has also authored three books: Branding India: An Incredible Story, The Path Ahead: Transformative Ideas for India, and Incredible India 2.0: Synergies for Growth and Governance.
The Modi administration extended his term for a third time in 2021, this time by one year, till June 2022, after the first extension till June 30, 2019, and the second time till June 2021.
Kant will be succeeded by Parameswaran Iyer, a former IAS officer from the 1980 batch of Uttar Pradesh Cadre. Iyer voluntarily left the Indian Civil Services in 2009. In the same year, Iyer was appointed as the World Bank’s manager of water resources. In 2016, he became the Secretary of the Ministry of Drinking Water and Sanitation. Iyer has also worked for the UN as a senior specialist in rural water sanitation. He became Secretary of the Ministry of Drinking Water and Sanitation in 2016 and oversaw the nationally launched Swachh Bharat Mission and the Jal Jeevan Mission, which aims to provide piped water supply to all households by 2024 through integrated grassroots water supply management.
Digital Payment Ecosystem
Considering Kant’s legacy and contributions as head of NITI Aayog, Iyer has pretty big shoes to fill in. For instance, Kant led a high-level panel to examine all potential digital payment methods across sectors after being sworn in to transition to a more cashless economy. According to NITI Aayog, this committee would find and implement user-friendly digital payment methods in all economic sectors as soon as possible.
In order to promote the quick adoption of digital payment systems and aid in the quick transition to the cashless, digital payments economy across all states and sectors, the committee would also regularly engage with central ministries, regulators, state governments, district administration, local bodies, trade, and industry associations, etc.
To make sure that about 80% of Indian transactions transfer to the digital-only platform, a framework for implementation was established and monitored with stringent deadlines.
A year later, at the NDTV-Mastercard Cashless Bano India, Kant stated that India’s digital payments infrastructure was at the time five years ahead of the United States. He was astounded by the fin-tech industry’s pace of innovation as well as the number of digital options available to Indians today, like UPI.
Due to the increasing use of smartphones due to network penetration, the fairly widespread availability of biometrics among Indians, and the likelihood that it was the only nation with a unified payments interface, this initiative alone transformed India from one of the largest informal economies in the world to a country supporting digital payments.
India’s Unified Payments Interface (UPI) had 330 banks participating as of June 2022, and it had recorded 5.86 billion monthly transactions totaling INR10,14,384 crore.
Fostering the Startup Culture
Under Kant’s leadership, NITI Aayog has also given the Indian startup ecosystem greater momentum, which has led to a notable expansion of the startup ecosystem in India over the past several years, particularly with the introduction of the government’s “Startup India” project. The number of these startups has also clearly shifted from Tier 1 cities to Tier 2 cities, indicating that innovation is increasingly taking a pan-Indian approach.
A growing young population that embraces fast-paced technology, an ambitious consumer market, the expansion of the investment-active middle-class and upper-middle-class segments, and India’s continued support for “Frugal Innovation” are some of the key drivers of the Indian startup ecosystem.
The government has also devised an action plan for startups, including establishing a new portal that allowed private organizations, particularly startups, to leverage public data from multiple ministries for innovation and the development of sector-specific solutions. Startups can employ artificial intelligence (AI) and data to address problems unique to India. Moreover, in order to help entrepreneurs obtain patents, NITI Aayog also assembled a team of attorneys from several patent offices. A startup hub has also been set up to help, support, and guide startups. NITI Aayog had also organized multiple funding events targeted at renewing funding for the Modi government’s Startup India scheme, thus enhancing the ease of doing business in India.
The country’s startup figure rose from double digits to tens of thousands as a result of Kant’s radical reforms. Even the development of unicorns at the height of the COVID-19 pandemic has been quite impressive. Speaking at a FICCI Ladies Organization (FLO) event in March, Kant added that there are currently 81 unicorns and more than 61,000 startups in India. In May, after receiving US$50 million from IIFL, the Bengaluru-based neo-banking startup Open became the nation’s 100th unicorn.
Kant also took a keen interest in ensuring women-based startups grew in prominence in recent years, to promote women’s empowerment as well as act as a catalyst for socio-economic transformations. He asserted that women-owned firms and enterprises are rapidly becoming the next major disruption in the Indian startup ecosystem and are already playing a significant role in society.
Kant was also a proponent of the notion that innovation stems from the nurtured minds of young students. Therefore, as part of the Atal Innovation Mission (AIM), the government, in collaboration with NITI Aayog built 500 tinkering labs in schools in 2016 and a considerable number of incubators at the college level in an effort to encourage the spirit of innovation in the youngest citizens. Another aspect of the project was making sure the IITs (Indian Institutes of Technology) and IIMs (Indian Institutes of Management) have research labs.
In the following year, an additional 1,500 schools were chosen by NITI Aayog to implement the Atal Tinkering Labs (ATLs) program. Last December, with the goal of empowering innovators and entrepreneurs across the country, NITI Aayog launched a first-of-its-kind Vernacular Innovation Program (VIP), which will provide innovators and entrepreneurs in India with access to the innovation ecosystem in 22 scheduled languages. AIM will train a Vernacular Task Force (VTF) in each of the 22 scheduled languages to develop the required capability for the VIP. Each task force is led by a regional Atal Incubation Center and includes vernacular language instructors, subject matter experts, technical writers, and subject matter experts (AICs). With the help of VIP, Kant plans to minimize the language barrier in the fields of innovation and entrepreneurship.
Roadmap to AI Dominance
Under Kant’s supervision, NITI Aayog published National Strategy for Artificial Intelligence (NSAI) as early as June 2018, making India one of the first nations to consider harnessing the emerging technology for social good by addressing inclusion and societal issues. In the document, it was noted that the vast geographic and cultural diversity presents specific developmental problems in the areas of agriculture, smart mobility, and healthcare, all of which might benefit from AI. Therefore, the Indian government anticipated that if the AI solutions were applied to the diverse demographics, they would be applicable elsewhere as well.
The NITI Aayog also mentioned the cloud platform AIRAWAT, which stands for AI Research, Analytics and knoWledge Assimilation plaTform, in their AI Strategy Report.
With a sizable, power-optimized AI computing infrastructure and cutting-edge AI processing, the AIRAWAT is deemed as a cloud platform for big data analytics and assimilation. The Indian government intends to address the issues brought on by limited access to computing resources through AIRAWAT. The government will soon start developing compute infrastructure that is specifically designed for AI to support the computing requirements of Innovation Hubs, International Centers for Transformational AI, and Centres of Research Excellence (COREs).
According to a recent study by Microsoft and the Internet and Mobile Association of India (IAMAI), the artificial intelligence (AI) market in India is anticipated to grow by 20% over the next five years. The nation is also one of the top three talent markets, producing 16% of the global AI talent pool. These milestones can be thought of as a result of the domino effect that started with the announcement of strategic plans to boost the AI-backed domestic industries.
The incubative environment for startups and AI technologies built by NITI Aayog reforms also helped India strengthen its healthcare industry. With Covid-19 triggering the need for pharmaceutical research, on-spot testing, mental health bots, and more, the healthcare sector was pushed to its limits to cater to the new demands and urgencies. This resulted in numerous success stories for this industry. For instance, a Bengaluru-based Software-as-a-service (SaaS) platform for doctors recently released the mobile version of its AI-powered electronic medical records (EMR) app called “EMR on Mobile.” The app is developed using the same architecture as the company’s premier desktop EMR powered by AI. It can be used to gain access to patients’ information in real-time anywhere.
Doctors in more than 350 cities, including Tier II and Tier III cities, have used the mobile EMR. Further, they can use “EMR on Mobile” to manage both in-person and online consultations across the same pool of patients in the wake of the Covid-19 outbreak.
Boosting the EV Market
Under Kant’s leadership, Aayog has been at the forefront of the government’s effort to encourage electric vehicles (EVs) in order to reduce pollution. Aayog was also a major force behind the FAME (Faster Adoption and Manufacturing of Electric Vehicles in India) program, which announced a number of incentives for the EV industry.
One of the major achievements of FAME was the launching of the e-Sawaari India Electric Bus Coalition in December 2021. This was possible due to NITI Aayog’s collaboration with Convergence Energy Service Limited (CESL), World Resources Institute, India (WRI India), and funding from the Transformative Urban Mobility Initiative (TUMI).
The central, state, and city-level government organizations transit service providers, original equipment manufacturers (OEMs), financing institutions, and ancillary service providers will be able to share knowledge and their lessons learned on e-bus adoption in India through the e-Sawaari India Electric Bus Coalition. Thus embarking on a new chapter in the electric vehicles market in India along with adherence to India’s decarbonization strategy.
In December last year, Kant stated that the government was attempting to lower the 18% GST on EV batteries at the time. In the end, the GST Council decided that EVs, battery packs and all, will henceforth be taxed at 5%. On June 28 and 29, the Union Finance Minister Sitharaman presided over the 47th meeting of the GST Council in Chandigarh, where the decision was announced.
Kant had previously claimed in February that the government was interested in Tesla manufacturing its cars in India. He explained that there are two different rates of duties in India: one is around 110% for luxury cars, and the other is about 60% for cars made in India. While Tesla is welcome to 110% duty, it will be beneficial for the company if it sets up a manufacturing and assembly plant in India.
Encouraging Data Interoperability
The NITI Aayog introduced the National Data and Analytics Platform (NDAP) for free public usage in the first half of this year. Kant claims that it will house fundamental datasets from multiple governmental organizations and offer analytics and visualization capabilities. Last August, a beta version was made available to a select group of people for testing and feedback.
To ensure that the datasets stored on the platform are curated to the needs of data consumers from different sectors like government, academia, journalism, civil society, and the corporate sector—NDAP will adopt a use-case-based approach. Because all datasets adhere to a uniform schema, NDAP makes integrating datasets and conducting cross-sectoral analysis simple.
Miscellaneous
On September 1 last year, NITI Aayog officially unveiled Ernakulam Karayogam’s “Clinic on Wheels,” a mobile medical unit. Bharat Petroleum Corporation Ltd has funded this project as part of its CSR initiatives. The Mobile Medical Facility unit, which is outfitted with cutting-edge medical facilities, would travel to the isolated and coastal villages in the districts of Ernakulam, Alleppey, and Idukki in diagnosing and treating the poor and rural residents right where they live.
Kant has also endorsed cryptocurrencies, stating that they are simply another asset class like bonds, gold, and mutual funds. He argued that because it is a new asset class and a large number of people are using it for transactions, the government would lose money if it did not tax it. He lauded the government’s move to tax cryptocurrency revenues in this year’s budget session.
The Estonia-based company Retouch4me launched a new neural network-trained plugin Heal OFX to retouch skin imperfections in a video. The video plugin is based on the technology found in standard photography retouching software. It is a suite of nine plugins that target different retouching tasks simultaneously and automate the process done manually before.
This plugin is different from other editing plugins in that it does not make the skin look fake or plastic-like. The company’s AI-powered tools provide excellent results across all photography and videography principles.
The company also provided a video demonstration of retouching skin in a video.
The tool removes skin imperfections while preserving textures and tones, which helps to produce a final image that looks realistic. Retouch4Me is pricey, but it accomplishes what it promises. PetaPixel, a photography news aggregator, recommended this plugin as the best for AI Portrait Retouching.
Although the technology for still photos was remarkable, it is evident that the company had bigger plans for its AI. Heal OFX for DaVinci Resolve transfers the fundamental effects from its picture tools to video, where retouching has traditionally been more difficult.
Oleg Sharonov, the leading developer of the project, said, “We decided to simplify the task, making the editing convenient and optimizing the work of the neural network for video. We are launching a new plugin that retouches directly in Da Vinci by pressing just one button.”
Videographers can quickly achieve natural-looking effects using Retouch4me’s Heal OFX plugin without investing a lot of time or money in the process.
A new study by a team of researchers at UK-based AI company DeepMind suggests that AI can make better public money decisions than humans. For this purpose, DeepMind has developed more capable problem-solving systems known as artificial general intelligence.
According to the research, AI can devise methods of wealth distribution that are more popular than systems designed by people. It also shows that machine learning systems are good at delivering more open-ended social objectives, such as the goal of realizing a prosperous and fair society.
The team trained an AI system to find a popular public fund distribution policy in a four player online game.The AI learned from more than 4000 people and computer simulations . Players voted on their favorite policies for distributing public money. The AI policy won more votes from human players.
Creating a machine that delivers beneficial results humans actually want is defined as value alignment. One problem with value alignment is that human society admits a plurality of views which makes it unclear to whose preferences artificial intelligence should align. For this, the AI discovered a mechanism that redresses the initial wealth imbalance and sanctions-free riders, thereby successfully winning the majority vote.
This new approach by DeepMind researchers combines artificial intelligence with human democratic deliberation to come up with better solutions to social dilemmas, such as public money distribution.
Rebellions, a South Korean AI chipmaker, raised US$22.8M as an expansion to Series A funding led by KT, a telecommunications giant in the region. The funding round also included participation from Temasek’s Pavilion Capital, Korea Development Bank, SV Investment, Future Asset Capital, and Cacao ventures.
As per a Rebellions spokesperson, the expansion would enable mass-production of the second AI chip prototype ATOM to be used by cloud platforms and data centers.
KT aims to develop AI-powered chips like Neural Processing Units (NPUs) for data centers, self-driving vehicles, fintech services, etc. With the same vision, it is KT’s second investment in an AI chipmaker firm to accelerate its semiconductor business. It aims to expand and innovate to stand out like other semiconductor companies like NVIDIA and Qualcomm.
KT Hyeon-MoKu CEO said, “AI semiconductors are one of the next big technologies. We hope that through our partnership with KT, Rebellions will become a global fabless company like NVIDIA and Qualcomm.”
Rebellions founder and CEO Sunghyun Park said, “We look forward to collaborating with KT, a leader in the cloud and Internet data center industry, and the strategic partnership will be the driving force behind Rebel’s new growth and business.”
The company plans to continue its strategic investment in startups rising from stealth in a challenging investment environment.
HUAWEI recently launched the 2022 HUAWEI GLOBAL AI CHALLENGE for students enrolled in higher education and is now open for registration. Jointly held by Huawei Consumer Cloud Service Department, the Jiangsu Association of Artificial Intelligence (JSAI), and Huawei Nanjing Research Center, the challenge aims to invite young developers with a passion for AI and intelligence.
In its interim of three years after commencement in 2019, the challenge has attracted more than 2,500 algorithm proposals from nearly 8,000 teams across 45 countries! Huawei specifically focused on the enrollment of several top universities and colleges from China, given its origin.
Wang Yue, President of Huawei Consumer Cloud Service Application Ecosystem BU, said that the company aims to inspire global campus talent in an era of “ubiquitous intelligence” to explore new AI technologies with a problem-solving approach.
Potential contest proposals were assessed by experts on the following criteria: practical value, risk assessment, technical innovation, popularity, and difficulty. They finally decided on three proposals namely, “knowledge-driven spoken dialogue”, “intelligent quality inspection of lane rendering data”, and “CTR prediction through cross-domain data from ads and news feed”.
The contest will include the preliminary and elite final stages. The preliminary stage will end on August 24. The experts will take another week for preliminary review and then announce the selected teams by September 3. Ultimately, the elite stage commences in mid-September. Seven teams shall be selected via an online competition in the preliminary stage, each of whose proposals would enter the elite final. A total of 21 teams for the three proposals will compete for the prize money of US$210,000 via an online competition and presentation.
The winner will receive US$35,000. The first and second runners-up will each receive US$15,000 and US$10,000, respectively. In addition, four teams will receive honorary mentions and US$2,500 each.
All student developers who are willing to take on new challenges are cordially invited by Huawei to present their best work to a larger audience.
A team of researchers at the Indian Institute of Science (IISc) has created a design framework to build next-generation analog computing chipsets. These chipsets are expected to be faster and require less power than the digital chips used in most electronic devices.
The team has developed a prototype of an analog chipset, ARYABHAT-1 (Analog Reconfigurable technologY And Bias-scalable Hardware for AI Tasks), using a new design framework.
Such chipsets can be very helpful for artificial intelligence-based applications like object and speech recognition devices or those that require massive parallel computing operations at high speeds. The chipset can provide orders of magnitude of improvement in power and size. In applications not requiring precise calculations, analog computing has the potential to outperform digital computing, as the former is more energy-efficient.
However, there are technological hurdles. Unlike digital chips, the testing and co-design of analog processors are comparatively complex. They must be customized individually when transitioning to a new application or the next generation of technology. Also, their designs are expensive.
To overcome such challenges, the team has designed this new framework that enables the development of analog processors that scale like digital processors. Their chipset can be programmed and reconfigured to port the same analog modules across different applications and generations of process design.
Also, different machine learning architectures can be programmed on ARYABHAT. Similar to digital processors, it can operate robustly across a wide range of temperatures. According to researchers, the architecture is bias-scalable, i.e., its performance remains the same even when the operating conditions like current and voltage are changed.
