Quora, a social question-and-answering website, has launched another AI platform called Poe. Poe (Platform Open for Exploration) will enable users to indulge in proper back-and-forth conversations and receive answers instantly. The service is currently available only for iOS users.
Poe functions like a standard text messaging application for AI–users text and talk with AI chat interfaces, somewhat like ChatGPT. The fundamental difference is that Poe is not designed to cater to technical or knowledgeable question-answering, like poetry, coding, etc. It has been designed for flowy conversations and standard chit-chat so that the parent platform remains out of hot water in case the AI outputs controversial answers, like that in ChatGPT’s case.
To chat with Poe, users can use key phrases like “writing help,” “problem-solving,” “nature,” etc. The idea is to make ordinary people familiar with language models and how they work.
As per a spokesperson, Quora launched Poe after over 12 years of research and experience building consumer internet products. They added, “We believe much of what we’ve learned can be applied to this new domain where people are interfacing with large language models.”
The project is yet in its infancy, and Quora is focusing on its scalability and feedback from users who tested the beta version of Poe.
Google hosted its 8th annual Google For India 2022 event on 19th December in New Delhi. The event was attended by Google’s CEO, Sundar Pichai, and Ashwini Vaishnav, Union Minister of Railways, Communications and Electronics & Information Technology, Government of India. Both talked about Google’s progress in building AI-enabled and India’s first innovations at the event. From breaking down language barriers to assisting millions of people to access data more naturally to keep them safe at every step, these innovations highlight Google’s commitment to making a digital India into reality. This article covers the important announcements made at the Google For India 2022 event to empower AI technology in its products used in India.
Introducing more natural ways to search
Today, search has moved beyond typing a text, and people are searching for information on the internet with the help of voice assistants and cameras to learn more about the world around them. But in the case of exploring advanced technologies information, you need to interact with it to get a complete understanding, just like you do in the real world. Therefore, to help people explore advanced technologies, Google has announced to focus on building unique search experiences that encourage people to interact with technology and explore information in the Google for India 2022 event.
Search and explore visually
You can often not describe what you are searching for on the web. Google helps you to explore search with your camera or images using Google Lens. Indians are already using Google Lens every month more than any other country worldwide.
To make the visual search even more natural, Google introduced Multisearch. Multisearch is a milestone in visual search that helps people to search with images and texts simultaneously. You need to open the camera in the Google app to take a picture or use an image and then add words to your search. Multisearch is currently available in the English language in India. Google announced in the Google for India 2022 event that Multisearch will be available in Hindi and other Indian languages in the following year.
Besides image searches, Google also talked about video searches at the event. Videos are a rich source of visual information, but finding something within a video requires time. With the search-in video feature, Google will help you quickly find anything you want within a video by entering a term beneath it. In the Google for India 2022 event, Google announced the implementation of a search-in video service in India to search for specific information within videos.
More natural search experience in Indian languages
Searching for information in your native language sounds more natural than searching for it in any formal language. Therefore, Google has announced making search pages bilingual for people who use many languages to search and explore information in the Google For India 2022 event. With advanced machine learning-based translation models and cross-language search technology, Google will serve you high-quality and relevant content in their local language, along with English results they wish to view. This type of innovation will be launched in India for the first time. Google has already started working on the Hindi language and will expand to other Indian languages like Tamil, Telugu, Marathi, and Bengali in the coming year.
Speech recognition technology for Hinglish speakers
Google is addressing India’s dynamic language needs regarding voice. Voice-based technologies have shown a transformational impact in India, which helped people find information by speaking in their language. Google wants to help more people, especially Indians, ask questions in search engines naturally in their voice. Therefore, it announced a new speech recognition model, which can understand people speaking in Hinglish at the Google for India 2022 Event. Google will lead this innovation using a new neural network-inspired speech recognition model that considers the user’s accent, surrounding sound, contexts, and speaking styles.
Help people with non-standard speech
Mostly, when people speak, it is not always necessary that they will be heard and understood. More than 250 billion people globally do not have standard speech or find it hard to communicate with others. Therefore, to solve this problem, speech recognition innovations are made. With years of research, Google has invented the Project Relate app, which is trained to identify unique speech patterns in non-standard speech. Google has been using Project Relate with English language users in India. It announced expanding the app to Hindi users in early 2023 at Google for India 2022 Event.
Easy access to verified official documents from Digilocker on the user’s android phone
Many people store multiple images of essential documents like aadhaar cards, PAN cards, and driver’s licenses on their phones. They often use these images regularly to fill out online passport applications, sign up for citizen benefits schemes, and more. Google announced a collaboration with the National eGovernance Divison (NeGD) to allow people to access their authentic documents directly from the Files by Google app on Android at the Google for India 2022 Event.
This functionality of Google is built on privacy and security, where documents stored in Files by Google will be in an isolated environment on the user’s device, and they can access using a unique lock screen authentication.
Google implements a unique AI model to scan and identify essential documents on a user’s phone and then leverages OCR technology and machine learning models to identify relevant texts and gather this information into a quick and simple view. According to Google, all these models run entirely offline and cannot be accessed by anyone, including Google.
Leveraging deep learning to improve fraud detection in Google Pay
Since Google Pay has completed five years in India, enhancing the extensive impact of digital payments for millions of people who do online transactions is challenging. As Google continues to find ways to streamline paying online, safety continues to strengthen those ways.
Google uses advanced methods in its safety engine to analyze aggregated transaction patterns of millions of users related to a graph of a billion nodes. This method helps identify fraudulent behavior and then uses its intelligent language engine of Google to flash the warning and notify the user in their preferred language. Occasionally, Google generates a quick vibration to get the user’s attention.
Google announced Google pay’s new feature at the Google For India 2022 Event for accessing users’ transaction history in simple language. With the new feature, you can ask Google pay queries in a natural language like ‘Show me how much I spent on coffee last week.’
Collaboration with the Indian Institute of Science on ‘Project Vani’
Google is working on its vision to provide an exclusive internet experience to every user in their choice of language via text or voice. As a part of this vision, Google has collaborated with the Indian Institute of Science on ‘Project Vani’ to collect and compose open-source speech data from the 733 districts of India and made it available through the Government of India’s Bhashini project.
Collaboration with the Government of Telangana to deliver India’s vision of Digital Agriculture
Google supports partnerships to empower AI’s impact on agriculture and share the perks of AI technology across the country. Google is collaborating with the Government of Telangana as it explores AI-based solutions for agriculture innovation for six million farmers in the state. Google is also partnering with Cropin, a global agricultural ecosystem intelligence provider, to improve applications to track land use and bring digital services to farmers.
Collaboration with IIT Madras with a grant of $1 million to a multidisciplinary center for responsible AI
Since India’s digital ecosystem increasingly adopts and uses AI, Google is committed to sharing the best practices it developed in 2018 when Google began campaigning for responsible AI. To build a foundation of interpretability, fairness, privacy, and security, Google supports establishing the multidisciplinary center for Responsible AI through the $1 million fund granted to the Indian Institute of Technology, Madras. The center for responsible AI is a first of its kind in India that aims to bring together legal experts, sociologists, technologists, and policymakers from academia, government, and industry to study all the aspects of responsible AI.
The chairman of Ram investment, Sheikh Rashid Bin Ahmed AI Maktoum, has collaborated with Europe’s AI venture studio and ecosystem, Merantixto, to build an AI city in Dubai.
AI city will be a global hub powered by smart city infrastructure aimed to develop, implement and scale impactful AI solutions. It will connect AI research institutions, universities, and talent with leading startups, corporations, and investors. The project will work closely with universities to attract and encourage talent.
With the AI city, a regulatory sandbox environment can be created to host world-leading computing infrastructures and data centers to enable the testing of wide-scale revolutionary and safety-critical AI concepts. The AI city will host an innovation lab, research centers, and co-working communities to connect and collaborate. It will also be open to the public to raise awareness and show the impact of AI.
Sheikh Rashid Bin Ahmed AI Maktoum mentioned that the AI city would catalyze new products and solutions for global issues developed in Dubai. It will also be a great place to bring AI closer to society. With AI city, AI can be more accessible, and people will be more aware of AI innovations and projects.
The Ministry of Culture and Sports, Spain, has decided to provide 8 million euros for developing video games and metaverse experiences on Tuesday. The 8 million euro fund is a part of the Spanish Recovery, Transformation, and Resilience Plan, which provides funds to incubate projects in the metaverse and video games.
The funds will be delivered as a yearly or multi-year grant to companies seeking modernization and digitalization of all aspects of the Spanish economy. Besides Spain, other countries like South Korea have already provided funds for metaverse investments.
Miquel Iceta, Minister of Cultural and Sports, Spain, stated that the video game industry had been essential in transforming the productive Spanish model. The funding will promote Spain’s projects in small and medium-sized enterprises and self-employed workers.
As per the Ministry, video games and metaverse experiences will now be considered cultural creations in Spain. The National Library of Spain will preserve all the video game and metaverse experiences as Spanish heritage, which can be used for future generations.
Since video games are the most popular audiovisual entertainment option in Spain, the Ministry wants to put Spain in a place where it can compete with other nations in video games and export its games.
Mahindra has unveiled the metaverse platform for Mahindra’s all-electric XUV400 in the virtual world, XUV400verse, powered by Metadome.ai.
The platform will serve as a universe that combines the brand’s virtual spaces to provide a unique customer experience with photorealistic graphics. The XUV400verse will allow Mahindra enthusiasts and customers to socialize and have an immersive product interaction.
The virtual showroom features an LED wall showcasing spatial sound interactive avatars, a futuristic design with guided and free-roam journeys, and zones with brand-led merchandise and accessories.
Visitors in the XUV400verse can create their own avatars and also invite their family and friends for a shared experience. Users can also engage with a realistic 3D car configurator, which enables real-time customization, switching colors, and exploring the SUV’s features.
The XUV400verse also hosts a unique virtual test drive with several modes and camera views, where the user may learn about the salient features of the SUV. The environment also has brand billboards and cinematics.
A Conservative MP, Luke Evans, has delivered an AI-generated speech in the House of Commons. He used a chatbot called ChatGPT to generate his speech. Evans started his speech by telling the House that it may be the first time someone has done this.
The AI program generated the speech for the prompt, a Churchillian speech on the state of the UK over the last 12 months. When questioned by Labour’s Kevin Brennan, the MP for Cardiff West, Evans emphasized the importance of AI and algorithms in shaping tomorrow’s society.
AI and algorithms are set to significantly impact education, society, the labor market, diagnosis, and tools, and it needs a far broader debate than today, Evans said. He noted that this year had seen a seismic change in which AI can be used differently.
However, he proposed for regulating algorithms and AI to run concurrently, in synergy with the technology the UK is currently developing.
Later, Brennan joked that even his speech was written by an AI program. Evans further said that they all must face the fact that he may not be the first member to have used AI to deliver speeches.
He said someone might have already done it and not confessed it to the House of Commons. He then cautioned the House about the potential impacts of this technology.
Perfect, an Israel-based human resource technology startup, has raised $13M in a seed funding round. Target Global, Joule Venture, and RTP Global led the seed funding round. Individual investors like Chemi Peres, the Kalish brothers, Dan Naftali, Uri Levy, Young Sohn, Leon Recanati, Oren Rozenbech, Tomer Pomerantz, and Nir Adler also participated in the round.
Perfect is an AI-enabled platform that helps organizations to fill staff vacancies by mapping the profiles of millions of technology-focused professionals. The company mentions that it wants to break LinkedIn’s monopoly on the talent-searching market, so it relies on generative AI to help companies search for the ideal candidates they are looking for. Perfect scans hundreds of public data sources, like social media and other databases, to search for ideal candidates whose skills match the company’s requirements.
According to Perfect, it has mapped profiles and career graphs of millions of US-based technology professionals. Perfect constantly scans public sources for live signals about companies, such as layoff indications, senior role changes, and other news.
CEO of Perfect, Eylon Etshtein, stated that Perfect sees itself as the ‘Spotify for Talent’ and it works by learning the hiring preferences of each company to adapt recommendations accordingly. Perfect’s list of customers includes Incredibuild Ltd, RTP Global, and RocketVisor Corp.
Sonio, a French AI startup that helps fetal medicine professionals in monitoring and securing pregnancy-related ultrasounds, has received funding from the EIC Fund, a leading funding organization that regularly supports startups. The funding comes in addition to the previously received €5m from Elaia, Bpi France, and OneRagTime.
Sonio, established a few years ago, benefits from five years of R&D and clinical research at the Necker Hospital in Paris, the PRAIRIE Institute, and Ecole Polytechnique. As of now, the company already offers a comprehensive set of solutions to assist over 250 practitioners, from beginners to the most experienced ones.
Within the last 2 years, Sonio has brought two significant products into the fetal medicine market: Sonio Diagnostics and Sonio Pro. Sonio DIagnostics is an ultrasound solution for professionals, and Sonio Pro is a more secure and efficient pregnancy diagnostic solution launched recently in October.
The new funding will be used to accelerate the development of newer features on the platform, like embeddable genomics and image recognition. Cecile Brosset, CEO and Co-founder of Sonio, said, “Our hard work has been fruitful. I am now more confident than ever that we can help secure pregnancy care worldwide with collective and artificial intelligence.”
Sonio is also partnering with several companies, like Samsung (France), one of the leading women’s health ultrasound service providers, Lifecell (India), Manager Systems (Brazil), and Dynamic Medicals (Nigeria).
The Department of Navy (DoN), United States, awards Amazon Web Services (AWS) a five–year enterprise software license contract of $700 million. According to the contract, AWS can produce a wide range of cloud services for the US Navy.
As per the contract, AWS will provide the US Navy access to its professional services, commercial cloud environment, and AWS training and certification courses. The US Navy specified that the agreement does not commit funds at the time of the award. However, the funds will be committed once the task orders are issued from various sources, including working capital, maintenance funds, and operations.
The agreement features the Naval Information Warfare Center Pacific in San Diego, California, as the contracting entity for the deal, which states that the work will be carried out for a maximum of five years between December 2022 to December 2028.
AWS has own lots of success in landing costly cloud contracts from the US government. Recently, the US Department of Defence announced its Pentagon cloud computing contract with four leading companies: Microsoft, AWS, Oracle, and Google. According to the agreement, the Pentagon will give each company about $9 billion for developing a joint new cloud architecture named Joint Warfighting Cloud Capability.
Quris, an Israeli pharmaceutical AI company, announced that it had secured an additional $9M in seed funding on Tuesday. The seed funding round was led by SoftBank Vision fund 2, along with private and institutional investors like GlenRock Capital, iAngles, Richter Group, and Welltech Ventures. With the seed funding, Quris can advance its Bio-AI platform, grow its team, encourage industry collaboration, and speed up its drug research.
Yossi Cohen, Director of Israel Operations for SoftBank Investment Advisers, stated that SoftBank believes in the early forecasting capabilities that Quris is developing and showing its impact on the global pharmaceutical industry. By utilizing the power of AI, Dr. Bentwich, the founder and CEO of Quris technology and his team will save the enormous cost of drugs failing in clinical trials.
Quris is the first startup to develop an AI-enabled ‘patient-on-a-chip’ system, an alternative to human clinical trials and animal testing in drug development. As per Quris, its platform can revolutionize drug development by simulating clinical trials instead of testing new drugs on animals or humans.
Dr Bentwich mentioned that his company has much to gain from the SoftBank partnership. SoftBank’s ecosystem consists of a cluster of companies leading in the field of AI for pharma. As an investment partner, SoftBank has given companies a unique perspective on AI to produce significant unicorns.
