Top AI News of the Day (10 January 2023)

By

-

January 10, 2023

DoNotPay’s AI-Powered Robot Lawyer to Defend a Human

An AI-powered legal assistant lawyer is set to be an advisor to a defendant in the US next month. Claimed to be the world’s first robot lawyer, developed by DoNotPay, will run on a smartphone, attend the hearing and then advise the defendant via an earpiece. DoNotPay, a legal service provider founded by Joshua Browder, uses artificial intelligence to leverage legal consultancy services. To see how DoNotPay works, check out Joshua’s recent conversation with Wells Fargo bank.

Microsoft’s VALL-E, a New Text-to-Speech AI Model

Microsoft has unveiled VALL-E, a new text-to-speech AI model that, given a three-second audio sample, can accurately mimic a person’s voice. After the model becomes familiar with a distinct voice, it can synthesize audio of that voice saying anything while maintaining the speaker’s emotional tone. Read more about VALL-E here.

100% Wireless Displace OLED TV Showcased at CES 2023

A technology upstart called Displace has surpassed LG’s signature wireless OLED (with just a power line) by making its 55-inch 4K OLED TV completely wireless. The TV runs on four hot-swappable batteries and a base hub that transmits entertainment channels. It feels like a dream come true, but at a whopping US$3000!

Microsoft in Talks with OpenAI Over US$10B Investment

According to Semafor, Microsoft is in talks to invest roughly $10 billion in OpenAI, raising the company’s estimated value to $29 billion. The funding includes other venture firms too. All prospective investors recently received the deal’s paperwork, with the goal of closing the round by the end of 2022.

Maruti Suzuki’s ExpoVerse at the Auto Expo 2023

The automaker announced its metaverse project at Auto Expo 2023, ExpoVerse, aimed to offer an immersive user experience. Maruti Suzuki also showcased its virtual pavilion with various experimental zones and over 1,100 NEXA and ARENA dealerships, accessible with VR devices throughout the country.

AR Rahman Steps into the Web3 World with His Metaverse Platform ‘Katraar’

The world-famous Music composer, producer, and singer AR Rahman is the latest Indian celebrity to foray into the web3 ecosystem with his metaverse platform ‘Katraar.’ Katraar is a digital platform that will feature upcoming musicians and artists. Rahman posted a teaser for his 56th birthday on 6th January, surprising millions of his fans while announcing the platform as a tradition and means of wisdom for the music industry.

Conversational AI Startup NLX Raises US$4.6M for Product Development

NLX, an AI startup working on conversational technologies, said it had raised US$4.6m after over a year of previous investment in a funding round led by IAG Capital Partners in collaboration with JetBlue Ventures and Flying Fish Ventures. The funding will help the budding startup to advance its next-gen conversational experiences on its end-to-end no-code platform.

Microsoft acquires Fungible, a DPU-based Solution Provider

By

Sahil Pawar

-

January 10, 2023

In a blog published on Monday, Microsoft announced the acquisition of Fungible, a maker of composable disaggregated infrastructure solutions for data centers.

Santa Clara, California-based Fungible’s solutions are centered on its Data Processing Units (DPUs), which are considered as “third-socket” components for data centers after graphical processing units and central processing units (CPUs).

The aim behind the Linux-based system-on-chip DPUs is to improve the inefficiencies of data-centric computations that exist with hyper-converged infrastructures.

According to Microsoft’s announcement, Fungible’s technologies will be used to strengthen Microsoft’s high-performance, scalable, disaggregated, scaled-out data center infrastructure with reliability and security. Fungible’s staff will join Microsoft’s data center infrastructure engineering teams as part of the deal.

Fungible, established in 2015, announced that it was excited to be a part of Microsoft. The announcement mentioned that Fungible first invented its DPU in the year 2016 to address the main problem of scale-out data centers, namely the inefficient execution of data-centric computations within server nodes.

Fungible’s solutions have been used by big companies like VMware and IBM. The terms of the deal weren’t described by Microsoft or Fungible.

Microsoft announces new text-to-speech AI model called VALL-E

By

Sahil Pawar

-

January 10, 2023

Microsoft new text-to-speech AI model VALL-E

Microsoft announced a new text-to-speech AI model called VALL-E on Thursday that can simulate a person’s voice closely when given a three-second audio sample.

Once the model learns a specific voice, it can synthesize audio of that same person saying anything and preserves the speaker’s emotional tone. Its creators claim that VALL-E could be used for high-quality text-to-speech applications and audio content creation when brought together with other generative AI models like GPT-3.

Microsoft describes VALL-E as a neural codec language model which builds on a technology called EnCodec. Unlike other text-to-speech methods that usually synthesize speech by manipulating waveforms, VALL-E creates discrete audio codec codes from text and acoustic prompts using EnCodec.

It basically processes how a person sounds, breaks that information into discrete components called tokens, and uses training data to know how that voice would sound in other phrases outside of the three-second sample.

Microsoft trained VALL-E’s speech synthesis system on an audio library called LibriLight, which Meta assembled. It contains about 60,000 hours of English language speeches from over 7,000 speakers, mainly from LibriVox public domain audiobooks. For VALL-E to create a good result, the voice in the sample must closely match a voice in the training data.

Australian universities to take pen and paper exams after students using AI to cheat

By

Sahil Pawar

-

January 10, 2023

Australian universities have changed how they conduct exams and other assessments amid concerns over students using emerging artificial intelligence softwares like ChatGPT to write essays.

Major institutions in Australia have added new rules which state that the use of AI for exams is cheating, with some students already caught using the software. In Australia, academics have expressed concerns over ChatGPT and similar technology’s ability to escape anti-plagiarism software while providing quick academic writing.

The Group of Eight leading universities, which are leading research-intensive universities across the country, said that they had revised how the assessments would be run this year due to the emergent technology.

Dr. Matthew Brown, the group’s deputy chief executive, said its institutions were proactively tackling AI through targeted technological detection strategies, staff training, and redesigning assessments.

He said, “Our universities have revised how they will run assessments in 2023. They will include greater use of pen and paper exams and tests and tests only for units with low integrity risks.”

The University of Sydney’s latest academic integrity policy now specifically mentions generating content using artificial intelligence as a form of cheating. The Australian National University has changed assessment designs to rely on laboratory activities and fieldwork. It will also time exams and introduce more oral presentations.

Minnesota startup claims to have made world’s first AI-generated online course

By

Sahil Pawar

-

January 10, 2023

Minnesota startup world's first AI-generated online course

A Minnesota startup, JazzJune founder and CEO Alex Londo claims he has made the world’s first online course that was entirely created by using artificial intelligence chatbot ChatGPT.

According to Lando, with various single-sentence prompts from him, the AI chatbot wrote an outline and all the lessons for the course. It even found images and detailed videos concerning the subject. The AI generated an entire course consisting of 19 lessons and a quiz.

The course is called Introduction to Artificial Intelligence. “Ten hours later, I had an entire ‘introduction to artificial intelligence’ course. I had artificial intelligence write a course about artificial intelligence,” Londo said.

Londo wanted to see if AI could be beneficial to his startup, which is a platform for users to create and share learning content. Londo believes AIs like ChatGPT can do more good than harm in the classroom.

“If we really learn how to embrace it, teachers learn how to use it with their students, learn how to use it for themselves in their lesson planning and course development, you can go from spending five hours on that curriculum to two,” he said.

Apple Set for a Huge Retail Push in India, Tata Group Planning to Reap the Windfall in Manufacturing and Supply Chain

By

Disha Chopra

-

January 10, 2023

Apple is anticipated to open two store outlets in India this year, first in Mumbai and later in New Delhi, upping its game to push retail in the country. The first flagship store will be launched in Mumbai’s Bandra Kurla Complex (BKC) district, followed by another in Saket, New Delhi, after a few weeks.

Apple’s plans to further the brand in India while maintaining its ownership got delayed due to the COVID-19 pandemic. But now, the company is all set to launch more flagship products and stores where people can simply walk in to avail of Apple’s products.

Besides retail, the company also plans to expand manufacturing in the country. It believes that an Indian manufacturer of its iPhones would boost the brand’s sales in India and have a strong competitive advantage over China.

China’s manufacturing advantage and political tensions with the US have jeopardized Apple’s supply chain by rolling out occasional COVID lockdowns and policy changes.

It was earlier reported in September 2022 that the Tata group was in talks with the Taiwanese supplier of Apple products to launch an electronics manufacturing venture in India. Later on, in November, Tata planned to buy its manufacturing facility in Karnataka for up to ₹5,000 crores.

24 Chinese researchers claimed to have broken 2048-bit RSA encryption using a quantum computer

By

Sahil Pawar

-

January 10, 2023

Chinese claimed broken 2048-bit RSA encryption quantum computer

According to a paper titled “Factoring integers with sublinear resources on a superconducting quantum processor, 24 Chinese researchers have reportedly succeeded in breaking 2048-bit RSA encryption using a quantum computer. This can prove to be a significant achievement as RSA encryption is extensively used in current security practices.

However, several experts, cryptographers, and computer scientists do not believe the researchers have made a significant discovery solely based on the scientific paper published. In Sept. 2022, the US Cybersecurity and Infrastructure Security Agency (CISA) warned that a post-quantum world was coming soon and stressed that contemporary encryption techniques could break.

Months later, i.e. recently, in December 2022, the scientific paper published by Chinese researchers claiming to have broken RSA encryption using a quantum computer came into the limelight. The paper follows a report from April 2022 that claimed that China was “taking the lead” in the quantum computing race.

Financial Times questioned a few experts about the paper, and Roger Grimes, a computer security expert and author, said that it is a huge claim and that it would mean that governments could crack other governments’ secrets.

While sharing his opinion on the latter, Bruce Schneier, a computer security and cryptography expert, said that we have no empirical proof that the new quantum algorithm overcomes the Schnorr scaling problem. He added that there’s no reason to believe it won’t, but there’s no reason to believe it will.

Mastercard Teams Up With Polygon To Empower Emerging Artists via Web3 Incubator

By

Preetipadma K

-

January 9, 2023

Mastercard polygon Artist web3 incubator — Source: Mastercard

Mastercard, a leading global payments and technology corporation, teamed up with Polygon to introduce a web3-focused incubator at the Consumer Electronic Show (CES) 2023, with the goal of supporting artists in establishing new channels of communication with their audiences. According to a statement from Polygon Studios CEO Ryan Wyatt, the new Mastercard Artist Accelerator project will equip upcoming artists with the web3 tools and skills to enable them to interact with their audience. As a result, artists can maximize their earnings and advertise innovative mainstream products.

The Mastercard artist accelerator initiative will guide the five selected artists on a variety of tasks that will enable them to cut out the middleman, mint their own NFT (non-fungible tokens) collections to increase online fan engagement, and take part in metaverse-based concerts and more, using the Polygon network and the Polygon platform. Mastercard’s chief marketing and communications officer, Raja Rajamannar, emphasized that the company will also offer incubees marketing and technical consultancy assistance.

This announcement indicates that the company is moving further into the world of digital assets. Ryan thinks Web3 has the ability to empower a new breed of artists who can build a following, support themselves, and develop cutting-edge platforms for self-expression and human connection. The Polygon platform was selected as the new program’s technical partner due to its excellent scaling capabilities and prior experience working with top Web3 companies.

This program’s five participating artists have not yet been chosen, and artists are still welcome to submit their applications. The program will begin in the spring of 2023, and fans will be able to accompany these artists as they travel while working together to understand how these technologies work. The accelerator will contain a special edition NFT from Mastercard called “The Mastercard Music Pass,” which the company claims would grant owners access to unique resources, physical and digital experiences, and exclusive music educational materials. Additionally, the company said that the accelerator program would end with a live-streamed artist presentation later this year.

TLC launches RayNeo X2 augmented reality smart glasses at CES 2023

By

Sahil Pawar

-

January 9, 2023

TLC RayNeo X2 augmented reality smart glasses

A leading display technology and smart experiences provider, TCL, has announced the launch of TCL RayNeo X2 augmented reality (AR) smart glasses at CES 2023. The new glasses possess binocular full-color Micro-LED optical waveguide displays along with an array of interactive features to create AR experiences for the users.

The TCL Ray Neo X2, powered by the Qualcomm Snapdragon XR2 platform, is designed to address the industry challenge of making augmented reality glasses look like conventional eyeglasses without compromising on the tech specs. The glasses maintain a lightweight and subtle body for everyday wear while having powerful all-in-one assistant features, from auto-translation and smart navigation to photography and music play, according to the company.

The glasses have display upgrades, including an image brightness of up to 1,000 nits and a high contrast ratio (CR) of up to 100,000:1. These features generate enhanced visual effects that make the AR glasses ideal for outdoor and indoor usage alike.

The device also has a smart GPS navigation system that makes use of simultaneous localization and mapping (SLAM) along with gesture recognition. Its mapping shows nearby landmarks as the wearer moves. Also, the Bluetooth-enabled on-screen messages and call notifications keep users updated wherever they go.

Moreover, TCL RayNeo X2 has an AI translation feature that interprets multiple languages in real-time. When speaking face-to-face, the augmented reality glasses automatically detect and translate in-person conversations, with the subtitles displayed on the screen. This can also be turned into an immersive language-learning experience for the wearer.

MIT Researchers Devised a New Quantum Computing Architecture for Large-Scale Devices

By

Disha Chopra

-

January 9, 2023

MIT Researchers have developed a new architecture for quantum computing that allows high-fidelity communication across superconducting quantum computers. While quantum computers can conveniently emulate systems, optimize more complex tasks and simulate quantum chemistry, building a large-scale quantum computer is challenging.

Traditional quantum computing architectures comprise various components that perform different functions. Electronic information is passed among these components, encoded and stored as bits, via interconnects (wires that facilitate electron movements in a quantum computer).

To make quantum computing more feasible, MIT researchers have published their work in Nature Physics, demonstrating the beginning step where single photons (information carriers for quantum computers) are deterministically emitted towards a pre-specified direction. The newly devised quantum computing architecture ensures that information flows are correctly directed 96 percent of the time.

Bharath Kannan, a co-author of the research paper, said, “This ‘directional transmission’ is what we demonstrated, and it is the first step toward bidirectional communication with much higher fidelities.” He added that the new quantum computing architecture might be a more straightforward method of scaling the same to larger systems rather than approaching with a single oversized chip.

This quantum computing architecture can be both a transmitter and a receiver, with multiple modules and just one physical connection. The researchers will now intend to connect many modules and use the procedure to emit and absorb photons now that they have successfully emitted photons in a particular direction. It would be a significant development for modular architectures combining small-scale processors for large-scale applications.