People Are More Likely to Believe AI-Generated Tweets

By

-

June 30, 2023

People More Likely Believe AI-Generated Tweets — Image Source: Phys.org

In a groundbreaking study published in Science Advances, researchers have discovered that tweets written by artificial intelligence (AI) language models are more convincing to people than tweets created by humans. The study raises concerns about the potential impact of AI-generated disinformation on public perception and trust.

To conduct the study, researchers turned to OpenAI’s renowned language model, GPT-3. The AI was assigned to generate tweets with either informative or misleading content on various topics prone to misinformation, such as vaccines, 5G technology, Covid-19, and the theory of evolution. Real tweets on the same subjects were also collected for comparison.

A total of 697 members participated in an online quiz, where they had to determine whether the tweets presented to them were AI-generated or collected from Twitter. They also had to assess the accuracy of the content. Surprisingly, participants were 3% less likely to believe false tweets from humans than from AI.

Giovanni Spitale, the lead researcher from the University of Zurich, expressed uncertainty about the reasons behind this phenomenon. However, the study suggested that the way GPT-3 structures information might contribute to the credibility of AI-generated content. Notably, participants found it difficult to differentiate between AI-generated and organic tweets, highlighting the advanced capabilities of GPT-3.

While participants were more effective at identifying misinformation from real Twitter users, GPT-3-generated tweets with false information managed to deceive them slightly more.

The study serves as a wake-up call, and the findings highlight the influence of AI-generated content and the challenges it poses in shaping public opinion. As AI technology continues to advance, it becomes crucial to develop strategies and safeguards to tackle the spread of AI-generated disinformation.

Government soon to Open Draft Digital India Bill for consultation

By

Tejaswini Kasture

-

June 29, 2023

Government to open draft Digital India Bill for consultation — Image Source: techobserver

The Ministry of Electronics and Information Technology (MeitY) held the first-ever consultation on the proposed Digital India Act on March 10th. Further, they discussed rules for handling data captured by invasive devices and stated the draft of the Act to be passed by April 2023.

The bill was due to be released this month but was held back for additional consultations. However, the much-awaited draft of the Digital India Bill for public consultation will be released within the next two weeks.

Currently, there are over 850 million internet users in India, marking a significant increase from the 5.5 million in 2000. Over the years, the internet has witnessed a rise in cyber crimes, violent content, and improper use of government-issued identity cards, and the old Act is inadequate to address these concerns. Hence, certain provisions are needed to update and re-evaluate rising situations.

Amid the growing threat of misleading information, the Indian government plans to enforce more innovation, and startups, at the same time, protect citizens’ data in terms of security, accountability, and trust.

The new Digital India Act 2023 aims to replace the existing Information Technology Act (IT Act) of 2000. It aims to provide a strict legal framework to support India’s techade goals, taking into consideration the need to protect users from the growing threat of AI-related misinformation. The Act will cover a range of topics, such as Artificial Intelligence, cybercrime, data protection, online safety, and deepfakes. The need for responsible and ethical use of online technologies will not only shape the technological future of India but also safeguard users’ confidential information.

Meta Introduces $250,000 Mixed Reality Fund for 5 Indian Startups

By

Sahil Pawar

-

June 29, 2023

Meta $250,000 mixed reality fund 5 Indian startups — Image Credits: AD

On Wednesday, Meta revealed a brand-new mixed reality (MR) programme for India that includes a $250,000 award for local companies and developers to create experiences and apps. According to a statement from the corporation, the fund aims to support creativity and create an ecosystem for extended reality (XR) technology in the nation.

The fund, according to the company, will assist developers in creating apps for the Meta Quest Platform and metaverse using Meta’s Presence Platform, which offers a wide range of machine perception and AI capabilities.

A chosen group of up to five Indian businesses and developers will benefit from the program’s financial grant as well as specialized mentoring and advice from the specialists at Meta Reality Labs. They will also have the chance to join Meta’s expanding developer ecosystem by uploading their creation to Quest App Lab, according to Meta.

At the conclusion of the programme, the top local venture capital funds will be introduced to the startups who made the short list. The fund is already accepting applications, encouraging established and up-and-coming entrepreneurs and developers to test out the capabilities of the Presence Platform.

“Meta is dedicated to creating the India XR ecosystem. According to Sandhya Devanathan, vice president for Meta in India, “The Presence Platform is a key component of the metaverse vision, seeking to make virtual experiences more tangible and accessible.”

“We think that this fund and programme will hasten the development and uptake of VR and MR experiences in the nation and provide Indian developers with a chance to make their innovations global,” Devanathan continued.

Infosys Using Generative AI to Handle 50 client Projects: CEO Salil Parekh

By

Sahil Pawar

-

June 29, 2023

Infosys using generative AI handle 50 client projects — Image Credits: Infosys

Salil Parekh, chief executive officer of Infosys, claims that the company, which just unveiled an AI-first service, is utilizing generative AI in 50 customer projects. Parekh said, “We currently have 50 active client projects where we are using generative AI, and this is increasingly becoming a part of the new world that will be defined on an AI-first basis.”

During Infosys’ annual general meeting on June 28, shareholders raised a number of important issues with the company’s leadership, including the usage of AI and concerns about productivity, hazards, and more.

The business is also doubling in on AI, which explains all the concerns. The company said that it was in discussions regarding Infosys Topaz, a new product that blends data analytics, artificial intelligence, and generative AI, with over 300 clients at the time of its debut.

“We will revolutionize our clients with AI by utilizing our Topaz infrastructure and AI-first approach. We’ll use responsible AI to accomplish that, making sure it abides by legal requirements as well as privacy and accuracy standards. With the right safeguards, we think AI can be a force for good,” Chairman Nandan Nilekani stated.

Chairman Nandan Nilekani issued a warning in his speech, pointing out that while generative AI has made significant strides, there are still a number of practical, ethical, and intellectual property-related challenges that need to be addressed.

“We are aware that scaling AI in the workplace is not going to be easy. But the AI-first approach we’ve adopted is already paying off for us. We have restructured our organization to carefully balance people, technology, and procedure while also segmenting work into cognitive tasks to carry out,” he stated.

Alphabet and Airtel Partner to Bring Affordable Internet to Indian Villages

By

Sahil Pawar

-

June 29, 2023

Alphabet Airtel partner bring affordable internet Indian villages — Image Credits: Taara

Google declared at the beginning of 2022 that it will invest $1 billion in Airtel, of which $700 million was to be used to acquire a 1.28% interest in the company. The parent firm of Google, Alphabet, is now working with Bharti Airtel with the aim of bringing reasonably priced internet to rural areas across the nation.

This company’s goal will be carried out through the Taara project, a component of Alphabet’s X innovation lab, which aims to offer faraway people high-speed internet. The plan is to employ light beams to distribute internet services, and according to the business, Taara’s light beams will give its users fiber-optic internet without the need for cables.

Taara’s executive director, Mahesh Krishnaswamy, claims that because of Airtel, the company is almost ready to begin a large-scale deployment of the technology and its services across the nation.

In addition to Airtel, Taara has agreements with another Indian internet service provider, Bluetown. Additionally, Project Taara is being developed in 13 additional nations, including Australia, Fiji, and Kenya.

Randeep Sekhon, CEO of Airtel, has said that the firm would be able to provide faster internet services in metropolitan areas as well with the aid of the most recent technology created by Taara and that it is less expensive than burying fiber-optic cables beneath the surface.

Taara lines provide an affordable and speedy deployment option for bringing high-speed connectivity to towns and cities. Taara linkages have the ability to assist thousands of people in accessing the educational, commercial, and communicative advantages of the web by filling important gaps to key access points like cell towers and WiFi hotspots.

Harvard University Employs AI Chatbot as CS Instructor

By

Amulya Reddy Avanaganti

-

June 29, 2023

Harvard University Employs AI Chatbot as CS Instructor — Image Source: Howtoai.ai

It seems that the jobs of teachers are in danger as AI continues to make advancements. Harvard University is taking bold steps in integrating artificial intelligence (AI) into its coding program. The university has announced its plan to employ an AI chatbot equipped with ChatGPT’s capabilities as an instructor in its renowned course “Computer Science 50: Introduction to Computer Science (CS50).”

According to course instructors, the AI teacher will be developed using OpenAI’s advanced GPT 3.5 or GPT 4 models. The program is set to launch in September, and enrolled students will be encouraged to utilize this AI tool.

Professor David Malan, who teaches CS50, expressed hope for the AI integration, stating, “Through AI, we can eventually approximate a 1:1 teacher-student ratio for every CS50 student, providing them with software-based tools that support their learning at their own pace and style.” He highlighted the challenge of providing personalized support to online students and emphasized the benefits of AI chatbots for both on-campus and off-campus learners.

This move comes at a time when AI tools are gaining immense popularity. OpenAI’s ChatGPT, launched in November 2022, quickly became the fastest-growing app of all time, attracting 100 million active users within two months. Users are drawn to ChatGPT’s versatility, from generating computer code to composing poetry and essays. Google DeepMind’s CEO recently announced the development of Gemini, a futuristic algorithm that can outperform OpenAI’s ChatGPT.

However, concerns over accuracy and reliability persist with AI-generated content. Google recently cautioned users that its AI-powered chatbot Google Bard may not always provide correct information.

Professor Malan recognizes these concerns and emphasizes the importance of critical thinking when interacting with AI-generated content. He encourages students to exercise their own judgment while evaluating information. Nevertheless, he remains optimistic about the future of these tools, highlighting the value of feedback from students and teachers in refining AI’s capabilities.

Harvard University’s employment as an AI chatbot instructor marks a significant milestone in integrating AI and education, paving the way for potential transformations in teaching and learning methodologies.

Celestial AI Raises $100M in Series B Funding Round

By

Sahil Pawar

-

June 29, 2023

Celestial AI raises $100M Series B funding round — Image Credits: Celestial AI

A $100 million Series B financing headed by IAG Capital Partners, Koch Disruptive Technologies, and Temasek’s Xora Innovation fund was announced by Celestial AI on Wednesday.

The additional funding, which increases Celestial AI’s total amount collected to more than $165 million, will be applied to expanding the company’s engineering, sales, and technical marketing divisions in order to assist the development of Celestial’s photonics platform. Currently, Celestial employs about 100 people, but founder Lazovsky predicts that figure will increase to 130 by the end of the year.

Technology experts David Lazovsky and Preet Virk, who have backgrounds in semiconductor engineering and photonics, realized several years ago that workloads for AI and machine learning would quickly run into a data movement problem. They projected that if AI models became larger than what could be stored on the die of any one memory chip, it would get harder to transfer data to and from computational devices.

They came up with an optical interconnect technique for compute-to-compute, compute-to-memory, and on-chip data transmission, which was designed by Phil Winterbottom, a former researcher at the prestigious Bell Labs. Lazovsky, Virk, and Winterbottom established Celestial AI as a firm to market the technology. The startup is now attracting significant investors.

“Compute and memory are now tightly connected,” Lazovsky said. “Whether the additional compute is required or not, the only method to add more high bandwidth memory is to add more compute. Memory disaggregation is enabled by Celestial’s technology.”

Grammy Awards Ban AI-Only Music, Encourage Human Creativity

By

Amulya Reddy Avanaganti

-

June 29, 2023

Grammy Awards Ban AI Music — Image Source: Whistler.ng

The Recording Academy, responsible for the renowned Grammy Awards, recently announced that only human creators would be eligible for the prestigious music event. In an effort to limit the use of artificial intelligence (AI) in the industry, the academy’s updated rulebook states that “a work that contains no human authorship is not eligible in any categories.”

However, music created with AI assistance may still qualify in certain categories as long as significant human authorship is involved. To earn a nomination for album of the year, music creators must now contribute at least 20% to the album. Previously, individuals with minimal input, such as producers, engineers, or featured artists, could receive nominations.

By revising the eligibility criteria for the Grammy Awards, the Recording Academy hopes to strike a delicate balance. While AI can be used as a tool to enhance music creation, the human contribution must remain substantial and meaningful. The decision encourages artists to rely on their creativity and personal touch rather than solely depending on artificial intelligence.

The rise of AI technology, particularly since the launch of OpenAI’s ChatGPT in November 2022, has sparked concerns across various creative industries. As AI applications continue to rise, concerns arise about their potential to displace human workers in various jobs, including animation, writing, and music creation.

The Writers Guild of America (WGA) and the Screen Actors Guild (SAG-AFTRA) grapple with AI’s role in creative fields. The WGA seeks to limit AI in screenwriting, while SAG actors aim to control their digital personas and secure fair compensation. The WGA went on strike in May, with AI usage as a key unresolved issue. Actors may also strike if SAG-AFTRA fails to address similar concerns. These industry organizations are navigating the complexities of AI’s impact on creativity and labor rights.

The Grammy Awards, established in 1959, have evolved into one of the music industry’s most esteemed accolades. This annual event was created to honor the accomplishments and artistic contributions of talented musicians, solidifying its reputation as a prestigious recognition over the years.

DarkBERT: New AI Trained on Dark Web to Combat Cyber Crime

By

Amulya Reddy Avanaganti

-

June 28, 2023

DarkBERT AI Dark Web Cyber Security — Image Source: Assetguardian

South Korean researchers have recently unveiled DarkBERT, a language model exclusively trained on dark web datasets. It was developed in collaboration between the Korea Advanced Institute of Science and Technology (KAIST) and Data Intelligence Organization S2W. DarkBERT utilizes the powerful BERT framework developed by Google.

Unlike popular chatbot models like ChatGPT or Google Bard, DarkBERT has been specifically trained to analyze and interpret data from the dark web.

Researchers developed DarkBERT with the aim of improving cybersecurity measures with the dark web. They fed the AI model a large amount of data for almost 16 days, divided into two sets.

The first set consisted of “raw” data, which was unedited content found on the dark web. The second set, known as the “preprocessed” data, underwent some modifications. Certain elements commonly found on the dark web, such as victim organization names, descriptions of leaked data, and threat statements with sample data, were removed from the preprocessed set. Including raw and preprocessed data offers a comprehensive understanding of the hidden aspects of the internet while maintaining ethical considerations and privacy standards.

DarkBERT’s advanced analytical capabilities enable the detection and monitoring of data leaks, ransomware sales, illegal drug trade, and more. By delving into these hidden operations, DarkBERT provides crucial insights to cybersecurity experts, equipping them to combat online wrongdoing more effectively.

While DarkBERT remains inaccessible to the general public, the research team recognizes the importance of academic access to its dataset. Scholars can request access to DarkBERT’s data, facilitating further research and advancements in cybersecurity, all while respecting the sensitive nature of dark web materials.

As we move forward into the future, one thing remains clear—DarkBERT’s influence and impact on the fight against cybercrime is undeniable. It is a beacon of hope, driving the path toward a safer and more secure digital future for all.

Google DeepMind’s CEO Says They Can Surpass ChatGPT

By

Amulya Reddy Avanaganti

-

June 28, 2023

Google DeepMind CEO Gemini Surpass ChatGPT — Image Source: AITELLIGENCE

DeepMind was acquired by Google in 2014, and it showcased impressive achievements by employing reinforcement learning in mastering basic video games. In subsequent years, Google’s DeepMind made history with its AI program AlphaGo, defeating a champion player in the board game “Go” in 2016.

Demis Hassabis, the CEO of DeepMind, has recently revealed the development of Gemini, an advanced AI system that draws inspiration from the techniques used in AlphaGo. Gemini is a large language model similar to GPT-4. This groundbreaking development aims to surpass OpenAI’s ChatGPT and redefine the capabilities of language models.

Hassabis anticipates that Gemini’s development will take several months and require a substantial investment, potentially reaching tens or hundreds of millions of dollars. In April, Sam Altman, CEO of OpenAI, disclosed that the development of GPT-4 cost more than $100 million.

Hassabis says, “At a high level, you can think of Gemini as combining some of the strengths of AlphaGo-type systems with the amazing language capabilities of the large models. We also have some new innovations that are going to be pretty interesting.”

In April, Google merged DeepMind with their primary AI lab, Brain, forming the powerhouse known as Google DeepMind. Hassabis believes that this collaboration between two influential entities will consolidate their expertise and significantly contribute to the continued advancements in the field of AI.

Hassabis also shared his perspective on the risks and benefits of AI, stating, “The recent, rapid advancements in language models have made many AI experts—including some building the algorithms—worried about whether the technology will be put to malevolent uses or become difficult to control.” However, he firmly believes in the positive potential of AI, adding, “If done correctly, it will be the most beneficial technology for humanity ever. We’ve got to boldly and bravely go after those things.”

He stated, “I can see the kinds of things we’re building into the Gemini series right, and we have no reason to believe that they won’t work.”