AI Headlines: Unveiling of Falcon 180b and More

AI Headlines: Unveiling of Falcon 180b and More

Introduction

AI headlines have been dominating the news this week, and we have gathered all the latest updates for you. In this blog post, we will cover the unveiling of Falcon 180b, a new large language model that is making waves in the AI scene. We will also explore the latest news on Chat GPT's website traffic, Opera's new AI chatbot for gamers, Mind Vis' groundbreaking system that turns brain scans into visuals, Microsoft's partnership with Paage for cancer detection, Snapchat's imaginative AI feature named Dreams, Ghost Writer's new AI-generated song, Dubai's new AI campus, Zoom's AI companion rollout, IBM's Granite series, and the introduction of the world's first humanoid pilot robot, Pibot. So, let's dive into this week's AI whirlwind!

Unveiling of Falcon 180b: The Game Changer

Let's start with the biggest news in the AI world – the unveiling of Falcon 180b. This new large language model, released by TI, is creating a buzz in the AI community. With a staggering 180 billion parameters, Falcon 180b is currently the largest open-source model available. It outperforms other models like Llama 270b and Open AI GPT 3.5 in certain tests and matches the performance of Google's Palm 2 in others.

The training process for Falcon 180b was a monumental task. TI used 3.5 trillion tokens from a dataset called Refined Web, and it took around 7 million hours on Amazon's powerful computers. The dataset included a variety of sources, such as web data, conversations, and technical papers, to ensure the model's understanding of different topics.

What sets Falcon 180b apart from its predecessor, Falcon 440b, is its specialization in chat and instruction-giving. Even when compared to models kept private by other companies, Falcon 180b performs exceptionally well. You can try it out on the Hugging Face website, but be aware of the rules for commercial use. Responsible and ethical use of Falcon 180b is of utmost importance in order to unlock its full potential.

Chat GPT's Website Traffic and Usage

Next, let's take a look at the recent trends in Chat GPT's website traffic and usage. In August, the number of people visiting the Chat GPT website decreased by 3.2%, with a total of 1.43 billion visits. This decline is consistent with the decrease observed in the previous two months. Additionally, the average time spent on the site decreased from 8.7 minutes in March to 7 minutes in August.

However, there is a silver lining. The number of unique visitors to the site slightly increased in August, with a rise from 180 million to 180.5 million. This boost can be attributed to the adoption of Chat GPT by schools, especially in the US, where students are using the chatbot for homework help. The summer slump in younger users seems to be reversing, and the launch of the Chat GPT app for iOS in May may have contributed to the decrease in website visits.

Opera's AI Chatbot for Gamers: Arya

Opera, known for its innovative browsing solutions, has introduced Arya, an AI chatbot designed specifically for gamers. Arya is built on Open AI's Chat GPT and utilizes Opera's composer technology. This AI chatbot can provide answers to various gaming-related questions, such as game launches, streaming career advice, custom character bios, and updates on favorite games.

Arya is available in Opera's main browser, Opera GX, which is tailored for gamers. It is also accessible through the Opera app for Android and iOS, reaching users in over 180 countries. However, users need to activate Arya by selecting the early bird option in Opera's settings and turning on the Arya extension and command line. This AI chatbot opens up exciting possibilities for gamers looking for personalized assistance and information.

Mind Vis: Turning Brain Scans into Visuals

In Singapore, researchers have developed a groundbreaking AI system called Mind Vis that can interpret brain scans and convert them into visuals. This system works by scanning a person's brain while they view specific images, capturing the brain's response. The AI then translates this response into a unique language and processes the information to recreate the images the person was seeing.

Although Mind Vis is an impressive achievement, it is important to note that it is not yet capable of understanding the thoughts of the general public. Every individual's brain structure and function vary, making it challenging to generalize mind reading capabilities. However, researchers remain optimistic about the potential benefits of this technology, especially for individuals with disabilities. Mind Vis could offer them a newfound sense of independence and connection by allowing them to control devices and communicate through their thoughts.

Open Interpreter: Your AI-Powered Digital Sidekick

Have you ever dreamed of having an AI-powered sidekick for your tech endeavors? Open Interpreter might just be the answer. Open Interpreter is a groundbreaking tool that bridges the gap between powerful AI models and everyday computing needs. It allows language models like Open AI's Chat GPT to run code on your computer, enabling you to have conversations and execute commands for various tasks.

With Open Interpreter, you can ask your computer to create or edit photos, videos, and PDFs, control a Chrome browser for research purposes, and simplify tech jargon. Unlike some cloud-hosted services, Open Interpreter operates in your local environment, giving you full internet access, no restrictions on package installations, and flexibility in software usage.

Safety is a priority when running code on personal machines, and Open Interpreter includes a built-in feature that seeks user confirmation before executing any code. While it offers the option to bypass this safeguard, caution is advised when executing commands that may change system settings or files. Open Interpreter opens up a world of possibilities for tech enthusiasts and simplifies complex tasks.

Microsoft's Partnership with Paage for Cancer Detection

Microsoft has joined forces with Paage, a company specializing in digital pathology, to create the world's largest AI model for spotting cancer. This AI model is being trained on billions of images and can detect both common and rare types of cancer. The goal of this collaboration is to assist doctors who are facing an increasing number of patients and a shortage of staff.

Paage has developed a tool called Full Focus, which allows pathologists to view tissue samples on a screen instead of under a microscope. This digital approach improves accuracy and reduces the chances of missing important details. By leveraging AI, Paage aims to enhance the efficiency and effectiveness of cancer detection and diagnosis.

However, digitizing tissue samples requires significant storage capacity, making it expensive. As a result, only larger medical centers have been able to adopt this digital approach. With the help of Microsoft, Paage is building an AI model using images from 4 million slides, making it the largest of its kind. This collaboration has the potential to revolutionize cancer care and improve patient outcomes.

Snapchat's AI Feature: Dreams

Snapchat, a leading social media platform, has recently launched an AI feature named Dreams. This feature allows users to modify their selfies and create imaginative images that alter their appearance. Users can edit up to eight photos for free, offering a fun and entertaining experience.

Jessica Melen, an expert in technology and innovation, acknowledges the fun aspect of Snapchat's Dreams feature, especially for younger users. However, she also highlights the potential risks associated with generative AI, such as the rapid spread of misinformation, particularly during elections.

While Snapchat has been pioneering AI tools for users, other social media platforms like Instagram, Facebook, and TikTok have yet to offer similar features. Melen believes that it is only a matter of time before these platforms catch up and emphasizes the importance of innovation and progress for the economy and quality of life. She also notes that Snapchat currently faces no major legal issues in Washington.

Ghost Writer's AI-Generated Music

Ghost Writer, an AI creator known for generating songs, has released a new track called "Whiplash." This song features the AI-generated voices of Travis Scott and 21 Savage. Ghost Writer previously gained recognition for creating a song in Drake's AI-generated voice.

The future of music could be shaped by AI-generated voices, eliminating the need for singers to physically perform. Ghost Writer has offered to share the earnings from "Whiplash" with Travis Scott and 21 Savage if they agree to its release. However, copyright issues have been a concern for Ghost Writer in the past, as his AI-generated song "Heart on My Sleeve," a Drake song, was removed from platforms like Spotify and Apple Music.

The rise of AI-created music raises legal challenges, and platforms like YouTube struggle to define what is allowed and what isn't. Despite the legal complexities, Ghost Writer aims to get "Heart on My Sleeve" nominated for Grammy awards, specifically in the categories of Best Rap Song and Song of the Year. The recording industry recognizes the efforts of human creators, but the eligibility of AI-generated music is still being debated.

Dubai's AI Campus: A Hub for Cutting-Edge Technologies

Dubai, known for its visionary projects, is building a new campus dedicated to cutting-edge technologies like artificial intelligence and web 3. The Dubai AI and Web 3 Campus, set to be completed by 2028, aims to host over 500 companies and create more than 3,000 jobs.

Muhammad Al Blushi, the head of Dubai International Financial Center Innovation Hub, shares that this campus is part of Dubai's strategy to become a world leader in AI and web 3 technologies. The campus, located in the DC Innovation One space, will span an impressive 100,000 sq ft and will offer modern research and development facilities, programs to accelerate business growth, and shared workspaces.

Dubai International Financial Center (DFC) is a major financial hub operating under English common law. With over 39,000 professionals working in more than 4,900 companies, the DFC is a magnet for industry talent. The DFC Innovation Hub has a track record of supporting new tech companies, particularly in the metaverse and AI domains. Recently, the DFC selected 10 startups for its DFC Metaverse Accelerator Program, further solidifying Dubai's position as a top business destination for tech companies.

The Dubai AI and Web 3 Campus is already in operation, with companies being granted licenses to operate there. The full launch of the campus is expected in the coming months, and it holds great promise for fostering innovation and driving the advancement of AI and web 3 technologies.

Call for a Federal Department of AI

AI specialists from top US universities are advocating for the establishment of a federal department of AI to oversee artificial intelligence. This recommendation follows a survey conducted by Axios Generation Lab and Syracuse University, which sought the opinions of leading computer science professors.

The survey revealed varied opinions on AI's future control, with most experts asserting that AI requires oversight. Many experts lack trust in the private sector's ability to self-regulate. President Biden emerged as the top individual trusted with AI issues. These insights come just before key AI discussions on Capitol Hill involving tech leaders.

Experts express concerns about AI exacerbating discrimination and disparities, potentially leading to job reductions in customer service but boosting sectors like technology and healthcare. While some experts support global organizations or Congress overseeing AI, a majority favor the establishment of a dedicated federal agency to regulate AI.

The White House has proposed an AI Bill of Rights, which represents an important step in addressing AI's ethical and legal challenges. It is worth noting that these views represent the opinions of experts and not the general public. The survey's main takeaway is the strong support for a specific AI department.

Zoom's AI Companion for Enhanced Productivity

Zoom, the popular video conferencing platform, is rolling out a new feature called the Zoom AI Companion. This smart assistant aims to make your work on Zoom easier and more efficient. The AI companion is included for free in paid Zoom plans, providing assistance in various aspects of video conferencing.

The Zoom AI Companion helps users catch up on missed meeting details, aids in chat responses, and summarizes recordings with highlights and actionable steps. In the future, it will expand its capabilities to assist with emails, chat message summaries, and quick replies. During meetings, the AI companion can help with real-time tasks like filing support tickets and looking up information. It can also provide feedback on communication style, helping users improve their skills.

Zoom prioritizes privacy and reassures users that their data will not be used to train AI models. Companies can choose which AI features they want to activate, and meeting hosts have control over the in-meeting AI tools. The Zoom AI Companion is available to users with paid Zoom plans and promises to enhance productivity and streamline workflows.

IBM's Granite Series: New Generative AI Models

IBM has recently unveiled its new generative AI models, the Granite series, on its Watson X platform. These models, similar to Open AI's GPT 4 and Chat GPT, utilize enterprise data and can be specialized for various tasks, including finance. IBM plans to share more about GR's training data by Q3 2023.

The Tuning Studio, a new feature in Watson X, allows users to customize AI models according to their specific needs. Additionally, Watson X offers a synthetic data generator to reduce risks in AI training. Through the Watson X.A interface, users can access AI tools via a chatbot, and the platform supports retrieval augmented generation.

IBM is also actively working on Watson X.G governance to protect customer privacy and detect model biases. Despite some revenue challenges, IBM remains optimistic about the future of AI, with many businesses already utilizing their services. The Granite series represents another step forward in AI technology.

Meet Pibot: The World's First Humanoid Pilot Robot

In a groundbreaking development, the world's first humanoid pilot robot, Pibot, has been introduced. Pibot has the unique ability to fully operate an airplane based on human language manuals. Unlike other systems that can only control specific aspects of flight, such as speed or altitude, Pibot can manage all aspects, from starting the engines to landing.

Human pilots possess contextual understanding and excel in handling complex situations. However, they may not always recall every detail from flight manuals, especially in emergencies. Pibot, on the other hand, can instantly access and remember all the information from the manual, ensuring precise and emotion-free responses, even in critical situations.

Pibot is designed in a humanoid form, allowing it to fit into spaces meant for humans, such as car seats, driver seats, cockpits, and other operator positions. This versatility enables Pibot to operate not just airplanes but also cars, tents, excavators, and submarines. The ability to recall detailed instructions and act without panic makes Pibot a valuable tool in various operations.

Conclusion

In this week's AI whirlwind, we covered a range of fascinating updates in the world of AI. From Falcon 180b, the largest open-source language model, to the introduction of Pibot, the world's first humanoid pilot robot, these advancements are pushing the boundaries of what AI can achieve. We also explored the latest developments in AI chatbots, brain scan interpretation, cancer detection, AI-generated music, AI campuses, and more.

As AI continues to evolve and shape various industries, it is crucial to embrace responsible and ethical use. These breakthroughs hold immense potential to improve our lives, but it is essential to navigate AI's challenges, such as privacy concerns and legal implications. By staying informed and promoting responsible AI practices, we can truly harness the power of AI for the benefit of society.

Post a Comment

0 Comments