• AI KATANA
  • Posts
  • A Developer Built an AI-Powered 'Propaganda Machine' to Highlight Disinformation Dangers

A Developer Built an AI-Powered 'Propaganda Machine' to Highlight Disinformation Dangers

Also: Venice Film Festival 2023 Is Cinematic AI’s Coming Out Party

Author

AI KATANA
September 04, 2023

Hi!

In this today's edition, we delve into the increasingly intricate landscape of Artificial Intelligence, capturing both its opportunities and ethical dilemmas. From a developer's eye-opening creation of an AI-driven "propaganda machine" to highlight disinformation dangers, to the glowing endorsements of AI's role in the future of cinema at the Venice Film Festival, the applications of AI are ever-expanding and deeply impactful. As AI steps up to aid in reducing the carbon footprint of buildings, and new supercomputing powers emerge in the UAE, we also examine updates in AI tools and venture capital movements that are shaping the industry. Whether you view AI as a tool for unprecedented growth or a Pandora's box of challenges, there's no denying its profound influence across multiple sectors.

Sliced:

  • 🤖 A Developer Built an AI-Powered 'Propaganda Machine' to Highlight Disinformation Dangers

  • 🎥 Venice Film Festival 2023 Is Cinematic AI’s Coming Out Party

  • 🌿 Artificial Intelligence Steps In to Lower Carbon Footprint of Buildings

  • 🇦🇪 UAE supercomputer powers up to fight global rivals

A developer who went by the pseudonym Nea Paw built CounterCloud, an AI-driven "propaganda machine" in two months, utilizing OpenAI tools like ChatGPT. Costing less than $400 a month to run, the system autonomously generates "convincing content 90% of the time," according to Paw. Intended as an educational experiment, the machine demonstrates how easily and inexpensively disinformation can be mass-produced. By inputting "opposing" articles and instructing the AI to produce counter-articles, the system can create fake stories and generate doubt about the accuracy of original articles. Although not set live on the internet due to ethical concerns, the project aims to raise awareness about the potential misuse of AI in spreading disinformation.

The 90th Venice Film Festival served as a pivotal moment for the integration of AI in cinema, contrasting Hollywood's recent trepidation toward the technology. Notable Hollywood figures like Tom Hanks, Robert Zemeckis, Joe Russo, and Darren Aronofsky endorsed AI's cinematic capabilities, with Hanks even contemplating an AI version of himself acting posthumously. The festival saw the launch of three AI film studios—EDGLRD, Pillars, and The Simulation—each pioneering various AI technologies in filmmaking, such as deepfakes, AI VFX, and text-to-episode conversions. Despite ongoing labor strikes in Hollywood centered around AI fears, these endorsements and technological advancements suggest a growing perception of AI not merely as a threat, but also as a transformative opportunity for the future of cinema.

AI is increasingly being employed to reduce the carbon footprint of buildings, which accounted for 26% of global energy-related greenhouse-gas emissions in 2022. Companies like JLL are investing in AI systems, such as Turntide's electric motors and Envio Systems' sensors, to make buildings smarter and more energy-efficient. These AI systems can help regulate lighting, heating, and cooling based on occupancy and usage patterns, potentially reducing a building's energy consumption by over 20%. Despite the promise, challenges exist in retrofitting older buildings and ensuring quality data collection, as AI efficacy relies heavily on accurate data. Governments, venture capital firms, and industry leaders see AI as a valuable tool for achieving more sustainable building practices, although it is considered one piece in the broader puzzle of environmental sustainability.

A new supercomputer, Condor Galaxy 1 (CG1), has powered the development of Jais, a 13-billion parameter language model proficient in Arabic and English. Created in collaboration between Mohamed bin Zayed University of Artificial Intelligence, Cerebras Systems, and UAE-based AI company G42, the model is reportedly more cost-effective and efficient than its global competitors. Capable of supporting up to 600 billion parameters, CG1 has drawn attention for its potential to challenge industry giants like Nvidia in high-performance computing. Jais aims to serve various sectors, including government and healthcare, in both Arabic and English, with plans for expansion into other languages and media formats. This development aligns with broader investment in AI and semiconductors in the Gulf states and poses an alternative in a market largely dominated by Nvidia.

🛠️ AI tools updates

Reddit has introduced Keyword Suggestions, a new feature in its Ads Manager designed to enhance ad targeting. Utilizing machine learning and natural language processing, the tool generates and ranks keyword suggestions by monthly Reddit views while also filtering out unsuitable content. This AI-driven approach aims to streamline the keyword research process for advertisers, enabling more efficient and targeted advertising campaigns by recommending keywords that resonate with Reddit's user base.

💵 Venture Capital updates

Sapeon, a Korean AI company spun out from SK Telecom, SK Hynix, and SK Square, has raised $46 million in a Series A round led by Ascent Equity Partners and other notable investors. The funding values the company at $380 million and will support its AI full-stack strategy, which spans AI chip development, software, and services. Sapeon has already launched its X220 AI chip capable of processing 6,700 images per second and plans to release next-generation chips with enhanced performance and self-learning capabilities. The company is collaborating with various partners to apply its technology in fields such as OCR for banking, image/video quality improvement, and smart transportation, among others.

🫡 Meme of the day

⭐️ Generative AI image of the day