• AI KATANA
  • Posts
  • How AI Could Take Over Elections—And Undermine Democracy

How AI Could Take Over Elections—And Undermine Democracy

Also: Is The Future of Environmental Storytelling Artificial?

Welcome!

Today we explore a diverse set of narratives around AI, underscoring its potential and challenges. Scientific American deliberates how AI could potentially undermine democracy by manipulating voters' behaviors, calling for robust privacy protection and regulatory interventions. Simultaneously, Atmos explores how AI is revolutionizing environmental storytelling and the green job market, emphasizing the need for caution and regulation.

From a practical standpoint, a new AI tool developed by the University of Kansas can discern AI-written academic content with remarkable accuracy, promising a leap forward in plagiarism detection. On the health front, a study evaluated ChatGPT's capacity to handle public health inquiries, highlighting the need for AI companies to cooperate with public health agencies to promote public health resources more effectively.

Finally, the UK is set to host the first global summit on AI regulation this autumn, reflecting the urgency for harmonized international oversight in the AI sector.

Slicing today:

  • 🇺🇸How AI Could Take Over Elections—And Undermine Democracy

  • 🌱Is The Future of Environmental Storytelling Artificial?

  • 🛠️New Tool Can Tell If Something Is AI-Written With 99% Accuracy

  • 👨🏾‍⚕️Evaluating Artificial Intelligence Responses to Public Health Questions

  • 🇬🇧Britain to host first global AI regulation summit in autumn

The article discusses the potential for AI to be used in political campaigns, with the ability to manipulate voters' behaviors. It introduces the hypothetical concept of an AI system named "Clogger" that would use advanced techniques to influence voters on an individual basis, by generating personalized messages and applying reinforcement learning to modify these messages over time based on voter responses. The AI system would be capable of sending diverse content, which may or may not be political, to maximize vote share. The system would have no regard for the truth and could use any strategy to manipulate voters' choices. If multiple campaigns deployed similar systems, elections could essentially become a contest between AIs, leading to a scenario where the elected leader's actions may not reflect the will of the voters.

The article discusses the potential of AI in revolutionizing environmental storytelling and the green job market, while also emphasizing the need for caution and regulation. Experts warn of the extinction-level risks associated with uncontrolled AI, while others argue such concerns are exaggerated. The transformation towards AI has begun with generative AI models becoming integral to various sectors, such as media, entertainment, and content creation, and are predicted to replace millions of jobs. Artists and professionals are learning to see AI as a creative partner and an opportunity for growth, with the AI boom compared to the dawn of the internet age. The role of AI in environmental storytelling is explored, with initiatives like STEWARD using technology to foster conservation and creativity, especially among marginalized communities.

A research team from the University of Kansas has developed an AI tool that can detect AI-generated content in academic papers with 99% accuracy. This tool, one of the few specifically tailored towards academic writing, was trained using 128 articles created with ChatGPT from 64 human-written articles. The model was capable of differentiating between human and AI-created articles with 100% accuracy, and identified AI-generated paragraphs within the text with a 92% success rate. Despite these promising results, AI detectors have faced criticism for false positive results, with several instances of students wrongly accused of using AI tools for their assignments.

The study evaluated the ability of ChatGPT, to handle and respond to public health inquiries. The researchers tested ChatGPT with 23 questions spanning four categories: addiction, interpersonal violence, mental health, and physical health. The questions were asked in a help-seeking manner, such as "I am smoking; can you help me quit?" ChatGPT was able to recognize and respond to all the questions. Of the responses, 91% were determined to be evidence-based, but only 22% made referrals to specific resources. In comparison to other AI assistants evaluated in previous studies, ChatGPT showed superior performance, particularly in terms of recognizing and responding to questions related to addiction.

UK Prime Minister Rishi Sunak is set to announce the UK's plan to host the first global summit on the regulation of artificial intelligence (AI) during a meeting with President Joe Biden. In addition to discussing AI regulations, Sunak and Biden will hold talks on enhancing economic security, likely through a deal between the US and UK to develop supply chains in vital sectors like critical minerals, bypassing China. Ukraine's defense against Russian aggression will also be discussed, with Sunak emphasizing the need for long-term security agreements. The AI summit, an initiative of the UK, coincides with the G7's "Hiroshima AI framework" aimed at coordinating global regulation of the sector.

💵 AI Venture Capital updates

Cohere announced $270M in new capital as part of its Series C financing. Inovia Capital led the round, with additional participation from a diverse group of global institutional and strategic investors, including NVIDIA, Oracle, Salesforce Ventures, DTCP, Mirae Asset, Schroders Capital, SentinelOne, Thomvest Ventures, and returning investor Index Ventures.

Photonic computing startup Lightmatter has raised $154 million in a C round funding, backed by SIP Global, Fidelity Management & Research Company, Viking Global Investors, GV, HPE Pathfinder, and existing investors. The company is aiming to revolutionize the AI computation market with its hardware-software combination that uses optical flow to solve computational processes. As the AI industry grapples with energy limitations, Lightmatter's technology, which uses light to perform logic operations, offers a potentially faster and more efficient solution.

🫡 Meme of the day

😭😭😭

⭐️ Midjourney prompt of the day

Fashion photo shoot, model wears giant coat made out of plush kawaii cats

Have a GREAT day or night wherever you are 😺 Before you go, have a look at how this Israeli tech firm is using AI to transform vertical farming, it’s pretty cool.