GT Protocol AI Digest №4: Innovations, Risks, and Global Shifts

Intro

We’re back with another packed digest, highlighting the biggest AI stories from the past week. From groundbreaking innovations to unexpected controversies and regulatory debates, AI continues to shape industries in ways we couldn’t have imagined just a few years ago. This edition captures the most important developments that are shaking up the AI world — there’s a lot to unpack! Let’s jump into the top stories you need to know about, carefully organized by theme so you can get a full view of the AI landscape.

1. AI in Cybercrime: Caught Red-Handed

• $10M Streaming-Royalty Heist: In an intriguing twist, the FBI has shut down an elaborate AI-powered scheme that saw a musician raking in over $10 million in streaming royalties fraudulently. By leveraging AI and bots to boost the streaming numbers of phony music tracks, the scam diverted revenue from legitimate artists. The scam’s complexity highlights just how AI can be weaponized, not just in hacking or security breaches, but in gaming digital systems designed for creators and influencers. Read more here.

2. AI and Creativity: Generation For/Against Сreativity

• Put Levy on Smartphones to Help Creatives Survive AI Threat: Some of the UK’s top creatives are pushing back against AI’s impact on the arts. They’ve proposed a smartphone levy to help safeguard livelihoods that are being eroded by generative AI tools. The idea is simple: charge a fee on devices that can generate AI art or music, and use that money to support artists whose work is being undermined. With AI rapidly encroaching on human creativity, these artists are taking a stand for their future in a digitized world. Read more here.

• Generative AI Backlash at Writing Event: Writers and literary figures are increasingly at odds with AI-generated content. This tension came to a head at an annual writing event, prompting high-profile resignations from those concerned that AI threatens to dilute the authenticity of human storytelling. The pushback reflects broader fears across creative industries, where the rise of generative AI is seen not just as an innovation but also as a potential existential threat. Read more here.

3. AI in Media: New Horizons and Dilemmas

• Audible’s AI Voice Cloning: Amazon’s Audible platform is giving narrators the ability to clone their voices with AI, potentially revolutionizing the audiobook industry. Narrators can generate voice models of themselves, allowing them to record audiobooks without ever stepping into a studio. While this could expand their reach and productivity, it also raises ethical concerns. Will human narrators eventually be replaced by their digital doppelgängers? Read more here.
• YouTube’s AI Detection Tools: YouTube is rolling out new tools to help content creators navigate the growing presence of AI-generated media. These tools aim to empower creators by helping them detect AI-generated content in their uploads and manage it appropriately. As deepfakes and AI-generated videos become more common, YouTube’s initiative is a step toward ensuring transparency and fairness in content creation. Read more here.

4. AI in Consumer Tech: Apple’s Latest Moves

• WatchOS AI Features: Apple’s latest watchOS update brings powerful AI-driven tools to your wrist, including real-time language translation. This AI feature promises to make travel and international communication smoother, all while being seamlessly integrated into Apple’s already robust wearable ecosystem. These advancements signal Apple’s growing reliance on AI to enhance user experience in everyday tech. Read more here.
• iPhone 16’s AI-powered Future: The iPhone 16 is set to ship as a “work in progress,” and AI is at the center of Apple’s plan for continuous improvement. While some of the more advanced AI features won’t be ready at launch, Apple is counting on AI to push the iPhone’s capabilities far beyond what users expect, evolving the device over time through software updates. Read more here.

5. AI Research: Complexities and Breakthroughs

• Teaching AI to Forget: Researchers at MIT have published a study exploring how to teach AI models to forget information. Through randomized fine-tuning, they’ve made headway on a major challenge — how to get AI to “unlearn” data, a process that could have major implications for privacy and data security. Read more here.

• Sony and AI Singapore’s SEA-LION LLM: In a significant collaboration, Sony and AI Singapore are working together to develop SEA-LION, a large language model (LLM) specifically designed for Southeast Asian languages. This project is an example of how AI can be tailored to specific linguistic and cultural contexts, making the technology more accessible and relevant to global populations. Read more here.

6. AI Policy: New Laws and Advocacy

• AI Reporting Requirements: In an effort to keep pace with rapid AI developments, the U.S. government is considering new mandatory reporting requirements for advanced AI developers and cloud service providers. These regulations would force companies to be more transparent about the capabilities and uses of their AI models, potentially improving accountability and ethical standards across the industry. Read more here.

• California AI Bill Gains Support: Some of the biggest names in AI, including employees from OpenAI, Anthropic, and Google, have come out in favor of a new California bill aimed at regulating AI development. This support shows the increasing momentum behind stricter AI governance, as both developers and lawmakers recognize the need for more oversight in an industry poised to reshape every sector. Read more here.

7. Global AI Affairs: Military and Geopolitics

• Blueprint for Military AI: At a recent summit, global officials laid out a draft blueprint for how AI could be used in military operations, though China notably declined to participate. The exclusion of such a significant player raises concerns about the lack of international collaboration in controlling the use of AI in warfare, a potentially destabilizing factor in future global conflicts. Read more here.

8. OpenAI: ChatGPT Model “o1” — 500% More Powerful Than GPT-4o

• OpenAI has launched the new “o1” model, which outperforms GPT-4o by 500% in computational power and operates at a PhD-level in exact sciences. Unlike
• GPT-4o, which struggles with complex tasks, “o1” excels in competitive programming, math olympiads, exact sciences, and even philosophy. This is a major leap toward AGI, as “o1” demonstrates human-like reasoning rather than simply pulling answers from a dataset.

Outro

This week’s news cycle has brought a mix of groundbreaking innovations, serious ethical debates, and significant strides toward regulating AI’s vast potential. As the boundaries between human creativity, AI, and regulation blur, it’s clear that this technology will continue to challenge and shape society in unpredictable ways. From policy discussions to revolutionary advancements in consumer tech, it’s an exciting — and sometimes contentious — time to be part of the AI world.

Stay tuned for more insights, and don’t hesitate to share your thoughts with the GT Protocol community. Together, we’ll navigate the ever-evolving AI frontier!