Google YouTube AI Training Clash, Meta Midjourney Partnership

Google is facing backlash from YouTube creators who are concerned about the use of their videos to train AI models like Veo. Creators worry about potential revenue loss and the replication of their styles without compensation, even though YouTube says it paid out over $70 billion to creators between 2021 and 2023. Meanwhile, the rise of AI-generated 'slop' – low-quality, often repetitive content created using tools like ChatGPT and Midjourney – is flooding the internet, attracting views and ad revenue despite criticism. Meta is partnering with Midjourney to enhance AI content on platforms like Facebook and Instagram, while also dealing with employee departures from its AI division, Meta Superintelligence Labs (MSL), despite offering substantial bonuses. Nvidia's success remains closely tied to the AI boom, with its chips being crucial for AI development, though some analysts question whether the AI market is overheating. Concerns persist about a potential AI bubble and the environmental impact of AI training, which consumes significant resources. Microsoft's analysis of Copilot conversations indicates that jobs involving knowledge and communication are most likely to be affected by AI. WhatsApp is introducing AI writing tools with a focus on data protection, though audits have revealed potential vulnerabilities. Agentic AI, which can make independent decisions, holds economic promise, particularly in Southeast Asia, but requires careful data management. The Google Pixel 10 Pro utilizes AI to anticipate user needs, raising privacy concerns due to the extensive data it collects. Securing AI systems is also becoming increasingly important, with companies like VMware by Broadcom offering tools to protect against AI-related cybersecurity risks.

Key Takeaways

• Google faces creator pushback over using YouTube videos for AI training, with concerns about compensation and control.
• AI 'slop,' low-quality AI-generated content from tools like Midjourney and ChatGPT, is prevalent online, earning revenue despite criticism.
• Meta is partnering with Midjourney to improve AI content on Facebook and Instagram, while also experiencing departures from its AI division.
• Nvidia's success is linked to the AI boom, but concerns exist about a potential AI bubble.
• Microsoft's study reveals knowledge and communication-based jobs are most impacted by AI.
• WhatsApp's AI writing tool emphasizes data protection, but audits found vulnerabilities.
• Agentic AI offers economic potential, especially in Southeast Asia, but requires careful data management.
• The Google Pixel 10 Pro uses AI to anticipate user needs, raising privacy concerns due to data collection.
• AI training has a significant carbon footprint, demanding greener solutions.
• AI cybersecurity risks are increasing, requiring robust compliance and resilience strategies.

YouTube creators clash with Google over AI training on their videos

Google is using YouTube videos to train its AI tools like Veo, upsetting many creators. These creators worry AI could replicate their styles and cut into their earnings, and they want compensation or an opt-out option. YouTube says it paid over $70 billion to creators from 2021 to 2023 and is developing tools to manage AI-generated content. Some creators feel their work is being used without return, while YouTube emphasizes its commitment to creator partnership and innovation.

AI 'slop' floods the internet: What it is and why it matters

AI 'slop' is low-quality content made easily with AI tools like ChatGPT and Midjourney. It's like spam, overwhelming social media and making it hard to know what's real. This slop targets algorithms to go viral, and some people make money from it. However, it can hurt real businesses and spread misinformation, making it harder for users to find reliable information. Experts suggest users should be aware and platforms need to address the issue.

AI 'slop' videos annoy, but earn big views and ad money

AI-generated videos, called 'AI slop,' are becoming common online, with some being odd or misleading. These videos are quickly made using AI and are often repetitive. Despite criticism, they attract many viewers and generate ad revenue. One creator, Mark Lawrence I Garilao, makes money from these videos on YouTube. Platforms like TikTok and YouTube are starting to address AI content, but aren't banning it completely.

AI videos take over the internet, making money despite criticism

AI-generated videos, known as 'AI slop,' are becoming popular on social media. Critics find them annoying, but creators use them to attract viewers and earn ad money. These videos are made quickly with AI tools, like those from Midjourney. Meta is partnering with Midjourney to improve AI content on Facebook and Instagram. Despite concerns, AI videos are profitable, with Midjourney earning millions through subscriptions.

Nvidia's AI success: Is the tech bubble safe for now?

Nvidia's success is closely tied to the growth of AI. The company's value has greatly increased due to its graphic processing chips used in AI technology. Despite high expectations, Nvidia's recent earnings have raised questions about the AI market. However, major tech companies are investing heavily in AI hardware, suggesting Nvidia still has room to grow. As long as AI demand stays high, Nvidia's tech bubble seems safe.

Is AI too big to fail like big tech firms?

Economists worry the AI bubble could burst, causing a crash worse than the dot com era. Companies are investing billions in AI, but profits haven't appeared yet. Despite this, some experts believe leading AI companies are too important to fail because they're tied to the economy and government. These companies are lobbying to gain support, and the government views AI as crucial for the U.S.

Meta's AI division sees employee departures after bonus offers

Meta, the parent company of Facebook, is experiencing employee departures from its AI division, Meta Superintelligence Labs (MSL). This comes shortly after offering $100 million bonuses to attract top AI talent. Eight people, including researchers and engineers, have left MSL, which was launched two months ago. A Meta spokesperson attributed the departures to normal employee turnover.

WhatsApp's AI writing help promises data protection, but audits found issues

WhatsApp is launching an AI writing tool with 'Private Processing' to protect user data. This system aims to keep messages private using techniques like anonymized requests and isolated processing. However, security audits found vulnerabilities that could have allowed de-anonymization and code injection. Meta fixed most issues before launch, but trust in Meta and its partners is still needed for the system to be secure.

Agentic AI: Promise, doubts, and its impact on Southeast Asia

Agentic AI, which can make decisions and act independently, could unlock billions in economic value. However, only a few companies have fully adopted it, and trust is declining. Experts say agentic AI works best when humans are involved. It can automate tasks like IT operations, but data and infrastructure must be properly managed. Southeast Asian companies should focus on data quality to benefit from agentic AI.

AI cheating panic misses the point, says expert

This article discusses the concern that Gen Z is using AI, like ChatGPT, to cheat in school. It is an opinion piece by Rachel Janfaza, the founder of Up and Up Strategies, a research and consulting firm focused on reaching Gen Z.

Microsoft reveals jobs most at risk from AI

Microsoft studied how AI impacts jobs by analyzing anonymous Copilot conversations. The study found that roles involving knowledge and communication have the highest AI overlap. Jobs like technical writers, editors, and customer service representatives are most affected. Physical labor and machine operation jobs have the least overlap. The study doesn't predict job losses but shows where AI can assist with tasks.

AI training's carbon footprint demands greener solutions

AI training consumes a lot of electricity, water, and carbon. Training OpenAI's GPT-3 used a large amount of power and water, creating tons of CO2. As AI models like GPT-4 and GPT-5 grow, these costs increase. Each AI query also adds to global emissions. Using renewable energy, auditing environmental impacts, and being transparent can help make AI more sustainable.

AI cybersecurity risks increase compliance difficulties

Cybersecurity in the age of AI requires making compliance and resilience part of AI strategy. VMware by Broadcom offers tools to secure AI models. Protecting AI is complex due to data distribution and changing regulations. Continuous compliance checks and strong data recovery are essential. AI can be used by bad actors, so VMware provides tools to protect against new and old threats.

Google Pixel 10 Pro review: AI phone saves time, but needs your data

The Google Pixel 10 Pro is an AI phone that uses your data to anticipate your needs. It accesses contacts, location, messages, and email to help save time. The phone can automatically share flight information and provide directions. However, using these features requires sharing a lot of personal data with Google. While helpful, this data sharing raises privacy concerns.

Sources

• Google is training its AI tools on YouTube videos. These creators aren't happy
• AI slop took over the internet. Here's what it is and why you should care
• 'AI slop' videos may be annoying, but they're racking up views
• AI-generated videos flood the internet, raking in ad revenue despite criticism.
• Nvidia: Tech bubble seems safe so long as AI demand remains high
• Is AI Already Too Big to Fail?
• Meta AI employees reportedly already jumping ship
• Writing Help: WhatsApp's AI writing help promises maximum data protection
• Agentic AI: Promise, scepticism, and its meaning for Southeast Asia
• Opinion | The AI cheating panic is missing the point
• Jobs that are most at risk from AI, according to Microsoft
• AI Training Fuels Massive Carbon Costs And Demands Greener Solutions
• AI poses new difficulties to compliance, cybersecurity
• Google Pixel 10 Pro Review: This A.I. Phone Can Save Time if You Surrender Your Data