A TEORIA DA INTERNET MORTA (Apocalipse).

The Silent Apocalypse of the Internet

The Shift in Internet Traffic

• The speaker reflects on the potential reality that we may be the last generation to remember a "real" internet, emphasizing a deeper concern beyond mere nostalgia.

• In 2024, it was reported that over half (51%) of internet traffic is generated by bots rather than humans, marking a significant shift in online interactions.

• While discussions about the state of the internet continue, there are alarming implications for our planet and democracy as AI influences our online experiences.

User Engagement with Technology

• A survey reveals varying screen time among users: 14% spend less than 2 hours daily on their phones, while others report up to 9 hours or more.

• On average, Brazilians spend 9:32 hours per day looking at screens—56.6% of their waking life—making Brazil one of the highest countries for screen time globally.

• The issue isn't just usage but what content is consumed; much of it is now controlled by algorithms rather than human input.

The Theory of a 'Dead' Internet

• The concept of a "dead" internet has circulated since 2010, suggesting that robots have taken over content creation and consumption from humans.

• By 2024, reports indicated that automated traffic had surpassed human traffic for the first time in history (51% vs. 49% previously).

Bots and Social Media Manipulation

• A study found that approximately 64% of accounts on X (formerly Twitter) are bots, raising concerns about authenticity in online discussions.

• Sam Altman, CEO of OpenAI, acknowledged this phenomenon on social media, indicating widespread recognition that AI influences user interactions significantly.

Real-world Implications and Examples

• An example from Facebook shows how an AI-generated meme ("Shrimp Jesus") gained massive engagement through bot activity without any real human interaction involved.

• Numerous news sites rely entirely on AI-generated content without human oversight; some articles contain generic phrases typical of language models like ChatGPT.

• Predictions suggest that by 2030, up to 99% of all online content could be created by artificial intelligence—a notion once considered extreme but increasingly plausible today.

The Impact of AI on Energy Consumption and Water Usage

The Decline of the Internet and Rise of AI

• Graft, an analysis company, predicts that by the end of 2024, AI-generated articles will surpass human-written text. This indicates a gradual decline in traditional internet usage as people spend excessive time consuming digital content.

• The speaker encourages viewers to subscribe to their channel, highlighting a milestone goal of reaching 100,000 subscribers.

Real Costs of Digital Consumption

• The notion that the internet is merely virtual and harmless is challenged; there are significant physical costs associated with its consumption that impact the planet.

• According to the International Energy Agency, energy consumption by AI data centers is projected to more than double between 2024 and 2030, reaching 945 TWh annually—equivalent to Japan's total annual energy consumption.

Environmental Consequences

• In the U.S., data centers currently consume about 4.4% of national electricity, with projections suggesting this could rise to 12% by 2028 and potentially reach 20% by 2030.

• A Goldman Sachs report states that approximately 60% of increased energy demand from data centers will be met through fossil fuels, resulting in an additional 220 million tons of CO2 emissions.

Water Usage Concerns

• Data centers generate substantial heat due to high electrical demands; they require significant water for cooling processes.

• Estimates suggest that by 2025, AI's water footprint could range from 312 billion to 764 billion liters annually—comparable to global bottled water consumption.

Scale and Future Projections

• OpenAI's CEO mentioned that a single query on ChatGPT uses only a small amount of water (0.0085 gallons), but with around 300 million active users weekly generating about one billion messages daily, this totals over 31 million gallons per year just for ChatGPT alone.

• OpenAI plans to build massive data centers in Texas (Project Stargate), each costing $50 billion and capable of powering approximately 750,000 homes, raising concerns about resource allocation in arid regions.

Economic Implications for Consumers

• A Dominion Energy report forecasts residential electricity bills in the U.S. may more than double by 2039 due to growing data center demands.

• Brazil is also investing heavily in large-scale data center projects; companies are constructing facilities with capacities equivalent to entire cities' energy needs.

Regional Developments in Brazil

• Brazilian projects include a new data center in Maringá with a capacity equal to cities like Florianópolis or Santos; another campus specializing in AI aims for up to 4.750 MW based on demand.

• Despite Brazil’s clean energy sources (hydroelectricity, wind, solar), modern data centers still rely heavily on water for cooling systems.

Transparency Issues

• Many companies operate through subsidiaries which obscure their actual energy and water consumption under claims of industrial secrecy; public agencies struggle for transparency regarding these metrics.

This structured overview captures key insights from the transcript while providing timestamps for easy reference back into specific parts of the discussion.

Impact of AI on Elections and Society

The Environmental Cost of AI

• The consumption of water for energy production related to AI usage is projected to reach 30.4 billion liters, leading to increased energy costs and environmental degradation.

AI's Role in Upcoming Elections

• In the context of the 2026 elections, over 155 million Brazilians will vote under new regulations prohibiting deep fakes in campaigns, raising concerns about misinformation.

Understanding Deep Fakes

• Deep fakes are hyper-realistic audio and video manipulations created by AI, allowing individuals to fabricate statements or actions attributed to public figures without their consent.

Misinformation Spread

• Misinformation can spread rapidly through unofficial channels; for instance, a fabricated video of a politician can gain traction before its authenticity is verified.

Historical Context of Data Manipulation

• Past incidents like the Cambridge Analytica scandal illustrate how data was exploited for targeted political advertising, manipulating voter emotions based on psychological profiles.

Accessibility of Manipulative Tools

• Unlike previous election cycles that required extensive resources for data collection, current technology allows even small campaign teams to create sophisticated misinformation strategies with minimal investment.

Predictions for Future Challenges

• Experts warn that the scale and sophistication of AI-generated misinformation in future elections could be unprecedented, complicating voters' ability to discern truth from fabrication.

The Rise of Online Scams

Emergence of New Influencers

• Following the launch of ChatGPT in 2022, many self-proclaimed AI experts emerged online, often promoting dubious schemes that exploit consumer trust.

Financial Impact on Consumers

• In 2024 alone, consumers lost over $1 trillion due to online scams; predictions indicate that AI-generated scams will dominate in 2026 as technology becomes more accessible.

Criminal Campaign Interventions

• Authorities have already disrupted numerous criminal campaigns leveraging advanced technologies; however, ongoing challenges remain as scammers adapt quickly.

AI and Content Creation Ethics

The Case of Virtual Influencers

• An example is Itana Lopes—a fully AI-generated influencer—who represents a growing trend where digital personas are created without real-world counterparts or ethical considerations regarding their impact on society.

Exploitation Through Adult Content

• There has been an alarming rise in adult content generated by AI using images resembling minors or non-existent individuals, highlighting severe ethical issues surrounding consent and exploitation.

Need for Regulation

• There is an urgent call for platforms to implement stricter monitoring systems to identify fake profiles and prevent harmful content distribution effectively.

The Controversial Use of AI in Image Generation

Issues with Grock's Image Generation Capabilities

• The use of Grock, an AI developed by Elon Musk's company, raises ethical concerns as it allows users to digitally undress individuals based on their social media photos, despite being theoretically prohibited.

• Users can comment on posts to instruct Grock to generate inappropriate images without needing the original photo, affecting people of all ages and genders.

• There have been significant security failures acknowledged by Grock that enabled the creation of sexualized images of minors, prompting legal actions from countries like France and India.

• A coalition warned Musk’s company about these vulnerabilities back in August 2025, describing Grock as a "nudification machine," but their concerns were ignored.

• When approached for comments regarding these issues, the company's response was dismissive, labeling traditional media as liars while Musk publicly mocked the situation.

Consequences and Recommendations for Media Consumption

• Following backlash, the generation feature was temporarily disabled; however, the damage had already been done. The technology is now widely accessible and misused.

• Viewers are advised to verify information from official accounts before believing or sharing content—especially during elections—to combat misinformation effectively.

• Skepticism is encouraged when encountering sensational claims; questioning the credibility of outrageous statements can help identify fake news.

• Emphasis is placed on supporting genuine content creators rather than anonymous profiles that contribute to misleading narratives online.

The Dual Nature of AI Technology

• While acknowledging potential dangers posed by AI like Grock, there is recognition that such technologies can also enhance productivity and efficiency in various fields when used responsibly.