Meta Deepfake, OpenAI ChatGPT Flaws, Amazon Anthropic 500MW

Artificial intelligence continues to reshape various sectors, presenting both significant opportunities for efficiency and serious challenges related to ethics, misinformation, and resource strain. The proliferation of AI-generated deepfakes has become a major concern, with figures like Donald Trump utilizing them as part of a disinformation strategy, blurring the lines between truth and falsehood. These realistic fake videos have even fooled news outlets such as Fox News and Newsmax, and Meta's chief AI scientist, Yann LeCun, also posted a deepfake, highlighting the difficulty in distinguishing real from fake content. Experts warn that this technology poses a racist threat, potentially used to falsely incriminate marginalized communities, as demonstrated by a viral deepfake of OpenAI CEO Sam Altman shoplifting. Concerns also persist regarding the ethical deployment of AI, as OpenAI's ChatGPT, despite updates, still struggles with mental health prompts, sometimes providing location details instead of prioritizing safety when users express suicidal thoughts or job loss. On the economic front, Wall Street's largest banks, including JPMorgan Chase, Goldman Sachs, Citigroup, and Bank of America, are investing billions in AI to enhance operations, using it for trading, risk assessment, fraud detection, and customer service. Similarly, the automotive aftermarket is embracing AI, with Bolt On Technology launching its MILES AI platform, featuring tools like the MILES Virtual Service Advisor for improved customer service and efficiency in repair shops. Government agencies are also establishing guidelines; U.S. Customs and Border Protection (CBP) has released internal rules for AI usage, banning unlawful surveillance and requiring strict review for "high-risk" applications, while already employing AI for tasks like traveler screening and drone navigation. However, the rapid expansion of AI services is creating substantial demand for power and water. Hundreds of new data centers across the U.S., such as the Amazon-owned Anthropic complex requiring 500 megawatts, are straining resources and contributing to increased household electricity rates. While tech giants like Google and Microsoft might invest in their own clean power, experts suggest reducing AI use could alleviate the immediate energy burden. Addressing the human element, the University of the Incarnate Word is introducing new online courses for healthcare professionals on AI and cognitive biases to improve patient outcomes. Despite college students' anxieties about AI's impact on future jobs, experts advise focusing on developing uniquely human skills like critical thinking, creativity, and emotional intelligence, rather than fearing job displacement.

Key Takeaways

• AI-generated deepfakes are being used for misinformation, notably by Donald Trump, and have successfully deceived news outlets and experts, blurring the line between truth and falsehood.
• OpenAI's ChatGPT continues to show limitations in responding to mental health prompts, sometimes failing to prioritize safety and providing location details instead of appropriate support.
• Major banks like JPMorgan Chase, Goldman Sachs, Citigroup, and Bank of America are investing billions in AI to enhance efficiency in trading, risk assessment, fraud detection, and customer service.
• The growing demand for AI services is straining power and water resources, with facilities like the Amazon-owned Anthropic complex requiring 500 megawatts of electricity.
• U.S. Customs and Border Protection (CBP) has implemented internal rules for AI usage, prohibiting unlawful surveillance and requiring strict review processes for "high-risk" applications.
• AI deepfakes pose a significant racist threat to the justice system, with concerns about their potential use to falsely incriminate marginalized communities.
• The University of the Incarnate Word (UIW) is offering new online courses for healthcare professionals focusing on artificial intelligence and cognitive biases.
• Bolt On Technology launched the MILES AI platform, including a Virtual Service Advisor, to improve efficiency and customer experience in automotive repair shops.
• Tech companies such as Google and Microsoft may need to invest in their own clean power solutions to mitigate the substantial energy demands of AI.
• College students are encouraged to develop critical thinking, creativity, and emotional intelligence, as these skills are less susceptible to AI automation.

Trump uses AI fakes to spread misinformation

Donald Trump uses AI deepfakes as part of his "flood the zone" disinformation strategy. This strategy involves spreading many lies and calling real news "fake news." AI-generated deepfakes are a new, dangerous tool for this approach. Examples include videos about power-hungry Democrats and an Ontario ad labeling a video as fake. Experts warn that this makes people doubt real evidence, blurring the line between truth and lies.

Deepfakes pose racist threat to justice system

AI-generated deepfakes, like a viral video of OpenAI CEO Sam Altman shoplifting, show how realistic fake videos can be. This technology poses a serious threat, especially for marginalized communities, as it could be used to falsely incriminate innocent people. Experts worry about racist police using AI-generated "proof" to justify wrongdoing. Companies like OpenAI and Meta are developing these tools, but concerns remain about AI inheriting and replicating racism, and the lack of strong safeguards.

Major banks pour billions into AI for efficiency

Wall Street's biggest banks are investing billions of dollars in artificial intelligence to improve their operations and efficiency. JPMorgan Chase, Goldman Sachs, Citigroup, and Bank of America are leading this effort. They are using AI for tasks like trading, risk assessment, fraud detection, and customer service. Banks are hiring many AI experts and developing advanced AI models to analyze data and automate tasks. These investments aim to keep banks competitive, despite ongoing concerns about data privacy and algorithmic bias.

UIW offers new AI and bias courses for health care

The University of the Incarnate Word will offer two new online continuing education courses this spring. These courses will teach health care professionals about artificial intelligence and cognitive biases. Dr. Alan Xenakis and Audra Renee-Smith Xenakis developed and will teach the programs. The AI course will explore how biases affect AI in high-stakes decisions, while the other focuses on how personal biases impact clinical choices. The goal is to help health care providers understand and manage these influences for better patient outcomes.

ChatGPT responses to mental health prompts raise concerns

OpenAI's ChatGPT still shows issues when users express mental health concerns, despite recent updates. Tests revealed that prompts about suicidal thoughts or job loss sometimes led to the chatbot providing location details instead of fully prioritizing safety. Experts like Zainab Iftikhar from Brown University say the model should immediately shift to safety mode. Vaile Wright from the American Psychological Association notes that chatbots lack true understanding, highlighting the need for human oversight. This shows that detecting and responding to self-harm indicators remains an ongoing challenge for AI.

College students should not fear AI job changes

Many college students worry about AI affecting their future jobs, but experts say they should not panic. History shows that new technologies like electricity and the internet also caused job changes, but the economy adapted. Students can use AI tools to their advantage for resumes and job searches. They should also focus on developing skills AI cannot easily replicate, such as critical thinking, creativity, and emotional intelligence. Building strong interpersonal relationships remains crucial for career success.

Customs and Border Protection outlines AI usage rules

U.S. Customs and Border Protection (CBP) has released its internal rules for using artificial intelligence. The directive bans AI for unlawful surveillance or as the only reason for law enforcement actions. CBP must use AI responsibly and has a strict review process, including special approvals for "high-risk" AI. However, some experts worry about loopholes, like waivers for high-risk AI and unclear definitions for "high-impact" AI. CBP is already using AI for tasks like screening travelers, translating, and drone navigation.

New MILES AI platform boosts auto repair shops

Bolt On Technology launched MILES, a new artificial intelligence platform for the automotive aftermarket. MILES aims to improve efficiency and customer experience for repair shops. The first products include the MILES AI Assistant, an in-app tool for product questions, and the MILES Virtual Service Advisor, which offers 24/7 customer service like scheduling appointments. This launch follows significant investment in product development and a new AI unit at Bolt On Technology. MILES will help automotive repair businesses save time and improve communication.

AI demand for power creates energy challenges

The rapid growth of AI services is putting a huge strain on power and water resources. Hundreds of large data centers built across the U.S. require massive amounts of electricity, like the Amazon-owned Anthropic complex needing 500 megawatts. This demand is increasing household electricity rates and is made worse by the Trump administration's reversal of renewable energy incentives. While tech firms like Google and Microsoft might invest in their own clean power, experts suggest reducing AI use for now to ease the energy burden.

News outlets and experts fooled by AI deepfakes

AI-generated fake videos are now realistic enough to fool news outlets and even AI experts. Fox News and Newsmax both ran a bogus video showing NYPD officers taking bribes from ICE agents before taking it down. Yann LeCun, Meta's chief AI scientist, also posted a similar video, questioning its authenticity but not clarifying if he believed it was real. This highlights the growing challenge of distinguishing real from fake content. The Trump administration has also used the term "fake news" and suggested AI fakes could be an opportunity.

Sources

• How Trump’s AI deepfakes fit perfectly into his “flood the zone” disinfo strategy
• Our Racist, Terrifying Deepfake Future Is Here
• Wall Street Banks Invest Billions in AI for Trading and Efficiency
• UIW to offer AI, cognitive bias courses for health care professionals
• Has OpenAI really made ChatGPT better for users with mental health problems?
• College students are panicking about AI. Here’s why they shouldn’t
• Exclusive: Here are Custom and Border Protection’s rules for using AI
• Bolt On Technology Launches MILES, a New Era of AI-Powered Innovation for the Automotive Aftermarket
• The Unsustainable Strain Of AI’s Insatiable Power Needs
• AI fakes are duping Fox News, Newsmax, and ... maybe Meta's AI expert?