The Artificial Intelligence Podcast

The US is harnessing artificial intelligence to transform modern warfare, with companies like Palantir and Govini leading the way in producing high-tech weapon systems. Recently, they showcased their cutting-edge technology to top military brass, demonstrating AI's ability to process vast amounts of data quickly and efficiently. AI can help with targeting and logistics, and even deter wrongdoing and prevent war if the US remains at the forefront of development. The Pentagon is seeing benefits, and Capitol Hill is urging Congress to spend $32 billion on AI research and development to ensure national security.

A recent survey has found that 78% of Americans are concerned that artificial intelligence will be used to manipulate or distort voters' perceptions of candidates in the 2024 election. They worry that AI will create fake social media accounts, videos, and targeted content to discourage certain demographics from voting. Most respondents believe AI will be used to generate fake accounts and drive conversations on social media, and only 5% think AI will have a positive effect on political discourse. Despite government efforts to address concerns, Americans are pessimistic about AI's influence on the election.

NASA has appointed its first chief AI officer, David Salvagnini, to develop and coordinate the agency's AI programs. With a 21-year career in the US Air Force and experience in engineering programs, Salvagnini is well-equipped to lead NASA's AI endeavors. The agency's use of AI dates back decades, but with rapid advancements, a dedicated leader is necessary to ensure responsible use and development. Salvagnini will develop AI programs with academia, industry partners, and government agencies, and ensure AI is used safely and securely, critical in the face of growing concerns about AI security.

The energy consumption of American AI data centers is expected to rival the output of new US solar farms, driven by the significant energy required to operate generative AI like Chat GPT. Estimating AI energy usage is complex, but Goldman Sachs' analysis of Virginia's commercial power consumption suggests a 37% increase from 2016 to 2023, likely due to data center growth. AI data centers currently account for 0.5% of US power demand, but this is expected to triple by 2030, driven by increasing AI use, electric vehicles, and industrial electrification, raising concerns about the environmental impact.

Google's demonstration of its call-scanning AI technology has sparked concerns among privacy and security experts, who warn that it could pave the way for centralized censorship and surveillance. The technology, powered by Gemini Nano, scans voice calls in real-time for conversational patterns associated with financial scams, but experts say it could be repurposed for social surveillance, monitoring, and blocking of certain activities. They caution that this could lead to a dystopian future of censorship by default, threatening basic values and freedoms.

Researchers at Delft University of Technology have developed a drone that uses neuromorphic image processing and control, inspired by animal brains, to fly autonomously. This approach enables the drone to process data up to 64 times faster while consuming three times less energy compared to traditional graphic processing units. The technology has the potential to enable tiny, agile drones for various applications, such as crop monitoring and warehouse management.

Colorado and Connecticut are pioneering efforts to regulate artificial intelligence, introducing groundbreaking bills to prevent AI-driven discrimination in crucial services like healthcare, employment, and housing. These bills come as over 40 states consider more than 400 AI-related bills, highlighting the challenges of governing this powerful technology. While Connecticut's effort stalled due to the governor's concerns about stifling innovation, Colorado's bill faces criticism from both consumer advocates and tech groups. As states navigate AI regulation, the need for safeguards against biased decision tools becomes increasingly clear.

The US and China held their first-ever dialogue on artificial intelligence in Geneva, discussing the risks and management of AI. US officials expressed concerns about China's misuse of AI, while China rebuked the US over restrictions and pressure on the technology. Both sides recognized the opportunities and risks posed by AI, including its potential for digital surveillance and disinformation. The talks aim to keep communication open on AI risk and safety, a key aspect of managing competition between the two nations. The success of the talks will be measured by whether they continue in the future.

A California-based startup, Atomic Canyon, is utilizing the world's fastest supercomputer at Oak Ridge National Laboratory to develop an AI tool that could revolutionize the nuclear power sector. The AI tool aims to efficiently assist nuclear companies in applying for construction and operating licenses, saving millions of dollars and countless hours. With access to the Frontier supercomputer, Atomic Canyon's technology will search through millions of documents from the Nuclear Regulatory Commission, providing valuable insights and data to nuclear engineers and operators. The long-term goal is to make nuclear power a cleaner and more viable source of energy.