The Artificial Intelligence Podcast

The New Hampshire state House has passed a bill aimed at regulating the use of artificial intelligence (AI) in political ads. The move comes after voters in the state received robocalls with an AI-generated voice pretending to be President Biden. The bill, which passed without debate, would require political ads using deceptive AI to disclose their use of the technology. The motivation behind the bill is concerns about the dangers of AI in political campaigns. The bill is part of a larger trend of states introducing legislation to regulate AI in election-related content. The incident has also prompted calls for federal regulations to safeguard against AI election disinformation.

Emad Mostaque, the founder and CEO of Stability AI, went missing in December 2022, causing concern among employees and investors. Although he reappeared and downplayed the incident, it highlighted deeper problems within the company. Mostaque's relationship with key investors, Coatue and Lightspeed Venture Partners, was crucial for the company's success. However, tensions escalated, ultimately leading to Mostaque's resignation in March 2024. Stability AI had gained attention for its open-source AI model, but issues with company structure, financials, and competition hindered its progress. The future of Stability AI is uncertain, and its downfall serves as a cautionary tale in the AI industry.

Researchers from the University Hospitals of Leicester NHS Trust have developed an artificial intelligence (AI) tool called VA-ResNet-50 that can predict the risk of lethal heart rhythm in individuals. The tool analyses Holter electrocardiograms (ECGs) taken from adults during their daily routines and can identify ventricular arrhythmia (VA), a condition that can lead to sudden death if not treated promptly. The AI tool correctly identified VA in 80% of cases where patients experienced lethal arrhythmias. Current clinical guidelines for assessing the risk of VA are not accurate enough, leading to avoidable deaths. The findings highlight the potential of AI to improve risk assessment and treatment decisions in healthcare.

Visa has announced the launch of three AI-powered risk and fraud prevention solutions, set to be available in the first half of 2024. These solutions will be a part of Visa's Protect portfolio and aim to address the challenges faced by issuers in the digital fraud landscape. The solutions include Visa Deep Authorization, which manages card-not-present payments, Visa Advanced Authorization, which expands fraud risk management tools, and Visa Risk Manager, which focuses on instant payment processes. By utilizing advanced deep learning AI, these solutions will provide real-time risk scores, improving fraud defense mechanisms and reducing operational costs for financial institutions.

Scientists from Stanford Medicine and McMaster University have developed an artificial intelligence model called SyntheMol that has generated structures and chemical recipes for six novel drugs to combat antibiotic-resistant strains of Acinetobacter baumannii bacteria. This bacterium is responsible for many antibiotic resistance-related deaths. The model was trained to construct potential drugs using a library of over 130,000 molecular building blocks and a set of validated chemical reactions. Out of the 58 compounds that were successfully generated and tested in the lab, six of them proved effective against A. baumannii and other antibiotic-resistant bacteria. The researchers plan to further test the compounds for toxicity and collaborate with other research groups for drug discovery in different areas.

AI search engines have made strides in improving their capabilities, but they still have a long way to go to compete with Google. These AI tools, such as Chat GPT, Google Gemini, and Microsoft Copilot, lack a true understanding of what a search engine is and how people use them. While they can find information, they struggle to replicate all the functionalities that Google offers. In tests comparing AI tools to Google, Google consistently provided faster and more accurate results. AI search engines excel in answering buried information queries, but fall short in responding to complex queries and recommending what to watch. Overall, Google's multifaceted functionality and speed make it difficult for AI search engines to truly compete. However, advancements in AI technology still hold potential for the future of search engines.

A recent case involving a Black Uber Eats bike courier, Pa Edrissa Manjang, who faced racial discrimination due to facial recognition checks and received a settlement from Uber. The use of facial recognition checks resulted in Manjang being unable to access the Uber Eats app, causing him to lose job opportunities. The case raises concerns about the lack of transparency and rushed implementation of automated systems, which can lead to biased outcomes. It also highlights the need for stronger enforcement of existing laws and the implementation of dedicated AI safety legislation to address AI bias and protect against discrimination and human rights abuses.

Microsoft has undergone a major shake-up, merging its Windows and Surface divisions under a new leader, Pavan Davuluri, as the company prepares for a focus on "AI PCs." This move follows frustrations from the top indicating that the previous split between the divisions was not successful. Additionally, Mustafa Suleyman, co-founder of DeepMind, has been hired by CEO Satya Nadella to oversee Microsoft's consumer AI push as CEO of Microsoft AI. This hiring suggests that the previous shake-up involving Windows and AI did not have the desired outcomes. The combination of Windows and Surface is expected to bring clarity to Microsoft's AI efforts for Windows.

Grindr, the popular gay dating app, is facing financial challenges and is under pressure to increase revenue. The company plans to monetize the app and introduce new in-app purchases, including the development of an AI chatbot that can engage in sexually explicit conversations with users. Grindr's push for monetization reflects declining stock prices in the dating app industry due to slowed growth. However, some employees are concerned about the impact on user trust and privacy. The company is also facing allegations of inflating the number of paying users. As Grindr considers putting previously free features behind a paywall, there are concerns about the negative impact on the user experience.

Amazon's recent $2.75 billion investment in AI startup Anthropic aims to solidify its position as a leader in cloud computing. However, analysts suggest that the move may not be enough for Amazon to maintain its dominance. While Amazon's cloud computing service, AWS, currently holds a significant market share, competitors like Microsoft Azure and Google Cloud are gaining momentum in the AI space. Analysts have criticized Amazon for its underwhelming efforts in generative AI and slow response compared to Microsoft's Copilot offering. Additionally, Amazon faces challenges in leveraging its infrastructure and accommodating AI-based TPUs. Nonetheless, analysts believe in AWS's leadership and resilience to overcome these challenges. As competition intensifies, Amazon must deliver innovative AI solutions to differentiate itself from rivals.