Week in Review (12/21/24)

If you’re a fan of NotebookLM, you may want to check out HeyGen! Below is a two-person HeyGen podcast featuring this week's AI news. While the AI hosts may feel a bit robotic, the possibilities are truly exciting.

The past few weeks have been very busy for OpenAI. Here's a summary of what was released over the 12 days of OpenAI:

Day 1: ChatGPT o1 along with a $200/month ChatGPT Pro subscription

Day 2: AI Reinforcement Fine-Tuning

Day 3: Launched Sora

Day 4: Unveiled Canvas

Day 5: Announced ChatGPT integration with Apple Intelligence

Day 6: Advanced Voice Mode gained visual capabilities

Day 7: Introduced Projects

Day 8: Expanded ChatGPT search to all, enhanced Voice Mode, and improved mobile

Day 9: Released new developer tools and model

Day 10: Made ChatGPT accessible via WhatsApp

Day 11: Updated the Mac ChatGPT app

Day 12: Previewed ChatGPT o3 and o3 mini models

Wishing you all a Happy Holidays! Enjoy this week's news.

Federal

• HHS reports 66% increase in AI uses as agencies release inventories HHS reports 271 AI use cases, a 66% increase, as federal agencies release expanded 2024 inventories, prioritizing risk management for safety- and rights-impacting applications per new OMB guidelines.
• DHS launches internal GenAI chatbot to leverage non-public data DHS launched DHSChat, an internal GenAI tool for 19,000 employees to draft reports, streamline tasks, and securely leverage non-public data, ensuring privacy and civil liberties compliance.
• Federal government discloses more than 1,700 AI use cases The U.S. federal government disclosed 1,757 AI use cases in 2024, over double last year's figure. Of these, 227 are rights- or safety-impacting, reflecting enhanced transparency and safety-focused governance.
• Treasury report on the uses, opportunities, and risks of AI in financial services The Treasury's report emphasizes AI's transformative potential in finance, recommending stronger collaboration, regulatory updates, and compliance measures to balance innovation with risks like bias, privacy, and cybersecurity.
• USDA release AI Strategy The USDA released its inaugural AI strategy, emphasizing ethical, transparent adoption of AI to boost efficiency, data-driven decision-making, and support for American agriculture, rural communities, and sustainability goals.
• House bipartisan task force on AI delivers report The bipartisan House AI Task Force unveiled a comprehensive report recommending strategies to maintain U.S. AI leadership while implementing safeguards against risks, emphasizing innovation, security, and consumer protection.
• CDAO, the Pentagon’s AI-accelerating office, undergoing restructuring before presidential transition The Pentagon's CDAO is undergoing restructuring to accelerate AI adoption for military needs. Key changes include the creation of an AI Rapid Capabilities Office and scaling proven technologies for broader use.
• AI and its potential effects on the economy and the federal budget The Congressional Budget Office report explores how AI adoption could influence economic growth, employment, and income distribution, with uncertain but significant implications for federal revenues and spending.

State/Local

• California says it’s ready to buy generative AI tools After a six-month trial, California launches procurement for generative AI tools to address housing, employment forecasting, and budgeting challenges, emphasizing responsible and innovative implementation statewide.
• TxDOT launches AI Strategic Plan TxDOT's AI Strategic Plan outlines 230 use cases and a roadmap to enhance road safety, streamline operations, and improve decision-making, emphasizing security, transparency, and workforce innovation.
• How AI is helping North Charleston law enforcement solve cases North Charleston police use AI-enhanced video analytics and Fusus information-sharing to accelerate investigations. The system cuts review time from days to hours, streamlining crime detection and response.
• New Arizona charter school will use AI in place of human teachers Arizona's Unbound Academy will use AI to tailor two-hour academic sessions, freeing up time for life skills development like teamwork and leadership. Adults provide emotional support, not instruction.
• Gov. Mills establishes task force to study potential impact of AI on Maine Gov. Janet Mills established a 21-member AI task force to prepare Maine's economy and workforce for AI’s impacts. Recommendations on safeguarding data, reducing bias, and leveraging AI are expected by October 31.
• Survey: Some Idaho government agencies use AI already A survey found 23% of Idaho state agencies using generative AI, enhancing efficiency but raising legal, ethical, and data privacy concerns. Lawmakers explore potential policies for responsible AI adoption.
• Washington police agencies tap AI to review body cam footage Washington police agencies, including Spokane County, leverage AI to analyze extensive body camera footage. AI and human-assisted tools aim to improve de-escalation tactics, accountability, and operational efficiencies.
• AI streamlines non-emergency dispatch calls in Kalamazoo County Kalamazoo County Dispatch introduces Ava, an AI-powered virtual assistant handling non-emergency calls. Designed for emergency centers, Ava routes calls and collects details, freeing staff for critical tasks.
• Illinois Supreme Court announces policy on AI The Illinois Supreme Court's new AI policy emphasizes ethical AI use in courts, requiring compliance with professional standards. Judges and attorneys must verify AI outputs to uphold justice and mitigate bias.

International

• China wants to dominate in AI — and some of its models are already beating their U.S. rivals Chinese firms excel with open-source large language models (LLMs) like Alibaba’s Qwen, despite chip restrictions. Their global strategy highlights innovation and expanding AI ecosystems.
• Security Council debates use of AI in conflicts The UN Security Council debated AI's dual-use potential, highlighting its transformative capabilities and risks, particularly in conflicts. Member States called for urgent international governance to prevent misuse and ensure equitable benefits.
• Zuckerberg: It’s ‘sad’ that EU is left behind on AI Meta CEO Mark Zuckerberg expressed frustration over EU regulatory hurdles, saying they prevent AI innovation and rollout in the region, leaving Europeans behind on technological advancements.
• China’s AI elite rethink their Silicon Valley dream jobs Tough U.S. visa policies and geopolitical tensions discourage Chinese AI talent from relocating, with many opting for Canada or staying in China, where AI ecosystems are rapidly growing.
• 'World-first' AI camera targets drink-drivers The innovative "Heads-Up" AI camera, tested in Devon and Cornwall, detects impaired driving behavior. Police use roadside tests to confirm, aiming to prevent accidents and save lives.
• An Australian-first AI Plan to boost capability Australia’s National AI Capability Plan aims to boost AI investment, strengthen industry capabilities, and enhance workforce skills, supporting economic growth while ensuring responsible adoption and inclusion.

Other

• Navigating the AI frontier: A primer on the evolution and impact of AI agents This white paper explores AI agents' evolution, benefits, and risks, emphasizing ethical considerations and governance to responsibly enhance efficiency in healthcare, education, and finance.
• Ex-Google CEO warns there's a time to consider "unplugging" AI systems Eric Schmidt warns of AI systems' self-improvement risks, urging safeguards like AI policing AI, as rapid development pushes boundaries and U.S.-China competition intensifies.
• AI agents: Why we should strategize on governance AI agents are advancing rapidly, promising personalized and business applications but posing governance challenges like accountability and ethical risks. Transparent frameworks, like Agent IDs, are vital for responsible deployment.
• United Airlines sets its flight plan for gen AI success United Airlines deploys generative AI models like "Every Flight Has a Story," boosting customer satisfaction by 6% through clear, empathetic updates on flight delays.
• ROI of AI IBM’s global survey finds 85% of IT leaders advancing AI strategies, with 47% seeing positive ROI. Open-source AI solutions boost ROI potential, with 62% planning increased AI investments in 2025.
• Move over, Siri and Alexa, AI agents are the digital assistants we've wanted Next-gen "agentic AI" promises smarter, proactive digital assistants, tailoring solutions from personal data and anticipating user needs. Tech leaders envision AI transforming smartphones, cars, and augmented reality interactions.
• How Liberty Mutual plugged generative AI into everyday work Liberty Mutual deployed LibertyGPT, an internal version of ChatGPT, to enhance workflows across teams. With robust governance and training, 25% of employees save 1.5 hours weekly using generative AI.
• AI and the future of workforce training AI's disruptive impact on jobs highlights the need for retraining and upskilling. Community colleges, alternative career paths, and AI-enabled learning can prepare workers while addressing skill obsolescence and equity challenges.