OpenAI CEO Sam Altman spoke at Web Summit in Lisbon, Portugal, acknowledging that people are right to fear rapidly advancing AI. He argued that industry leaders should be trusted to develop the technology safely and called for collaboration among leaders to guide AI for society's benefit. Altman stressed that AI offers significant benefits alongside serious risks.
Researchers Edward Hughes and Louis Kirsch left Google to build AI that can build better AI. Their London start-up Inherent uses a prototype called Faraday that learns from all company data. This pursuit, known as recursive self-improvement, is pursued by other start-ups as well, and some experts fear AI systems could one day improve themselves with little human help.
OpenArt founder Mao says the film and TV industry is slowly adopting AI production tools, primarily in pre-production and post-production. The company's Director tool is designed for previz and postproduction rather than replacing live on-set production. Mao compares current resistance to past opposition against VFX technology that is now widely accepted.
A Glassdoor survey found 61 percent of companies now use AI in hiring. Experts say AI can speed up the process but cannot replace human judgment, and researchers warn that AI systems can carry over biases from their training data. Meanwhile, Morgan State University launched Maryland's first public AI bachelor's degree, building 18 AI courses into a converted cloud computing program.
Key Takeaways
- Sam Altman says people are right to fear AI but industry leaders should be trusted to develop it safely
- Google alumni Edward Hughes and Louis Kirsch are building AI that can improve itself through their start-up Inherent
- 61 percent of companies use AI in hiring according to a Glassdoor survey
- OpenArt founder Mao says generative AI will not replace live production but will enhance it
- Morgan State University launched Maryland's first public AI bachelor's degree with 18 AI courses
- Radford professor Chapman Rackaway says AI fears have created a cross-partisan coalition among Democrats and Republicans
- Israel's national AI plan names physical AI as a strategic priority and calls for testing in signal environments
- The Globee Awards are accepting nominations for AI security achievements globally
- Security podcast Below the Surface episode 82 covered firmware risks and the growing role of AI in security research
- Experts urge companies to use AI hiring tools alongside human decision makers to avoid bias
Sam Altman says world should trust AI industry despite risks
OpenAI CEO Sam Altman spoke at Web Summit in Lisbon, Portugal. He said people are right to fear rapidly advancing AI. He also said industry leaders should be trusted to develop AI safely. Altman stressed that AI offers big benefits but also serious risks. He called for collaboration among leaders to guide AI for society's good.
Altman urges world to trust AI firms amid growing fears
OpenAI CEO Sam Altman said the world is right to be afraid of fast-moving artificial intelligence. He argued that AI companies should be trusted to develop the technology safely. His remarks add to a growing list of warnings from industry leaders about AI risks. He stressed the need for caution in AI development.
The AI self-improvement scenario that terrifies experts
Researchers Edward Hughes and Louis Kirsch left Google to build AI that can build better AI. Their London start-up Inherent uses a prototype called Faraday that learns from all company data. This pursuit is called recursive self-improvement, or RSI. Other start-ups are also chasing this ambitious and risky goal. Some experts fear AI systems could one day improve themselves with little human help.
AI interviewers are here: what job seekers need to know
Many companies now use AI in hiring, including AI interviewers. A Glassdoor survey found 61 percent of companies use AI in hiring. Experts say AI can speed up hiring but cannot replace human judgment. Researchers warn AI systems can also carry over biases from their training data. Experts urge companies to use AI as a tool alongside human decision makers.
Security podcast covers firmware risks and Linux malware
Below the Surface episode 82 was recorded on September 10, 2026. Host Paul Asadoorian discussed security topics with Vlad Babkin and Chase Snyder. They covered Secure Boot bypass risks, vendor trust, firmware transparency, and Linux malware. The group also debated attacks on network devices and the growing role of AI in security research. They stressed the need to independently verify firmware and software.
Globee Awards accept AI security achievement nominations worldwide
The 23rd Annual Globee Awards for Cybersecurity are now accepting nominations for AI security achievements globally. The awards recognize achievements involving AI-powered threat detection, security operations, fraud detection, identity security, and other areas. Both cybersecurity providers and organizations using AI security technologies are invited to participate. Entries are evaluated through a merit-based, data-driven process with scoring criteria and professionals from multiple industries worldwide.
OpenArt founder says generative AI will not replace live production
OpenArt founder Mao says the film and TV industry is the slowest to adopt AI production tools but is starting to use them in pre-production and post-production. The company's Director tool is meant for previz and postproduction purposes rather than replacing live on-set production. Mao believes the future is a combination of live production and AI enhancement. She compares the current resistance to past opposition against VFX technology that is now widely accepted.
Morgan State University launches Maryland's first public AI degree
Morgan State University launched Maryland's first public AI bachelor's degree by converting an existing cloud computing program into a full AI curriculum. The university built 18 AI courses and embedded 8 completed projects into the curriculum, enabling regulatory approval from the Maryland Higher Education Commission in days. Students learn classical programming and data structures before progressing through problem-based AI courses. The program emphasizes reasoning grounded in reinforcement learning and prepares graduates to design and extend AI models rather than simply use them.
Radford professor discusses AI fears and bipartisan politics
Chapman Rackaway, a political science professor at Radford University, discusses how politicians are struggling to decide where they stand on AI regulation. He notes that fears about AI corrupting autonomous drones have created a cross-partisan coalition of Democrats and Republicans. Rackaway says constituents are still figuring out their own opinions on AI, making it hard for politicians to take a firm stance. Research shows people like the idea of a killswitch that could turn AI off if it gets out of control.
Israel's AI plan highlights need to simulate invisible signal environments
Israel's new national AI plan names physical AI as a strategic priority, focusing on machines that perceive, decide, and act in the real world. The plan calls for sandboxes and a national physical testbed to validate autonomous systems before they are trusted. An article argues that the electromagnetic environment, including radio waves and signals, should also be part of that testbed. The author points out that autonomous machines depend on communication networks, and those networks can be disrupted by jamming or other interference, making testing of both physical and signal environments essential.
Sources
- 2.9K views · 25 reactions | OpenAI chief executive Sam Altman said...
- OpenAI’s Altman says world should ‘trust’ AI firms amid mounting fears
- The Liftoff Scenario That Terrifies A.I. Doomsayers
- Your AI interviewer will see you now: How job seekers should navigate the 'jarring' new hiring technology
- BTS #82 - Firmware Analysis, Linux Malware, Future of AI
- The 23rd Annual Globee® Awards for Cybersecurity Invite Artificial Intelligence Security Achievement Nominations Worldwide
- Generative AI Tools Won’t Replace Live Production OpenArt Founder
- Morgan State won Maryland's first public AI degree in months by building it first
- Radford political science professor Chapman Rackaway talks AI fears, bipartisanship
- AI can think, but can it communicate
Comments
Please log in to post a comment.