The clock is ticking on AI safety

In partnership with

Most people will look back at these moments in history and wish they’d paid closer attention. You didn’t scroll past it. 😤

Build real AI and tech skills, faster

Udacity helps you build the AI and tech skills employers actually need—fast. Learn from industry experts through hands-on projects designed to mirror real-world work, not just theory.

Whether you’re advancing in your current role or preparing for what’s next, Udacity’s flexible, fully online courses let you learn on your schedule and apply new skills immediately. From AI and machine learning to data, programming, and cloud technologies, you’ll gain practical experience you can show, not just list on a résumé.

Build confidence, stay competitive, and move your career forward with AI and tech skills that are in demand.

Learn More

📣 Our AI Community, called the AI Business Boosters, is officially launching for everyone TOMORROW! Early access folks are already joining... Each day counting down to launch, we have been revealing one feature of the community. Today’s feature:

AI SAFETY RESEARCHER SAYS WE ARE RUNNING OUT OF TIME 😳

What’s up? A leading AI safety researcher has warned that the world may be running out of time to prepare for the risks posed by rapidly advancing AI.

Want the details? David Dalrymple, an AI safety expert at the UK’s Advanced Research and Invention Agency (Aria), says modern AI systems are improving so quickly that safety research and regulation may struggle to keep up. According to recent testing by the UK’s AI Security Institute, top AI models are now able to complete many complex, human-level tasks faster and more cheaply than people, with performance in some areas doubling roughly every eight months. Some advanced systems can even carry out long, multi-step tasks on their own and have shown early signs of being able to copy themselves—an important safety concern. Many ethicists are working on ways to limit risks, especially in critical areas like energy and infrastructure. However, if AI capability races ahead of safety, it could destabilize economies, security, and decision-making. The concern isn’t science fiction, but speed: powerful systems may arrive before society is ready to control them.

Learn AI in 5 minutes a day

What’s the secret to staying ahead of the curve in the world of AI? Information. Luckily, you can join 1,000,000+ early adopters reading The Rundown AI — the free newsletter that makes you smarter on AI with just a 5-minute read per day.

Sign up to start learning.

CLAUDE CODE SHOCKS GOOGLE ENGINEER 🤯

What’s up? A senior Google engineer said Claude’s AI coding assistant recreated, in about an hour, work her team had spent nearly a year developing.

Want the details? Jaana Dogan, a principal engineer at Google working on the Gemini API, tested Anthropic’s Claude Code by giving it a short, high-level description of a complex engineering problem. The task involved “distributed agent orchestrators,” systems that coordinate multiple AI agents working together. Dogan intentionally avoided using any internal Google knowledge and kept the prompt to just three paragraphs. Despite this, Claude Code produced a solution strikingly similar to what Google engineers had built after months of experimentation. While the output wasn’t perfect and still needed refinement, Dogan said it captured the core ideas her team had struggled to finalize.

What does this mean? AI is now collapsing years of expensive engineering work into hours. For Google, it’s a wake-up call: competitors’ models can now match internal progress at shocking speed, threatening traditional advantages like scale and resources. It also means that individual engineers and small teams can now build systems that once required tech giants—shifting power, pressure, and expectations across the entire industry.

AI HELPS RESEARCHER TRACK SEALS 🦭

What’s going on? AI is being used to rapidly count and track grey seals at a protected beach in Scotland, cutting data analysis time from hours to seconds.

How does it work? At Newburgh beach in Aberdeenshire, marine biologist and PhD student Claire Stainfield uses drones to film hundreds, sometimes thousands, of seals resting on the sand. Traditionally, she had to manually count seals from this footage, a slow and demanding process. Now, an AI model she trained herself can automatically identify and count seals in drone images almost instantly. The system also records GPS locations, helping her track where seals gather across seasons. This data supports her research into whether increasing tourist numbers affect seal behavior and population size, while keeping disturbance to a minimum.

Why is this significant? This shows how AI is making wildlife research faster, more accurate, and less invasive. By saving time, scientists can study animals more frequently, across more locations and species, improving conservation decisions as human activity and wildlife increasingly share the same spaces.

HAS AI REACHED SINGULARITY? … CHECK OUT THE FRY METER BELOW:

Your feedback on these daily polls helps us keep the newsletter fresh—so keep it coming!