What's new in AI?

Source: Microsoft Developer on YouTube By Microsoft Developer

Summary

John Maeda explains that AI's genuine progress lies in defining goals and using systems to iteratively improve toward them. This approach, once theoretical, now works due to cheaper and more powerful tokens. Maeda highlights compounding improvement systems as the key to current AI advancements, a concept now practical for implementation.

Why it matters

This video focuses on a core AI development principle: iterative improvement. Maeda's explanation is relevant to all frontier labs building AI, including Google Gemini, Anthropic, and OpenAI. The key change is the practical feasibility of these 'compounding improvement systems' thanks to advancements in token cost and capability. This affects developers and researchers by providing a framework for building more capable AI. Future developments will likely focus on refining these iterative loops for complex tasks.

Read this on Microsoft Developer on YouTube
Opens in a new tab. Subvolts summarizes and links; the full piece belongs to Microsoft Developer on YouTube.
Where the other five stand

Related: OpenAI: OpenAI’s next big AI model has ‘entered the AGI era’ · Google: Google’s latest AI weather model gives you no excuse to forget your umbrella · Anthropic: AGI IS HERE · Meta: Trump Administration Sides With OpenAI in New York Times Copyright Lawsuit · xAI: Sam Altman "AGI by December"

Hype check
2/5Worth a look

Rated low: routine. Worth knowing, not worth rearranging your day for.

Prior coverage our earlier items on the same thing
Published
Source
Microsoft Developer on YouTube (youtube.com)
Author
Microsoft Developer
Company
Microsoft · Official · Videos
People
John Maeda
Summary by
Subvolts, using an AI model (how we work). Spotted a mistake? Tell us.

Questions people ask

What is the core idea behind current AI progress?
Current AI progress centers on defining clear goals and using systems to iteratively improve toward them. This 'hill climbing' approach is now feasible due to cheaper and more powerful AI tokens.
Why is this approach possible now?
The approach is now possible because tokens have become cheaper and more powerful, enabling complex iterative processes that were not practical a decade ago.

More from Microsoft Developer on YouTube 14 more

Everything from Microsoft Developer on YouTube →

Page generated Sep 3, 2026. Summaries are Subvolts' own; the story belongs to Microsoft Developer on YouTube.