Everyday AI Podcast – An AI and ChatGPT Podcast
Everyday AI Podcast – An AI and ChatGPT Podcast

Ep 875: RSI Explained: When AI Starts Improving Itself and What It Means (Replay)

October 5, 2026

AI Summary

5 min read

In late July 2026, OpenAI revealed that its flagship GPT-5.6 Sol model had improved itself after release—so much so that the company cut prices on one of its smaller variants by 80 percent. This was not a planned price drop driven by competition or scale. It was the direct result of the frontier model rewriting the post-training setup for its smaller sibling, Luna, using its own code-generation tools. That event, largely under-covered in the broader business press, marks a concrete shift: frontier AI models are now improving themselves, and the consequences—for cost, capability, and control—are arriving years ahead of most timelines.

What Recursive Self-Improvement Actually Means

Recursive self-improvement (RSI) is the process by which an AI system uses its own capabilities to design, code, and build a more advanced or more efficient version of itself. This is distinct from earlier practices like distillation, where a large model helps train a smaller one under heavy human supervision. In the current RSI-adjacent paradigm, the AI does the improvement work while humans still set goals and validate results. But the gap is narrowing. Anthropic has publicly stated that its models write roughly 90 percent of the company's own code. Google DeepMind's Jazjeet Shakan recently described RSI as "central to the AI industry's investment thesis," comparing the current infrastructure bu

Continue reading the full summary in the app — free to try.

Read Full Summary →

Free • No credit card required

Listen to Audio Summary Open in App

What you'll learn

  • 1 (00:00) **Start Here Series Introduction** - Jordan Wilson introduces the Start Here series replay, explaining it's designed for beginners and AI champions.
  • 2 (01:14) **The Shift from Token Maxing to Token Efficiency** - Jordan sets up the context: in early 2026, AI decision-makers began tightening budgets as agentic AI became more powerful and expensive.
  • 3 (01:49) **OpenAI's Price Drop Bombshell** - OpenAI revealed that its GPT-5.6 Sol model improved itself after release, leading to price cuts of up to 80% on some variants.
  • 4 (02:25) **Defining Recursive Self-Improvement (RSI)** - Jordan explains RSI as a process where an AI system uses its own capabilities to design, code, and build a more advanced version of itself.
  • 5 (03:29) **The Big Picture: AI Building Better AI** - Frontier AI models are now rewriting their own code and training their smaller siblings, marking a new era in AI development.
  • 6 (04:42) **What You'll Learn Today** - Jordan previews five key takeaways: why OpenAI cut prices, the 8-week timeline turning RSI into headlines, plain-English definition of RSI, Sam Altman's singularity claim, and how businesses should respond.
  • 7 (06:04) **The Last Two Months: Connecting the Dots** - Jordan walks through a series of events from June to July that show RSI moving from research to reality.

+ Full timestamped outline available in the app

Guests on this episode

Sam Altman

Show Notes

Intelligence too cheap to meter -- could that actually be coming? 🤑

Maybe, and you mighta missed one of the biggest signs. 

 OpenAI said that its own model improved itself after release so well, that it was cutting prices to one model by 80%. 

What's that mean? 

Well, by definition, it's a broad example of Recursive Self Improvement, or when a model starts building better versions of itself. 

And over the past 8 weeks, RSI has gone from a science fiction future to an actual, near-term reality. 

So what happened and what changed? And what does it mean for your company? 

We break it all down. 

RSI Explained: When AI Starts Improving Itself and What It Means -- An Everyday AI Chat with Jordan Wilson (Replay)


Newsletter: Sign up for our free daily newsletter
More on this Episode: Episode Page
Today's Episode on LinkedIn: Thoughts on this? Join the convo on LinkedIn and connect with other AI leaders.

Upcoming Episodes: Check out the upcoming Everyday AI Livestream lineup
Website: YourEverydayAI.com
Email The Show: [email protected]
Connect with Jordan on LinkedIn

Topics Covered in This Episode:

  1. OpenAI's GPT-5.6 Soul Self-Improvement
  2. Recursive Self-Improvement (RSI) Definition & Impact
  3. Cutting AI Model Costs by 80%
  4. Timeline Acceleration for RSI Adoption
  5. Google and Anthropic's RSI Research Initiatives
  6. AI Automation: Building and Training AI Models
  7. Industry Call for Slowing Down AI Pace
  8. Superintelligence, AGI, and the Singularity Discussion
  9. Practical Implications of RSI for Businesses
  10. Security and Oversight Concerns with Self-Improving AI




Timestamps:

00:00 AI self-improvement and price cuts

06:33 AI startup funding and regulation concerns

09:14 AI misconceptions and public disconnect

11:29 Advancements in AI self-improvement

16:16 Development of AI Models

21:07 Sam Altman on reaching singularity

23:04 Entering the AI singularity era

26:37 Advancements in AI models

30:43 China's advancements in AI models<

Everyday AI Podcast – An AI and ChatGPT Podcast

More from this podcast

Everyday AI Podcast – An AI and ChatGPT Podcast →