- Nextool AI
- Posts
- OpenAI Is Preparing for AI That Improves AI
OpenAI Is Preparing for AI That Improves AI
Plus: Jensen Huang Puts AI Extinction Odds at Zero
Jafar Najafov
September 22, 2026 • Read this in 3 minutes
In partnership with
AI is entering a more complicated phase. Amazon is blocking Meta’s shopping agent, Nvidia’s Jensen Huang is rejecting AI doomsday fears, and OpenAI is calling for global standards as AI systems grow more autonomous. Together, these stories point toward the same question: as AI gains more power and independence, who decides where the boundaries should be?
In today’s post:
AI could soon help build better AI
Amazon doesn’t want AI agents shopping for you
Jensen Huang puts AI doomsday fears at 0%
SPONSORED BY
The best marketing ideas come from marketers who live it.
That’s what this newsletter delivers.
The Marketing Millennials is a look inside what’s working right now for other marketers. No theory. No fluff. Just real insights and ideas you can actually use—from marketers who’ve been there, done that, and are sharing the playbook.
Every newsletter is written by Daniel Murray, a marketer obsessed with what goes into great marketing. Expect fresh takes, hot topics, and the kind of stuff you’ll want to steal for your next campaign.
Because marketing shouldn’t feel like guesswork. And you shouldn’t have to dig for the good stuff.
What’s Trending Today
UPDATE
AI could soon help build its own successors
Image credits: OpenAI
OpenAI is preparing for a different phase of AI. Models won’t just answer questions. They could increasingly conduct AI research themselves. That could accelerate progress dramatically. It also creates a difficult problem: keeping humans in control.
OpenAI is working toward an automated AI researcher. Humans would remain involved in the improvement process.
The bigger concept is recursive self-improvement, or RSI. AI helps develop increasingly capable generations of AI.
Fully autonomous RSI isn’t happening today. OpenAI says it shouldn’t happen until it can happen safely.
The danger is losing meaningful human oversight. Research could become too fast or complex to understand.
OpenAI believes AI could also strengthen its own safeguards. Automated researchers could accelerate alignment and defensive research.
That makes shared measurements increasingly important. Labs need ways to measure capabilities, autonomy, and serious incidents.
OpenAI proposes international technical standards for frontier AI. Governments would still decide how those standards become law.
The biggest AI shift may not be smarter chatbots. It may be AI becoming part of AI research. That changes the speed of everything. Humans currently set much of the research pace. Automated researchers could compress years of work dramatically. That could unlock extraordinary scientific progress. It could also make mistakes harder to catch. The challenge is simple to describe. AI can accelerate. Human understanding needs to keep up.
PROFITS
Amazon just drew a line around AI shopping
Amazon has blocked Meta’s new Muse AI agent. Not because Muse recommends competing products. Because Muse can shop Amazon for you. That creates a much bigger question. Who controls shopping when humans stop doing the browsing?
Meta’s Muse can handle multi-step shopping tasks. It compares products, fills carts, and asks before purchasing.
Amazon says Muse accessed its store without permission. It also says the agent didn’t identify itself while browsing.
Meta says Muse protects passwords and payment information. Credentials stay secured and hidden from the agent itself.
The fight goes beyond privacy and security. Amazon made over $68 billion from advertising last year.
AI agents could change that advertising model. They may skip sponsored listings and hunt directly for value.
Amazon has already challenged other outside shopping agents. Perplexity, Google, and OpenAI have faced similar resistance.
Yet Amazon is building agents itself. Its Buy for Me feature can purchase products from external websites.
The interesting battle isn’t Amazon versus Meta. It’s websites versus agents. The internet was designed around humans visiting pages. Those visits created attention, advertising, and customer relationships. Agents could remove that entire layer. When software becomes the shopper, websites become infrastructure. And whoever controls the agent may control the customer.
RESEARCH
Jensen Huang says AI won’t end the world
Nvidia CEO Jensen Huang just made a bold prediction. He says AI won’t destroy humanity by 2030. Not probably. Not hopefully. He puts the chance at zero. But his reasoning matters more than the prediction itself.
Huang argues AI doomsday claims lack scientific grounding. He says frightening people without evidence is irresponsible.
His argument starts with incentives. AI companies lose enormous value when their technology causes harm.
Huang believes existing laws should remain important guardrails. Cybersecurity, liability, and damages already create accountability.
That challenges calls for entirely new AI regulations. Huang favors applying existing frameworks before creating replacements.
He also believes unsafe products simply shouldn’t ship. Companies should hold systems back until they’re ready.
Still, Nvidia sits inside this debate financially. Its chips power much of today’s AI development.
That creates an interesting tension. AI builders must accelerate innovation while proving acceleration remains safe.
The useful question isn’t whether AI ends humanity by 2030. Nobody can verify that future today. The better question is what happens before then. AI systems are becoming more capable every year. That creates extraordinary opportunities and unfamiliar risks. Doomsday predictions can distract from practical safety problems. So can dismissing every warning as fear. Progress needs optimism. But optimism works better when paired with evidence.
Free Guides
🚀 Founders & AI Builders, Listen up!
If you’ve built an AI tool, here’s an opportunity to gain serious visibility.
Nextool AI is a leading tools aggregator that offers:
500k+ page views and a rapidly growing audience.
Exposure to developers, entrepreneurs, and tech enthusiasts actively searching for innovative tools.
A spot in a curated list of cutting-edge AI tools, trusted by the community.
Increased traffic, users, and brand recognition for your tool.
Take the next step to grow your tool’s reach and impact.
That's a wrap:Please let us know how was this newsletter: |
Reach 150,000+ READERS:
Expand your reach and boost your brand’s visibility!
Partner with Nextool AI to showcase your product or service to 140,000+ engaged subscribers, including entrepreneurs, tech enthusiasts, developers, and industry leaders.
Ready to make an impact? Visit our sponsorship website to explore sponsorship opportunities and learn more!