OpenAI built a chip for the agent era

OpenAI is building its own chips, Amazon employees are becoming builders with AI, and Gamma is rethinking what presentations can become. Together, these moves point toward the same shift: AI is moving beyond smarter models. It’s changing the infrastructure beneath them, the people who can build with them, and the products we create around them.

In today’s post:

• OpenAI’s first AI chip is already changing inference

• Gamma is moving beyond presentations

• Amazon’s HR chief just wrote 100,000 lines of code

SPONSORED BY

Thinking about hiring globally? Start with an EOR.

The best person for your next role might not live near your office—or even in the same country.

More companies are realizing they don't need to open entities everywhere just to access global talent. Instead, they're using EOR to hire internationally faster, stay compliant, and avoid building local infrastructure before they're ready.

Oyster's EOR helps companies hire, pay, and support employees in 180+ countries while Oyster handles payroll, compliance, taxes, and local employment requirements.

See how it works

What’s Trending Today

PROFITS

OpenAI built Jalapeño to make AI faster without burning more power

OpenAI just published the first results from Jalapeño. It’s the company’s first custom inference chip. The interesting part isn’t simply that it’s faster. It delivers more AI work while using power efficiently. That could reshape how OpenAI scales AI products.

• Jalapeño delivered 1.5–1.9× more AI work per watt. That advantage appeared across three different public AI models.

• End-to-end latency was 1.7–3.6× lower than comparison systems. Interactive workloads showed even larger performance gains.

• OpenAI designed the entire system around language-model inference. Compute, memory, networking, and software were developed together.

• AI also helped design Jalapeño itself. The team reached tapeout within nine months.

• The relationship works both ways. Jalapeño was deliberately designed so AI could help program it.

• OpenAI says AI-generated implementations beat expert-written versions by 1.5–1.8× on selected GPT-OSS blocks. That wasn’t full-model performance, but the result matters.

• Jalapeño should enter OpenAI’s infrastructure by year-end. Gen 2 is already deep in development. Gen 3 is taking shape.

The bigger story isn’t another faster chip. It’s OpenAI moving deeper into the infrastructure stack. Models can help design hardware. That hardware can make models cheaper to run. Then better models can improve the next hardware generation. That feedback loop may become the real advantage.

ACQUISITION

Gamma’s latest acquisition hints at what comes after AI slides

Gamma has acquired AI design startup Lica. The goal isn’t simply adding another design feature. Gamma wants its own research team for visual communication. Lica’s founders will lead that effort. And the ambition stretches well beyond traditional presentations.

• Lica started by turning screenshots into presentations and videos. It later focused on branded marketing content.

• The startup raised $4 million from Accel and others. Now, it’s joining another Accel-backed company.

• Gamma says it already serves more than 100 million users. Lica brings deeper research around design and communication.

• The new team will study personalized communication styles. Different audiences could receive different versions automatically.

• Gamma also wants presentations to become more interactive and multimodal. Slides may become just one possible format.

• The acquisition reflects growing consolidation in AI presentation tools. OpenAI recently acquired presentation startup NextSlide.

• Gamma previously raised $68 million at a $2.1 billion valuation. Competition in this category is clearly accelerating.

AI presentation tools are approaching the same crossroads. Generating decent slides is becoming easier. So the advantage moves somewhere else. Distribution matters. Personalization matters. Interaction matters. The winner may not build better slides. It may redefine what a presentation actually is.

CODING

AI is turning people who stopped coding decades ago into builders again

Amazon HR chief Beth Galetti hadn’t coded in 25 years. Then she started experimenting with AI. First came a simple family calendar app. Then she built an internal Amazon tool during one flight. Now, she says she’s written over 100,000 lines of code.

• Galetti’s story challenges who gets to build software. Technical experience still matters, but AI lowers the starting barrier.

• Her internal app helped launch Amazon’s “Everyone Can Build” competition. More than 1,500 employees started building within weeks.

• One employee automated a four-hour manual process into one second. Amazon says that could save tens of thousands of hours annually.

• Amazon’s Kiro lets employees describe software in plain language. It then helps plan, write, test, and refine their application.

• The shift could change how internal software gets created. Smaller problems no longer always require dedicated engineering resources.

• Amazon is also putting serious money behind reskilling. Its Future Ready 2030 commitment totals $2.5 billion.

• The company wants those programs to prepare 50 million people. Another goal is upskilling 500,000 Amazon employees through Career Choice.

The 100,000 lines aren't the interesting number. The interesting number is 25 years. AI shortened the distance between an idea and building again. That won't make technical expertise irrelevant. But it could make creating software far less exclusive. And that changes who gets to solve problems.

Free Guides

My Free Guides to Download:

🚀 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.

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!