OpenAI upgrades Codex with GPT-5

PLUS: How to generate studio-grade AI images with ChatGPT

Good morning, AI enthusiasts. OpenAI has supercharged its coding agent with a dynamic new model, GPT-5 Codex, designed to think longer and smarter on tough programming problems.

PLUS: How to generate studio-grade AI images with ChatGPT

TODAY IN AI

OpenAI is rolling out GPT-5 Codex across Codex products, GitHub, IDEs, terminals, and ChatGPT, giving developers more powerful tools for writing, debugging, and reviewing code. Unlike earlier versions, GPT-5 Codex can flexibly spend anywhere from a few seconds to seven hours on a task.

→ Dynamically adjusts thinking time
→ Outperforms GPT-5 on coding benchmarks
→ Fewer incorrect code review comments
→ Available now on ChatGPT; API coming soon
→ Competes with Claude Code, Cursor, Copilot

Why is important: With its adaptive reasoning, GPT-5 Codex can tackle both quick fixes and complex multi-hour builds, setting a new bar for AI coding tools in speed, accuracy, and autonomy.

Nvidia’s shareholder CoreWeave Inc. announced a cloud services agreement valued at $6.3 billion, supporting AI adoption across industries.

→ Contract runs through 2032 and covers excess capacity not used by Nvidia customers
→ Guarantees AI computing power will be available on demand
→ Part of Nvidia’s broader strategy to accelerate AI infrastructure and adoption

Why is important: Securing dedicated cloud capacity ensures Nvidia can scale AI workloads efficiently, strengthening its leadership in AI computing and supporting enterprises that rely on high-performance AI infrastructure.

HOW TO AI

How to generate studio-grade AI images with ChatGPT

This is a step-by-step method to write a single prompt that produces a professional, magazine-quality product shot every time. We’ll use a VolTX energy drink as the example, but the same structure works for portraits, fashion, tech products, or any commercial image.

🧰 Who is this for

  • Creators and marketers looking to design product ads

  • Startups and small businesses without a photography budget

  • Students and freelancers who want polished visuals fast

  • Anyone wanting to upgrade AI images from “AI-generated” to “studio-quality”

STEP 1: Start with the Subject

This is the foundation of your image. Clearly describe who or what is in the picture. Be specific about gender, age, clothing, and what they’re doing.

Let’s imagine we’re creating a product shot for an energy drink brand. The goal is to make the drink look bold, refreshing, and powerful, something that could fit right into an ad campaign.

Here’s an example subject prompt:

“A confident dynamic young man holding a sleek VolTX Energy drink can near his lips, just before taking a sip. His posture is energetic yet stylish, subtly drawing attention to the can’s bold metallic design with electric blue and silver accents. The energy drink can remains the visual anchor, symbolizing power, focus, and unstoppable energy.”

The clearer your subject, the better the AI will understand the scene.

STEP 2: Add Camera and Composition Details

AI responds best when you add camera directions, just like a real photographer would. Mention angle, framing, depth of field, and perspective.

For Example:

"Camera Angle: Eye level angle 50 mm prime lens crop from waist up centered on the can and his energized expression."

This tells the AI not just what to show, but how to show it.

STEP 3: Set the Mood and Lighting

Lighting and mood completely transform the image. Use terms photographers use, like:

“Soft golden hour light” “Dramatic studio lighting with shadows” “Neon city glow at night” “Dark cinematic tone with high contrast”

For example:

"Mood: Bright directional daylight Vibrant tones with electric blue, neon orange, and silver highlights Urban backdrop with a sleek, sporty edge (like blurred motion trails or a cityscape hint."

STEP 4: Add Realism with Fine Details

The details are what make an image feel real. Add elements like:

Skin texture, pores, freckles, or makeup details Fabric texture on clothes (silk, denim, leather) Image quality keywords: “ultra-detailed, sharp focus, 8K, studio grade”

These micro-descriptions remove the “AI feel” and make it look like a pro shot.

For example:

"Quality: Ultra-realistic 4K editorial lifestyle style Subtle film grain for grit and energy Crisp sharpness to emphasize power and refreshment.

Additional:
Slight sweat for freshness and intensity Defined jawline and athletic build Casual sporty jacket or fitted tee Cold condensation and droplets on the can Subtle mist or splash effect in the air Vertical aspect ratio for print or social."

STEP 5: Generate Image

Here's the final prompt:

"A confident dynamic young man holding a sleek VolTX Energy drink can near his lips, just before taking a sip. His posture is energetic yet stylish, subtly drawing attention to the can’s bold metallic design with electric blue and silver accents. The energy drink can remains the visual anchor, symbolizing power, focus, and unstoppable energy.

Camera Angle: Eye level angle 50 mm prime lens crop from waist up centered on the can and his energized expression.

Mood: Bright directional daylight Vibrant tones with electric blue, neon orange, and silver highlights Urban backdrop with a sleek, sporty edge (like blurred motion trails or a cityscape hint.

Quality: Ultra-realistic 4K editorial lifestyle style Subtle film grain for grit and energy Crisp sharpness to emphasize power and refreshment."

Results with this template:

ESSENTIAL BITES

  • ByteDance introduced Seedream 4.0, a new image generation and editing model that competes with Nano Banana with 4K outputs and multimodal capabilities.

  • Google announced a £5B investment in UK AI, opening a new data center near London and projecting 8,250 jobs annually, ahead of Trump’s state visit.

  • Sequoia’s former Asia arms invested in Marq Vision, a U.S. startup tackling unauthorized AI chip and product sales, in a $48M Series B round led by Peak XV and HSG, bringing total funding to $90M.

⚙️ Hot AI Tools:

  • 🧠 Oboe: Instantly create AI-powered learning courses

  • ⚙️ Replit Agent 3: Build production-ready apps autonomously

  • 🎨 Nonverbia: Turn video meetings into clear, actionable insights.

  • 🤝 Lumro: Deploy AI agents for top-notch customer experience.

That's it! See you tomorrow

- Dr. Alvaro Cintas