ChatGPT-4.5 is more human than humans!

Welcome to a new version of The AI Business!

This week, we have:

● Midjourney finally launches a new model after nearly a year.

● Create better presentations in no time using AI. (FREE)

● ChatGPT-4.5 is more human than humans!

● ChatGPT’s new image generation model is a game-changer.

AI news
Major News in The World of AI Were Announced This Week:

● Midjourney launches V7, its first new model in a year: Midjourney has unveiled V7, its first new AI image model in a year, featuring a revamped architecture and default personalization based on users’ rated images. The model offers sharper textures and more coherent visuals, especially in hands, bodies, and objects, and introduces a fast, low-cost Draft Mode. Some key features like upscaling are still on the way, but V7 already sets a new bar for creative control and prompt responsiveness. (Try Now)

● OpenAI revives o3 and delays GPT-5 as it preps for high demand: OpenAI has reversed course on its o3 model, now planning to release both o3 and the new o4-mini within weeks while pushing GPT-5’s debut to later this year. CEO Sam Altman cited challenges in unifying advanced tools and preparing infrastructure to meet the massive expected demand for the next-gen model. GPT-5 will include multimodal features and tiered intelligence levels, as OpenAI faces mounting pressure from more open rivals. (Read Post)

● Amazon unveils AI agent that shops third-party sites for you: Amazon is testing “Buy for Me,” a new AI-powered feature that can browse external websites and complete purchases on a user’s behalf directly from the Amazon app. Powered by Amazon Nova and Anthropic’s Claude, the agent fills in your details to check out securely—even on non-Amazon platforms. While the move could expand Amazon’s e-commerce dominance, it also raises trust questions about handing purchasing power to an autonomous AI. (Read Article)

● OpenAI raises $40B to scale AI research and build massive data center network: OpenAI has secured a staggering $40 billion in funding at a $300 billion valuation, with SoftBank leading and existing investors like Microsoft joining in. The investment will supercharge AI research and expand infrastructure, including $18 billion earmarked for the ambitious Stargate data center initiative. With 500 million weekly ChatGPT users, OpenAI is doubling down on scaling transformative AI at an unprecedented pace. (Read Article)

● OpenAI backs AI defense startup in its first cybersecurity investment: OpenAI has made its first move into cybersecurity by co-leading a $43 million round in Adaptive Security, a startup that simulates AI-driven hacks to train employees. The platform uses spoofed emails, texts, and calls to expose weak links and fight social engineering threats made easier by generative AI. With over 100 customers already, Adaptive plans to scale up engineering to stay ahead in the fast-evolving AI threat landscape. (Read Article)

You + AI = Unbeatable Presentations

Prezi AI turns your ideas into unique, dynamic presentations in seconds.

Just enter your prompt and you’ll get a stunning, fully-custom presentation—no templates or generic layouts. It’s just like having a professional designer by your side. Prezi AI even helps you refine your text, find the perfect image, and layout your content.

Get polished presentations that look like they took hours—done in seconds, every time.

Try for free today

AI Study
OpenAI’s GPT-4.5 Outstamrts the Turing Test, And You Might Not Know the Difference:

The Study That Redefines “Humanlike” Intelligence:

A recent study out of UC San Diego has put OpenAI’s GPT-4.5 model to the test—literally—and the results are turning heads. In a controlled setting mimicking the famous Turing test, GPT-4.5 was judged to be human 73% of the time when prompted to act with a “humanlike” persona. That means it was more convincing than actual people, who were mistaken for bots more frequently than the AI.

For business leaders, this isn't just a quirky stat—it’s a glimpse into the future of customer interaction, brand communication, and even internal operations.

What This Means for Customer Experience?

If a chatbot can now pass for human more often than a human can, customer-facing AI is no longer a stopgap or secondary channel—it’s a frontline asset. GPT-4.5’s performance shows that with the right prompting, AI can handle complex, conversational engagements that were once the domain of trained service reps.

For marketing and CX teams, that translates to scalable, personalized engagement—without hiring overhead. But it also raises the bar: if your competitors are deploying “superhuman” AI that feels natural and empathetic, sticking with robotic, rule-based bots could quietly lose you customers.

The Power of Prompting: Why Framing Matters?

Interestingly, GPT-4.5 only hit that 73% human-likeness rating when explicitly told to “adopt a humanlike persona.” Without that instruction, its pass rate dropped to 36%. This reveals a key insight: how you prompt your AI matters as much as which model you use.

Smart prompt engineering could become a core competency for marketing and operations teams—whether that means training AI to mirror your brand voice, defuse customer complaints, or write ad copy that converts.

Why It Matters for Business Leaders?

This study offers more than academic curiosity—it’s a signal to reevaluate how your organization is approaching AI:

● Should your sales or support functions integrate high-fidelity AI agents?

● Are you investing in the right prompt design and training to optimize results?

● And critically—how will your brand maintain trust when AI is indistinguishable from human?

As OpenAI’s models grow more sophisticated and mainstream, companies that lean into AI-human collaboration—not just automation—will be better positioned to innovate, engage, and scale.

AI tools
ChatGPT’s New Image Generation Model is a Game-Changer for Businesses:

A Powerful Visual Tool, Now Built Into ChatGPT:

OpenAI’s latest image generation feature—powered by GPT-4o—is now available to all ChatGPT users, making it easier than ever to create high-quality, on-demand visuals from text prompts. Unlike previous tools, this new model offers faster rendering, improved image quality, and a more seamless user experience, directly inside the ChatGPT interface.

For businesses, this means less reliance on stock imagery or external design resources—and more control over brand-aligned content.

Why It Matters for Business Leaders?

AI-generated images are no longer experimental—they’re practical, fast, and commercially viable. Marketing teams can use them to:

● Accelerate content creation for social media, email campaigns, and product launches

● Visualize products or ads before development begins

● Reduce costs on graphic design or stock photo subscriptions

This tool brings design closer to where strategy happens—within the same workflows used for planning, brainstorming, and content writing.

Best Practices to Maximize ROI:

To get the most out of ChatGPT’s image generator, keep these in mind:

● Stay on-brand by refining prompts that reflect your visual identity

● Use images as prototypes to spark creative direction, not final deliverables

● Fact-check and review for visual accuracy or potential bias

Looking Ahead:

From ecommerce and real estate to education and B2B marketing, the potential use cases are broad—and growing. As AI-generated imagery becomes indistinguishable from real photos, businesses that adopt and experiment early will gain a creative edge.

Now’s the time to rethink how your team creates and communicates visually—with speed, scale, and strategy.

That’s it for today, thanks for reading till the end. 😊 

Stay connected on LinkedIn for the latest updates.

See you next week.