When OpenAI launched ChatGPT on November 30, 2022, it quickly turned from a simple experiment into a global phenomenon. In a record-breaking five days, it amassed over 1 million users—a milestone that took Instagram months and Netflix years to reach. By January 2023, it had soared to 100 million monthly users, making it the fastest-growing consumer app ever.

Fueled by this explosive adoption, OpenAI rolled out GPT-4 just several months after launch, delivering stronger reasoning, image support, and more reliable outputs. GPT-4 quickly became the foundation for tools like Microsoft’s Word and Excel Copilot—and countless innovations across business and creativity.

Now, with GPT-5, we’re entering the next phase. From quirky launch to a reinvented productivity powerhouse, ChatGPT’s evolution reflects how AI is reshaping how we learn, create, and work.

When GPT-4 launched in 2023, it set the bar for what an AI assistant could do. It was better at reasoning, more reliable, and, for the first time, multimodal, meaning it could process both text and images. For many, GPT-4 felt like a true digital companion rather than just a clever chatbot.

GPT-5, however, marks another leap. While it builds on the same foundation, three key differences make it stand out:

In short, GPT-4 impressed us with its intelligence. GPT-5 builds on that intelligence with depth, memory, and a smoother, almost human-like flow.

For many people, the big question isn’t just how advanced GPT-5 is, but what that actually means for daily use. With GPT-4, we saw AI become a study buddy, coding partner, and creative tool. GPT-5 takes all of that further — and opens new doors.

In other words, GPT-5 isn’t just a tool you “use” — it’s something you can work with. The line between human and AI collaboration is blurring, and GPT-5 makes that shift more tangible than ever.

It’s easy to get swept up by GPT-5’s strengths—but it still has important limitations worth knowing.

Not a Reliable Source – Nick Turley, Head of ChatGPT, warns that despite its improvements, GPT-5 can still produce confident but incorrect answers (“hallucinations”). He recommends treating it as a second opinion, not a trusted fact source. 

No Continuous Learning or AGI – Sam Altman clarifies that GPT-5 is not a step toward AGI. It doesn’t learn from ongoing use or adapt to changing contexts—so it won’t grow smarter with time. 

Underwhelming Upgrade Hype – Rather than a leap forward, many users found GPT-5’s improvements incremental. The Verge called its shift “underwhelming,” citing a more robotic tone, and OpenAI had to reinstate older models like GPT-4o in response. 

Loss of Comfort and Tone – Earlier versions felt warmer and more empathetic. With GPT-5’s shift to safer, neutral responses, some users feel a loss of emotional connection. Sam Altman acknowledged this eroded the bond many had with the older model. 

Mistakes in Simple Tasks Still Occur – Even now, GPT-5 can fumble with basic math and geography. One user highlighted a small decimals slip during an exchange—so it’s still imperfect for precise tasks. 

Struggles with Complex, Layered Tasks – GPT-5 doesn’t excel at multipart or deeply nested reasoning unless tasks are broken into smaller chunks. 

Academic Research Restrictions – Research workflows, especially in areas like biological data analysis, have been hampered. OpenAI’s new policies now prevent GPT-5 from handling even pre-clinical datasets without verified credentials, frustrating many academic users.

Spatial Reasoning Remains Limited – Even with multimodal advances, GPT-5 still falls short of human performance in spatial tasks, according to a new empirical study. 

For a different approach to what GPT-5 still can’t do read this Medium article here.

GPT-5 is a major step forward from GPT-4—faster, more versatile, and capable of tackling a wider range of creative and analytical tasks. While it still makes mistakes, struggles with nuance, and doesn’t learn from experience, it’s already a powerful tool for writing, problem-solving, and exploration. The key is using it wisely: take advantage of what it can do, stay aware of what it can’t, and let it push your own ideas further.

Keeping up with AI is hard – we get it!

That’s why over 1M professionals read Superhuman AI to stay ahead.

Sign up and start mastering AI

If you’re intrigued and don’t want to miss the full post this Sunday, don’t forget to subscribe if you still haven’t.

What you’ll get by subscribing:

In the meantime, sit back, relax, and start reading all our previous articles tailored just for you.