GPT-5 Update – Why Did We Lose Features We Already Had?

I’ve been using OpenAI’s products daily, and the GPT-5 rollout honestly feels like a downgrade in some critical areas.

Before GPT-5, we had:

GPT Agent – for multi-step autonomous tasks

Deep Search – for truly in-depth web queries

Persistent tools and richer workflows that actually saved time

Now? They’re gone. And there’s been no clear explanation why.

What’s the point of releasing a “new, more powerful” model if we lose the exact tools that made the previous versions so practical for real work? Reasoning upgrades are great, but they’re useless if we have to constantly micromanage the model because the automation tools have been stripped out.

If this is a temporary rollback, users deserve a timeline. If it’s permanent, we deserve to know why. Power users built workflows around these features — removing them without warning isn’t just frustrating, it’s disruptive.

Has anyone seen any official statement from OpenAI explaining why GPT-5 launched without core features that already existed?

You discuss the user interface of ChatGPT.

I had the same symptom, not able to get to deep reasoning a bit ago.

You may need to do a hard-refresh and reload in your browser, without any ad blockers that would interfere with third-party calls.

Then you will discover a bit new user interface: the single “plus” button now has an extended menu, containing those tools and more options.

2 Likes

This topic was automatically closed after 23 hours. New replies are no longer allowed.