OpenAI / Elyse Betters Picaro / ZDNET
ZDNET’s key takeaways
- OpenAI has made GPT-5 warmer and friendlier.
- Many users say they don’t want a friendlier ChatGPT.
- Challenge lies in finding a balance between simplicity and choice.
Get more in-depth ZDNET tech coverage: Add us as a preferred Google source on Chrome and Chromium browsers.
Looking for a kinder, gentler AI model? Then you might like the latest flavor of GPT-5. Other people, though, not so much. On Friday, OpenAI announced an update to its latest AI model designed to improve its personality and quell its critics.
Also: Is ChatGPT Plus still worth $20 when the free version offers so much – including GPT-5?
“We’re making GPT-5 warmer and friendlier based on feedback that it felt too formal before,” OpenAI said on X. “Changes are subtle, but ChatGPT should feel more approachable now. You’ll notice small, genuine touches like ‘Good question’ or ‘Great start,’ not flattery. Internal tests show no rise in sycophancy compared to the previous GPT-5 personality. Changes may take up to a day to roll out, more updates soon.”
(Disclosure: Ziff Davis, ZDNET’s parent company, filed an April 2025 lawsuit against OpenAI, alleging it infringed Ziff Davis copyrights in training and operating its AI systems.)
The reactions so far
OK, that sounds like a worthwhile change, one aimed at giving the AI a more affable and friendly personality without being too obsequious about it. But based on the user reaction so far, that’s not necessarily what people want.
“This is not at all what people want. I do not want it to pretend to be human, I want it to dispense information parsimoniously,” said one X user in response.
“Do not speak to me imitating a human. You are a machine, act like it,” said another.
“I don’t want this, I want it to be my assistant, not my friend,” said a third.
“Please don’t put this on Pro. I just want the thing to answer my questions. If I wanted genuine human interactions I’d go outside instead,” added a fourth.
Also: Don’t like GPT-5? You can still use GPT-4 and other legacy models in ChatGPT – here’s how
Mixed in with the complaints were responses from a few people who liked the change. Others suggested offering “personality” as a toggle, leaving it up to each user to decide how the AI should act. In general, though, most just want direct answers to their requests without any fuss or frills.
Ongoing grievances
Since the debut of GPT-5 earlier this month, OpenAI has been stung by a wave of protests over the new AI. Some users who were hyped about the release of GPT-5 have found it a letdown. Others have bemoaned the sudden loss of GPT-4 and consider GPT-5 less than capable in comparison.
To address the ongoing grievances, OpenAI has frantically been tweaking GPT-5 and ChatGPT. Trying to placate people who preferred the older models, the company reinstated the legacy GPT-4o, GPT-4.1, o4-mini, and o3 models, but only for Plus and Pro subscribers.
Also: I tested GPT-5’s coding skills, and it was so bad that I’m sticking with GPT-4o (for now)
GPT-5 was also supposed to be a more efficient model that would decide on its own how best to respond to your request, choosing a quick answer, longer and deeper thinking, or another variation. But after people balked at that change, OpenAI added four different modes — Auto for people who don’t want or need to tell the AI how to respond, Fast for quick answers, Thinking for longer thinking, and Thinking mini mode for quicker thinking.
Rocky reception
The rocky reception that’s greeted GPT-5 brings up questions of what we want in an AI and how far it can go at being all things to all people. Do we want a simple cut-to-the-chase robotic assistant that answers our questions as quickly and briefly as possible? Or do we want a friendlier voice that can provide warmth and courtesy along with the responses? In many cases, that depends on the request.
If I’m asking the AI to answer a factual question, generate code, or provide deep scholarly research, I want it to be quick and efficient without wasting time on pleasantries. But if I’m seeking advice on a personal matter or in need of a sounding board, I’d like some personality splashed in with the response. ChatGPT even allows you to customize it to help it keep track of personal details that you wish to share.
Also: Here are all the GPT-5 updates OpenAI has rolled out since launch
With GPT-5, OpenAI tried to make ChatGPT simpler and more efficient, eliminating the vast array of models and modes from which to choose. Now, the company is going in the opposite direction by bringing back all the different choices. Since an AI like ChatGPT can’t be all things to all people, the challenge is to strike a balance between simplicity and user choice.
Maybe further tweaks will hit the right spot.
