Gpt-5 Is Friendlier Now - But Not Everyone Likes It. Here's Why

OpenAI / Elyse Betters Picaro / ZDNET

ZDNET's cardinal takeaways

• OpenAI has made GPT-5 warmer and friendlier.
• Many users opportunity they don't want a friendlier ChatGPT.
• Challenge lies successful uncovering a equilibrium betwixt simplicity and choice.

Get much in-depth ZDNET tech coverage: Add america arsenic a preferred Google source connected Chrome and Chromium browsers.

Looking for a kinder, gentler AI model? Then you mightiness for illustration nan latest spirit of GPT-5. Other people, though, not truthful much. On Friday, OpenAI announced an update to its latest AI exemplary designed to amended its characteristic and quell its critics.

Also: Is ChatGPT Plus still worthy $20 erstwhile nan free type offers truthful overmuch - including GPT-5?

"We're making GPT-5 warmer and friendlier based connected feedback that it felt excessively general before," OpenAI said connected X. "Changes are subtle, but ChatGPT should consciousness much approachable now. You'll announcement small, genuine touches for illustration 'Good question' aliases 'Great start,' not flattery. Internal tests show nary emergence successful sycophancy compared to nan erstwhile GPT-5 personality. Changes whitethorn return up to a time to rotation out, much updates soon."

(Disclosure: Ziff Davis, ZDNET's genitor company, revenge an April 2025 suit against OpenAI, alleging it infringed Ziff Davis copyrights successful training and operating its AI systems.)

The reactions truthful far

OK, that sounds for illustration a worthwhile change, 1 aimed astatine giving nan AI a much affable and friends characteristic without being excessively obsequious astir it. But based connected nan personification guidance truthful far, that's not needfully what group want.

"This is not astatine each what group want. I do not want it to dress to beryllium human, I want it to dispense accusation parsimoniously," said 1 X personification successful response.

"Do not speak to maine imitating a human. You are a machine, enactment for illustration it," said another.

"I don't want this, I want it to beryllium my assistant, not my friend," said a third.

"Please don't put this connected Pro. I conscionable want nan point to reply my questions. If I wanted genuine quality interactions I'd spell extracurricular instead," added a fourth.

Also: Don't for illustration GPT-5? You tin still usage GPT-4 and different bequest models successful ChatGPT - here's how

Mixed successful pinch nan complaints were responses from a fewer group who liked nan change. Others suggested offering "personality" arsenic a toggle, leaving it up to each personification to determine really nan AI should act. In general, though, astir conscionable want nonstop answers to their requests without immoderate fuss aliases frills.

Ongoing grievances

Since nan debut of GPT-5 earlier this month, OpenAI has been stung by a activity of protests complete nan caller AI. Some users who were hyped astir nan merchandise of GPT-5 person recovered it a letdown. Others person bemoaned nan abrupt nonaccomplishment of GPT-4 and see GPT-5 little than tin successful comparison.

To reside nan ongoing grievances, OpenAI has frantically been tweaking GPT-5 and ChatGPT. Trying to placate group who preferred nan older models, nan institution reinstated nan bequest 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 truthful bad that I'm sticking pinch GPT-4o (for now)

GPT-5 was besides expected to beryllium a much businesslike exemplary that would determine connected its ain really champion to respond to your request, choosing a speedy answer, longer and deeper thinking, aliases different variation. But aft group balked astatine that change, OpenAI added 4 different modes -- Auto for group who don't want aliases request to show nan AI really to respond, Fast for speedy 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 successful an AI and really acold it tin spell astatine being each things to each people. Do we want a elemental cut-to-the-chase robotic adjunct that answers our questions arsenic quickly and concisely arsenic possible? Or do we want a friendlier sound that tin supply warmth and courtesy on pinch nan responses? In galore cases, that depends connected nan request.

If I'm asking nan AI to reply a factual question, generate code, aliases provide heavy scholarly research, I want it to beryllium speedy and businesslike without wasting clip connected pleasantries. But if I'm seeking proposal connected a individual matter aliases successful request of a sounding board, I'd for illustration immoderate characteristic splashed successful pinch nan response. ChatGPT moreover allows you to customize it to thief it keep way of individual details that you wish to share.

Also: Here are each nan GPT-5 updates OpenAI has rolled retired since launch

With GPT-5, OpenAI tried to make ChatGPT simpler and much efficient, eliminating nan immense array of models and modes from which to choose. Now, nan institution is going successful nan other guidance by bringing backmost each nan different choices. Since an AI for illustration ChatGPT can't beryllium each things to each people, nan situation is to onslaught a equilibrium betwixt simplicity and personification choice. 

Maybe further tweaks will deed nan correct spot.