OpenAI capitulated to feelings and returned warmth to ChatGPT

OpenAI released GPT-5.1, and this is not just an update. This is a capitulation to human feelings. The updated lineup includes 2 models: GPT-5.1 Instant and GPT-5.1 Thinking.

Instant – this is the version for those who missed warmth. After the release of 5, a huge mass of people complained: the model became soulless and cold. Everyone was nostalgic for GPT-4o. And now OpenAI surrendered. Quote: “GPT-5.1 Instant, the most used ChatGPT model, has now become warmer and more conversational by default. Judging by early test results, it often surprises people with its playfulness while remaining understandable and helpful”.

Thinking – this is pumped-up reasoning. The model became better at handling instructions and adapting: quickly responds to simple questions and thinks even longer on complex ones. Quote: “We’re also modernizing GPT-5 Thinking to make it more efficient and understandable in everyday use. Now it more accurately adapts thinking time to the question”.

Very unusual that the first point in the company’s address talks about the model’s friendliness. There are no benchmarks at all.

It turns out, for years the industry chased benchmarks. Who’s smarter, who’s faster, who’s more accurate. And OpenAI takes and releases an update with no numbers at all. Only the promise of warm conversations. It turned out users miss friendly AI more than brilliant AI. This is unexpected.