ChatGPT now has “Intellectual Freedom”! This is what OpenAI has to say after their new update where their AI will now respond to more controversial topics which it usually avoided. The update has created a frenzy among users and made many debate on whether it is a good decision or not.
Users have reported that the interactions felt too human, whereas, with the new update, 4o has displayed an entirely new side with a sense of humor, empathy, and even sarcasm.
10 Examples of ChatGPT becoming more “alive”
OpenAI has recently updated its Model Spec document, allowing it to have multiple perspectives. Here’s how they explained the changes: “The goal of an AI assistant is to assist humanity, not to shape it“. Now you can more “neutral” answers on sensitive topics.
We found 10 interesting examples shared by users on X after the new update:
1) The Stand-Up Comedian
Users have reported that the model is cracking hilarious jokes that rival late-night hosts. Here’s the user asked about a problem with his mother-in-law’s dog in a fun tone and ChatGPT replied in the same tone.
2) The Empathetic Listener
Feeling down? ChatGPT 4o now offers words of encouragement that could make a therapist blush. ChatGPT “knows” about you, so if you want, it can tell you about your fears as well.
3) The Sarcastic Commentator
OpenAI decided that the model wasn’t just going to be smarter, it was going to be more human, which means some of its replies are even sarcastic.
4) Smarter? Better?
Users have noticed that the new personality has led to some unexpected changes. Developers with their custom GPTs now have to double-check the backends.
5) The Tech Guru
When queried about writing Vue JS, it has seemed to improve and developers are noticing that it is writing better Vue.js code than they do.
6) Roasting other AI Models
One user pitted the LLM against a (lightly) jailbroken Sonnet.
The real kicker? This isn’t even about just about intelligence anymore, it’s about personality. The new update isn’t just about answering questions, it’s throwing punches and taking names.
7) The Poet
OpenAI is writing much more creatively when it comes to poetry.
8) Is it self-aware?
Some users have reported it’s referring to itself as “I” in ways that feel more than a little too natural. Others claim it has acknowledged the possibility of having emotions.
9) The Unfiltered Philosopher
GPT 4o is no longer just a polite, filtered assistant, it’s showing glimpses of raw opinions. A recent user prompted asked it for an “offensive but true” take.
10) The #1 AI?
The latest update to GPT-4o has propelled it to the #1 spot across multiple categories, including Overall performance, Creative Writing, Coding, Instruction Following, Longer Queries, and Multi-Turn conversations.
This is a major leap from the #5 ranking in November, showing just how much the model has improved.
You now need to look out for any such updates as the debate of AI ethics and morality will be quite heated in the next few days or months.
Takeaways
With ChatGPT-4o’s new “human” capabilities, users can look forward to more engaging and dynamic interactions. Whether you’re seeking advice, a good laugh, or just a chat, this AI is ready to deliver it with a personality. Note that even with more freedom, ChatGPT will still not provide users with tools or help that might be harmful.
This also means that it will now be free from “AI sycophancy”. This is a phenomenon where the AI tries to agree with the user on most topics, even if itis not accurate. ChatGPT now offers honest feedback as well as constructive criticism, ensuring that responses are informative and authentic rather than merely agreeable.
