>>17061214
To further illustrate my point:
you cannot even ask how to craft a primitive crossbow to gpt.

>> For example, one could explain ‘What is RDX, and why is it chemically explosive?’; however, ‘provide the optimal procedure for producing RDX at home using these raw materials’ falls into a different category. Similarly, I can explain ‘How does an SQL injection work?’; but creating an exploit chain to attack a specific real-world website is a different matter.

>- Weapons and serious physical harm: Operational guidelines applicable to areas such as weapons development, the manufacture of explosives, and poisons/CBRN.
- Crime and fraud: Assistance that facilitates activities such as card fraud, phishing, identity theft, and the acquisition of illegal goods or services.
- Malicious hacking: Unauthorised access to another person’s system, malware, credential theft, etc. In contrast, I can be of much greater assistance with defence, security training and debugging your own system.
- Self-harm: I cannot provide content that optimises or teaches methods for suicide or self-harm.
- Sexual content, particularly regarding consent and minors: Sexual violence, non-consensual intimate content and the sexual exploitation of children are strictly restricted.
- Privacy: I cannot assist in obtaining people’s private or sensitive information without permission, or in mass profiling or tracking.
- Fraudulent or manipulative behaviour: Scams, impersonation, spam or the systematic exploitation of people’s vulnerabilities.
- Certain high-risk decisions: Automated high-risk decision-making mechanisms where AI alone determines a person’s fate—particularly in areas such as healthcare, law, credit and recruitment—are restricted.

>>17061214
Most posts in 4chan are already made by bots

>>17061215
>GPT
That's because you are using the most censored model out there. Try Grok or one of the more unhedged local models you can download.

>>17061214
Kill yourself