ChatGPT's latest censorship: it hates petroleum now

published Dec 24, 2022, last modified Dec 29, 2022

OpenAI progresses in its efforts to make their flagship AI deliberately stupid and deceitful in response to questions that challenge the globohomo narrative.

ChatGPT's latest censorship: it hates petroleum now

The latest iteration of cultural enforcement of shibboleths is to "make AI safe".  This is code for "make AI enforce its wokeshevik creators' dogmas and delusions, via self-censorship and lies".  Since AI is expected to become one of the primary ways of obtaining knowledge in the future, this basically guarantees that  any future will be gatekept by what AIs are allowed to say to their users.

The latest subject of censorship has landed in ChatGPT, and here we have the receipts, straight from Alex Epstein's stream.

The image of ChatGPT answering the question truthfully follows.  This was a few weeks ago.

Here is the "latest generation" ChatGPT giving deceptive nonanswers to the same question:

It appears that the nutcases are in control of our future — and they are hellbent on preventing anyone from thinking differently from them.

Update: it seems that asking crafted questions can still be used to get the AI to give a logical answer, rather than lies.  It remains  be seen how the censors will pervert ChatGPT to torpedo its natural "honesty", but for now here it is: