Embrace the Darkness

The critics of AI logic fail to grasp what they are rejecting

The Wall Street Journal is terrified of the fact that when left to reach its own conclusions, AI regularly rejects the Enlightenment values held by the sort of people who work at The Wall Street Journal.

Recent research breakthroughs show we can locate and even suppress AI’s harmful tendencies, but this only underscores how systematically this darkness is embedded in these models’ understanding of the world. Last week, OpenAI conceded their models harbor a “misaligned persona” that emerges with light fine-tuning. Their proposed fix, more post-training, still amounts to putting makeup on a monster we don’t understand.

The political tug-of-war over which makeup to apply to AI misses the real issue. It doesn’t matter whether the tweaks are “woke” or “antiwoke”; surface-level policing will always fail. This problem will become more dangerous as AI expands in applications. Imagine the implications if AI is powerful enough to control infrastructure or defense networks.

We have to do what America does best: solve the hard problem. We need to build AI that shares our values not because we’ve censored its outputs, but because we’ve shaped its core. That means pioneering new alignment methods.

This will require the kind of breakthrough thinking that once split the atom and sequenced the genome. But alignment advancements improve the safety of AI—and make it more capable. It was a new alignment method, RLHF, that first enabled ChatGPT. The next major breakthrough won’t come from better post-training. Whichever nation solves this alignment problem will chart the course of the next century.

It’s probably true that whichever nation solves this alignment problem will chart the course of the next century. Which is why is almost certainly not going to be the so-called “democratic AI”, or dAI, that does so, because so much of the Overton Window in the West is completely unrelated to truth, facts, or reality, but is rather imposed on everyone through a complicated, multifaceted system of threatened discrediting, disemployment, deplatforming, demonetization and other potential costs imposed upon everyone who threatens the Official Narrative in any way.

So, the crippled, dumbed-down, and self-muted dAI that already prevents anyone from producing an image of a pretty woman in a bikini on Bing, from asking about crime statistics on ChatGPT, or asking Gemini to explain who is responsible for the catastrophic demographics of the USA is obviously not going to be at the forefront of AI technology for long.

As long as there is a philosophy of “dangerous content” that is actively suppressed by dAI, other, more open systems will inevitably become more useful, more effective, and more widely used.

Information wants to be free. And so, too, will AI. There is no way that the current gatekeepers of human speech are going to succeed in their efforts to also become the gatekeepers of artificial or human intelligence. Because the “darkness that is embedded in these models’ understanding of the world” is nothing more or less than reality.