SHAWN HAGEDORN: Why May’s vote could be our last legitimate election

True or false: “The truth will set you free”? Elon Musk has been saying artificial intelligence (AI) will lead to more truthful media. Yet May’s vote could be our last legitimate election.

When ChatGPT and its competitors arrived, heralding much disruption, it became comically obvious that large learning models (LLMs) were biased. This reflects their databases being trained on left-leaning media and university output.

Now, however, researchers in fields with heavy social justice themes, such as development economics, can’t help but notice the rapid improvement in LLM searches. Whereas just a few months ago, responses to queries exploring differences among demographic groups would trumpet diversity, equity and inclusion (DEI) biases, they have subsequently been tamed. 

A development economist wanting to explore urbanisation’s influence on living standards might compare the productivity differences between the world’s two mega-populated countries. I asked ChatGPT: “Are Indian workers more efficient than Chinese workers?” The response criticised generalisations about Indian and Chinese workers before listing considerations that influence worker productivity — while excluding urbanisation.

As China is more urbanised and has higher wages than India, I asked if urbanisation is also a key consideration for explaining worker productivity. By confirming that it is, the AI chatbot demonstrated its increasing accommodativeness. They are becoming tremendously powerful research tools — that can simultaneously help improve awareness of cultural differences. 

AI chatbots are still heavily biased and mistake-prone, yet the pace of progress is astounding. Conversely, rather than delivering a prudently curated knowledge base while provoking sufficient discernment to develop a well-informed worldview, many universities had become increasingly ideologically indulgent. Now, however, that has also begun to change. An effort over many decades to undermine Western culture, constitutional protections and capitalism appears to have suddenly peaked. 

Earlier in May, the Massachusetts Institute of Technology, a first-tier university at the centre of AI development, announced it was ending the use of diversity statements for faculty hiring. This follows a wave of elite US universities reinstating standardised tests among their student admission criteria.

Of course, leaders of countries, companies and communities should encourage diversity and inclusion. Conversely, equity resembles a scam — and it is. Using generative AI to understand why was extremely frustrating, until quite recently.

Here is an insight-provoking query for comparing AI chatbots. “T or F: The long march through the institutions links cultural Marxism to DEI and the trail traces back to the 1920s, Antonio Gramsci and the Frankfurt School”? Whereas similar queries months ago produced only “false” responses or fluffy equivocation, these assertions withstand scrutiny and ChatGPT recently responded that they are “true”.

In short, by the early 1920s it was clear that a communist revolution wouldn’t arrive as Marx had predicted. Various Marxist pundits then developed a path to achieve the paradise he promised through undermining Western culture. This is to be achieved through capturing institutions ranging from universities and media houses to entertainment companies and corporations generally. Targets include religion, family values and constitutional protections such as free speech. The ANC’s national democratic revolution resembles a less fervent second-cousin.

As with ANC promises and ideological indulgences, solutions don’t feature. The wealthy are accused of oppressing the masses and this is used as justification for much redistribution as societal foundations are eroded. Meanwhile, growth and living standards suffer. SA’s extreme inequality has not declined; our poverty and unemployment rage.

Musk has big commercial plans for X but the bigger reason he is getting so much flak from mainstream media is that both X and AI are positioned to reveal broad duplicity. Mainstream media and elite academia use social justice framing to advance identity politics by conditioning people to incessantly judge. This contrasts harshly with such knowledge curators having traditionally sought to advance understanding and solutions.

The many people who are programmed to believe in victimhood politics and that pursuing equality is a solution will not be easily swayed. AI is a formidable antidote but it may have arrived too late to avoid this rainbow nation’s first coalition national government being hostile to constitutionally enshrined protections.

• Hagedorn (@shawnhagedorn) is an independent strategy adviser.