Artificial intelligence’s potential for rapid disruption is unprecedented. Given our country’s trajectory, the sudden acceleration of its adoption may have begun just in time.

AI could eventually make humans irrelevant. But, in the here and now, it also has the potential to overcome South Africa’s governance failures by making it easier to frame issues with an orientation toward solutions.

Human communication has always privileged emotions and interpreting intentions. AI algorithms, like ChatGPT, can now simulate comparable capabilities while being much better than humans at unemotionally deciphering how complex, high-volume data sets intersect. As the Covid pandemic demonstrated, societies struggle with such deciphering.

Our 1990s transition prioritised ideals and political remedies. It also led to ANC intentions being routinely misinterpreted. The party was too frequently given the benefit of the doubt. While revelations during the second half of Jacob Zuma’s reign interjected much realism, his being replaced by Cyril Ramaphosa initially triggered a pivot back toward idealised expectations. There has since been a fresh resurgence in realism yet misperceptions persist –particularly regarding managing our economic challenges.

As South Africa has long benefited from sophisticated capital markets expertise, we have become accustomed to using the economic metrics favoured by market participants to gauge our economy. They prioritise the metrics which feed into their models for pricing securities. Normally, such metrics also accurately gauge the health of a country’s economy. But our economy is far from normal.

It was comforting to learn last week that the economy did not contract in the first quarter. But how rational can it be to focus on near-term GDP data when we have the world’s worst youth unemployment crisis and no plan to remedy it?

Everyone, including our poorly educated children, can see that we need to fix Eskom. Conversely, there is remarkably little acknowledgement of how dangerous our youth unemployment crisis is or how difficult it will be to fix. Countries go to great lengths to avoid our level of youth unemployment as it is so economically destructive and politically destabilizing.

Economically doomed

For every one of our employed young adults, there are currently nearly three who are unemployed.  A majority of our young adults are economically doomed and forecasts suggest further deterioration.

It seems reasonable to presume that if we fix our politics and attract enough investment, our economy will then grow sufficiently to eventually resolve our unemployment issues. Unfortunately, such thinking is far more risky than realistic.

Seeking a one-to-one ratio of employed to unemployed young adults is not ambitious. If, however, this is to be achieved through growing the economy, this modest goal would likely remain elusive until today’s school leavers approach middle age. Meanwhile, crime and social upheaval would percolate to the point of tempting dreadful scenarios.

As our economic goals are framed by our ruling party around justice-defined ideals, and we navigate using capital market metrics, it shouldn’t be surprising that we were slow to recognise the damage compounding from Zuma’s reckless embrace of patronage and Ramaphosa’s inability to tame it. Mixing idealistic goals with capital market metrics also helps to explain why none of our leaders can articulate a plan which could deliver broad prosperity.

The AI-supporting self-driving car continues to advance but it is already on a par with human capabilities. As AI adoption is swiftly gaining much momentum, the next time there is a viral pandemic, AI will play a major role in assessing how various tradeoffs should be managed. Here in South Africa, we urgently need such objectivity to realistically prioritise our economic goals by more precisely appreciating the tradeoffs involved.

Humans aren’t great at driving cars but our capacity to process visual inputs has proven difficult to replicate. Conversely, AI is vastly superior to humans at processing the meta data a modern pandemic – or economy – produces. Yet a key takeaway from the Covid pandemic should be that almost all societies failed to manage complex tradeoffs like closing schools to slow the spread of the virus versus diminished child development.

Open societies

Managing tradeoffs is what effective political systems are designed to do. But such processes require access to the relevant facts. Legal systems place high reliance on rules of evidence being heeded. Similarly, open societies rely on media organisations to competently curate the news to adequately inform public debates.

Notwithstanding copious communication channels, the Covid crisis did not spur governments – or societies – to respond as ‘learning organisations’. The arrival of the internet and then social media has been extremely disruptive for news organisations. Many found that their survival required bonding with their audiences by shaping their reporting to suit shared narratives.

The thumbs up ‘like’ symbol, in effect, symbolises the desire by audiences to have their preferred narratives validated. Maximising news media revenue is consistent with maximising partisanship and promoting outrage. This fuels political polarisation which is contrary to preparing societies to efficiently manage tradeoffs.

Over a million Americans died from the pandemic, and its lockdowns were managed not by the national government but at the state and local level. This means there is a tremendous volume of data regarding key tradeoffs – for instance, excess deaths versus scholastic outcomes. But processing that data remains elusive because, in the parlance of US politics, there are Republican ‘red narratives’ and Democratic ‘blue narratives’.

How excess deaths should be balanced against child development is not obvious. AI doesn’t have the answers but it can counter, to a significant degree, the political shaping of news and this can make such balancing much more manageable.

One person who seems to appreciate how media partisanship is undermining politics is Elon Musk. He seeks to have Twitter become a town square conducive to open debates. The challenges are huge and the project has progressed unevenly. But his long having been a major player in the world of AI would have been integral to his vision for Twitter. It is far from outrageous to presume that AI can counteract media herding.

Robust media scrutiny

As someone who has been writing articles about South Africa’s politics and economics for over twenty years, I am well aware that it wasn’t until Mandela had died and the Gupta emails were leaked that media houses stopped shielding the ANC. While it is easy to recognise that partisan reporting and the pursuit of likes has fueled rampant polarisation in the US, our 1990s transition also inhibited robust media scrutiny.

US media houses use bias-defined playbooks to appeal to either Democrats or Republicans and this exacerbates the challenges of bridging differences. Here in South Africa the 1990s transition provoked faith in this country’s exceptionalism and its ability to prosper by coming together.

This country is exceptional but prosperity, in terms of adequate employment and broad economic growth, requires far greater global integration – most particularly with the West. AI can help transcend the barriers to our appreciating this. This can’t happen fast enough.